A $1.2B Penalty and a FICC Air Pocket Break the Beat Streak, but the Spread Engine Gets a 2027 Guide Above Range: Maintaining Outperform, Fair Value Cut to $55–66

Results vs. Consensus

The quarter split cleanly into two halves. Every line tied to lending, funding and credit met or beat the bar. Every line tied to capital-markets activity missed it, the capital ratios slipped on AOCI, and the regulatory charge sat on top. Consensus below reconciles five sources: the EPS bar clustered at $1.10–1.11 and we use $1.10; the GAAP revenue bar clustered at $30.58–30.62B and we use ~$30.6B. For the two markets lines we grade against management’s own mid-September guide, which is the bar the stock had already traded down on.

Q3 2026 Scorecard

Year-over-Year Comparison

Sequential Comparison (vs. Q2 2026)

The sequential view shows exactly where the quarter went. Revenue fell $1.6B. Global Markets alone lost $2.2B of revenue, so everything outside it grew by roughly $550M.

Revenue

Revenue of $29.96B rose 3% and missed by roughly $640M. NII contributed $1.22B of growth; noninterest income subtracted $0.30B. Inside the fee lines the story is two-sided. Asset management fees grew 17% to $4.63B on market levels and $20B of AUM flows. Market-making revenue fell $843M year over year and $1.82B sequentially, and investment banking fees fell 11% to $1.80B on weaker advisory (−19%) and debt underwriting (−14%). Card income fell 4% and missed by about $80M. Outside Global Markets, revenue grew 5.6% year over year, which is the clearest single read on the franchise’s underlying growth.

Assessment: The revenue miss is concentrated almost entirely in the line we had already told readers not to extrapolate. In July we wrote that a record trading quarter is not a base. October is what the absence of one looks like. The 5.6% growth outside markets is decent, not exceptional, and it is lower than the 7–8% that the FY27 NII guide implicitly needs from the banking book.

Expenses and Operating Leverage

Reported expense was $19.7B. Excluding the $1.2B charge, adjusted expense was $18.51B, up 6.8% year over year and down 0.6% sequentially. Against revenue growth of 3.2%, adjusted operating leverage was roughly −360bps, after +290bps in Q1 and +660bps in Q2. The efficiency ratio excluding the charge rose to 62% from 58.7% in Q2, back outside the 55–59% target band it had just entered. Global Markets is the mechanical reason: its expense rose 7.7% while its revenue fell 5.8%, because the variable costs tied to record prime-brokerage activity in equities do not fall when FICC client activity stops.

Assessment: The Q2 guide raise to 300–400bps was built on revenue outrunning a cost base that grows with it. We said at the time that the mechanism “inverts quickly” if revenue stops compounding. It inverted in one quarter. The cost base itself is behaving: adjusted expense has held near $18.5B for three quarters and headcount fell by 3,473 year over year. The problem is the revenue line, not cost discipline.

EPS

Reported EPS of $0.91 fell 13%; adjusted EPS of $1.08 rose 4% against $1.04 a year ago. Adjusted net income actually fell 2% year over year, to $8.1B from $8.3B, so the entire adjusted EPS increase came from a 5.3% lower diluted share count. The reported tax rate of 23% reflects the non-deductible charge; excluding it, the rate was about 20%, in line with the full-year framing and resolving the 21.5% question we carried out of Q2.

Assessment: This is the first quarter in our coverage in which buybacks did all of the EPS work. That is not a quality earnings outcome. It is also not a structural one: Global Markets net income fell $413M year over year, while the other three segments together grew net income by $264M. The earnings base below the markets line still grew; it did not grow enough to cover a normal trading quarter and a compliance bill at the same time.

Segment Performance

Global Markets was the only segment whose revenue declined year over year. But only Wealth produced positive operating leverage, and the markets reversal turned the segment mix inside out compared with Q2.

Segment Results — Q3 2026

Segment revenue is reported on an FTE basis and sums to $30.12B; the GAAP total is $29.96B. Excluding the charge, All Other lost $153M against $292M in Q2 and $103M a year ago.

Global Markets Revenue Mix — Q3 2026

Global Markets: The Reversion, Arriving Through FICC

Revenue of $5.87B fell 6% and net income of $1.24B fell 25%, with the efficiency ratio at 72% and return on allocated capital at 9%. The decline came almost entirely from the micro side of FICC. Mortgages, municipals and credit are the franchise’s largest fixed-income businesses, and a quarter in which rates rose 85bps left clients unwilling to trade them. FX and commodities grew, and the macro businesses rose 4%, but not enough to offset. Equities, the line we worried about in July, held up better than FICC: a record third quarter at $2.51B (+10%) on client financing. Average trading-related assets still rose 8% year over year to $729B, which means the balance sheet was there and the clients were not.

“Equities delivered a record third quarter, and reduced client activity and changing market conditions, including a major shift in interest rates, significantly affected our leading FICC micro businesses, including credit mortgages and municipals. Those declines more than offset growth elsewhere in trading.”
— Alastair Borthwick, CFO

Global Markets fell to 19.5% of FTE revenue from 25.3% in Q2, and to about 15% of segment net income before All Other from about 28%. Year to date the segment is still at a record (revenue +11%, net income +14%), and sales and trading of $18.5B is up 13%.

Assessment: The bear pillar we opened in July said a normal reversion in markets would consume the operating-leverage raise. It did, and faster than we expected, and through FICC rather than equities. The franchise is not broken: year-to-date numbers are records and the trading balance sheet is larger. But the quarter settles the question we asked in July. Q2 was a peak, not a plateau. We now model sales and trading at $5.0–5.5B a quarter, not the $6.5B+ that a plateau would have required.

Global Wealth & Investment Management: Still the Cleanest Compounder

Revenue of $7.12B (+13%) and net income of $1.54B (+22%), with a 29% pretax margin, up from 27%, and return on allocated capital of 28%. Client balances passed $5T for the first time at $5.06T (+9%), AUM rose 12% to about $2.4T, AUM flows were $20B in the quarter and $74B over four quarters, and average loans rose 13% to $279B on securities-based and custom lending. The segment added about 5,000 net new $500K+ relationships across Merrill and the Private Bank.

“This business continued to improve profitability through strong client flows, positive operating leverage, and higher margins.”
— Alastair Borthwick, CFO

Assessment: Wealth was the only segment to produce strongly positive operating leverage (revenue +13%, expense +9%), and its 29% pretax margin is now close to the 30% Investor Day target that we called conservative in July. The segment is market-sensitive through asset-management fees, so it is not a hedge against a risk-off quarter. It is, however, the best evidence in the print that the franchise compounds when markets are merely normal.

Global Banking: Lending and Treasury Up, Investment Banking Down

Revenue of $6.21B was flat and net income of $1.99B fell 5%, with return on allocated capital at 15%. Investment banking fees for the whole corporation were $1.80B, down 11%, at the top of the $1.6–1.8B range the CEO gave in mid-September. Advisory fell 19% and debt underwriting fell 14%; equity underwriting rose 9%. The rest of the segment grew: Global Transaction Services revenue rose 11% to $3.04B, treasury service charges rose 10%, average deposits rose 6% to $669B and average loans rose 8% to $419B.

“Total company investment banking fees, excluding self-led transactions, were about $1.8 billion. Fees declined from a strong year-ago period, driven by lower advisory and debt underwriting fees. Looking forward, our pipelines remained healthy.”
— Alastair Borthwick, CFO

Assessment: In July we said the level mattered more than the growth rate and put $2.0–2.1B a quarter in the model. The level fell to $1.8B. Year to date, fees are $5.78B and up 16%, so this is a quarter, not a cycle. The treasury business is the part of Global Banking that compounds without regard to deal flow, and an 11% revenue gain there is the line we would point to. We move the IB model to $1.7–2.0B a quarter.

Consumer Banking: Steady, With the Positive-Leverage Streak Paused

Revenue of $11.57B (+4%) and net income of $3.52B (+2%), with a 50% efficiency ratio and a 31% return on allocated capital. Expense rose 4.6% against revenue growth of 3.6%, which ends the five-quarter run of positive operating leverage the segment reported in Q2. Management attributed the gap to marketing that stepped up in Q4 2025 for sponsorships and the BofA Rewards relaunch. The segment added 200,000+ net new checking accounts, opened more than 1 million new card accounts, and grew average card balances nearly 5% to $105.8B. Combined card spend rose 8% to $266B. The risk-adjusted margin on card fell to 6.74% from 7.48% a year ago.

“Year-over-year expense comparisons are tougher because we increased marketing beginning in the fourth quarter of 2025, ahead of new sponsorships and the relaunch of BofA Rewards, with spending steady for the past four quarters.”
— Alastair Borthwick, CFO

Assessment: The card scorecard management volunteered in Q2 (1%, 2%, 3%, 4%, toward 5%) reached “nearly 5%.” That is a promise kept. The cost is visible in a lower card risk-adjusted margin and a pause in segment operating leverage, both of which should ease as the marketing comparison laps in Q4. Consumer deposits were flat year over year at $949B, which is the soft spot we take up under Key Topics.

All Other: The Charge, and a Smaller Underlying Loss

All Other lost $1.35B, including the $1.2B non-deductible settlement charge. Excluding it, the loss was $153M, narrower than Q2’s $292M, which we had flagged as unexplained. The segment also carries the $11B of higher-cost deposits that management deliberately ran off over the past year.

Key KPIs

Key Topics & Management Commentary

Overall Management Tone: Management was more defensive than at any of the five earlier calls we have covered, and it was defensive in a controlled way: the settlement was handled with a scripted three-point statement, and the misses were explained as mix rather than franchise. The forward-looking content was confident and specific on NII, where a 2027 guide above the medium-term range was offered unprompted, and noticeably thinner on operating leverage and capital, where the answers deferred to the January guide and to final capital rules. The shift from the relaxed, guide-raising posture of July was obvious, and the call ran long on capital and compliance questions that the July call never had to field.

1. Grading the Ten Q2 Commitments

In July we left the quarter with a ten-item checklist. Five were delivered, two partly, two missed, and one is still pending. The two misses are the ones that moved the stock.

Assessment: The scorecard says the spread, credit and funding promises held and the markets-dependent promises did not. That is consistent with how we framed the risk in July, which is some comfort. It also says that the operating-leverage raise we called “the single most important forward-looking disclosure on the call” three months ago is gone, and we have to own having put weight on it.

2. The $1.2 Billion Penalty: Willful Violations, and a Lookback Still Ahead

On October 13 the OCC and FinCEN announced consent orders covering the bank’s Bank Secrecy Act, anti-money-laundering and sanctions programs. FinCEN found the violations willful, and the bank admitted them: monitoring and due-diligence gaps meant that more than $5B of suspicious activity went unreported in time, including activity tied to money-laundering networks and the southwest border. The OCC’s $750M civil money penalty is credited against FinCEN’s $1.2B. The orders require a third-party lookback for unreported suspicious transactions and independent reviews of the monitoring and data programs. This resolves the penalty attached to the OCC’s December 2024 cease-and-desist order; it does not close the remediation that order began.

“We understand the seriousness of compliance, and we are responsible for getting it right. Second, we have invested billions of dollars in technology and dedicated thousands of employees to combat financial crime. The team is far along with and committed to completing the work required to be addressed by the regulators. Third, we have committed the resources, and the ongoing cost remediation is embedded in our company's expense base.”
— Brian Moynihan, CEO

Management drew two lines under the matter. There is no asset cap and no constraint on business activity. And the remediation cost, a couple of hundred million dollars in Q3 combined with the software-security work, is already inside an adjusted expense base that has held near $18.5B for three quarters.

Assessment: The dollar amount is absorbable: $1.2B is about 15% of one quarter’s adjusted earnings and 0.6% of CET1 capital. The finding is the harder part. A willful violation, admitted, at the second-largest US bank, more than a year after a cease-and-desist order, is a governance fact rather than a cost fact. As one question on the call pointed out, management said in January 2025 that the work was in the run-rate, and a $1.2B charge followed. We believe the run-rate claim on expenses, because the expense line supports it. We do not assume the lookback is clean, because nobody can know that yet. This opens Bear #7.

3. Markets Missed Management’s Own September Number

On September 14, at an industry conference, the CEO guided Q3 sales and trading to roughly flat year over year, about $5.4B, and investment banking fees to $1.6–1.8B. The stock fell more than 5% that day. Two days later the Federal Reserve raised rates for the first time since 2023. Sales and trading came in at $5.06B excluding DVA, about $350M below the guide given with only two weeks left in the quarter. Investment banking landed at the top of its range.

“Secondly, the markets activity, because of our strong business mix towards fixed income, hit us. We still made a lot of money trading this quarter. We made it very few days with low revenues. We feel very good about it.”
— Brian Moynihan, CEO

Management did not reconcile the result to its own mid-quarter guide, and the explanation offered was the 85bps rate move hitting its strongest micro-FICC businesses. For Q4, the CFO declined to guide but said the extreme conditions have moderated in the first weeks of October and that seasonality may be smaller than usual.

Assessment: Missing a guide set in the last third of the quarter is more informative than missing a consensus. It says the September back half was worse than management could see in mid-month, which fits a rate shock, not a franchise problem. It also tells us how much of the trading line is mix exposure the bank cannot hedge: the strength in mortgages, munis and credit that drove the 17-quarter streak is the same concentration that drove this quarter. Management says it is building rates, FX and commodities to balance it. That is a multi-year project, not a Q4 offset.

4. NII: The Pillar That Held, and a 2027 Guide Above the Range

NII on an FTE basis was $16.61B, up 8% year over year and $450M sequentially, and it beat the $16.55B consensus. The net interest yield rose 8bps to 2.16%; excluding Global Markets it rose 8bps to 2.65%. Sequential drivers were one more day, fixed-rate asset repricing, loan growth and Global Markets activity. Management held the FY26 guide at about 8% and, unprompted, gave a first look at 2027.

“We still expect to hit that full-year revised NII growth guidance of approximately 8%. For 2027, we currently expect full-year growth of NII to be 8% or more, higher than 2026, and again, higher than our original 5%-7% range.”
— Alastair Borthwick, CFO

The history matters here. At Investor Day the medium-term NII growth framework was 5–7%. FY26 was raised twice during the year, and FY27 now starts above the top of the original range. The CEO added the yield target trajectory: “this march from 2.15% towards 2.30% really just taking place as we look out over the next several quarters.”

Assessment: This is the most important disclosure of the quarter and the reason the rating holds. NII is roughly 55% of revenue and the line on which we upgraded the stock in January. It grew 8–9% year over year in every quarter of 2026, and the first look at 2027 is a raise rather than a fade. The open question is what the guide assumes, which is the next topic.

5. What the 2027 Guide Asks of the Core Book

Two assumptions sit underneath “8% or more.” The first is rates. The outlook assumes three further 25bps hikes, in December, March and June, after September’s. The disclosed sensitivity is now asymmetric: +100bps above the forward curve adds $1.4B of NII over twelve months, and −100bps subtracts $1.9B. The second is mix. Global Markets NII has been flat at about $1.9–2.0B for three quarters and is liability-sensitive, so management expects it could fall if rates rise.

“Yeah. To be precise, the 8% includes effectively a drag on markets because it hasn't been growing for the last couple of quarters. It's been relatively flat. So it's an all-in number overcoming that, just to be clear.”
— Brian Moynihan, CEO

The arithmetic follows. Global Markets NII grew from $1.5B to $2.0B over the past year, which is roughly $0.5B of the $1.2B of total NII growth. NII excluding Global Markets grew about 5%, from $13.9B to $14.6B. For total NII to grow 8% or more in 2027 with markets NII flat or lower, the core banking book has to accelerate from about 5% to something above 8%. Management agreed with that direction under questioning, without giving a number.

Assessment: The guide is credible in its drivers (fixed-rate repricing, roughly $20B a quarter of loan growth, noninterest-bearing deposits growing 4–5%) and generous in its rate assumption. If the three hikes do not come, the disclosed sensitivity implies roughly $1.4B less NII over twelve months, about two points of growth and roughly $0.16 of EPS. Our FY27 base uses 8% exactly, the floor of the guide, and our low case removes most of the hikes.

6. Operating Leverage: The 300–400bps Guide Quietly Left the Room

In July management raised the FY26 operating-leverage guide to 300–400bps after a first half above 450bps. Q3 delivered about −360bps excluding the charge, leaving roughly 190bps for nine months. Reaching even 300bps for the year would need Q4 adjusted expense of about $17.5B on ~$30.3B of revenue, roughly $1B below the Q3 level. On the call the full-year number was not mentioned. The replacement language was a quarter and a medium-term range.

“No. No change there. We've obviously had a very good first half. Q3 was slower, and we're back on creating operating leverage in Q4. I think as we get into Q4, Chris, I will likely provide guidance for next year, but it's not going to change from the 200-300 basis points. Based on what we see, it could be higher. But we'll see as we get closer.”
— Alastair Borthwick, CFO

Assessment: A guide raised in July and abandoned in October without being named is the clearest credibility cost of the quarter, more so than the penalty. We said in July that a self-healing expense framework is “comfortable for management and less useful for underwriters,” and this is what that looks like in practice. We take 200–300bps for 2027 as the working range, model 250bps, and wait for the January guide before giving credit for “could be higher.”

7. Expenses, AI and Mythos: Where the $1.2B of Growth Went

Asked about AI productivity, the CEO decomposed the year-over-year expense increase in unusual detail. Of roughly $1.2B, about $450M was wealth-management incentive pay tied to revenue, about $300M was brokerage, clearing and exchange costs in markets, and of the remaining ~$400M, a couple of hundred million was remediation: AML work, plus work on what he called “the Mythos implementation,” for which he said the bank brought in about 1,000 consultants. That program patches software vulnerabilities found with advanced AI models. Underlying core expense growth, by this account, was 1–2%.

“AI has helped us. We are about 130 different projects implemented. All of them have a payback, and they are very detailed projects. None of them are huge contributors. Overall, the benefits are 2x-3x over time of the expense you are giving out. We expect that number to go from the $250, $300 this year to $700 million next year in terms of projects we put in the pipeline.”
— Brian Moynihan, CEO

Assessment: In July we wrote that AI would move from supporting the operating-leverage case to driving it when management attached a dollar figure to the savings. It has now done so: $250–300M this year, about $700M next. That is a real milestone and also a modest number, under 1% of a ~$74B adjusted expense base. The more useful disclosure is the decomposition itself. If 1–2% is the true core growth rate, the revenue-linked and remediation pieces explain the rest, and the remediation piece should fade.

8. Capital: CET1 Down 37bps, the Buyback Held, and Two Constraints Ahead

CET1 fell to 10.9% from 11.2%. CET1 capital fell $3.2B to $198.4B and risk-weighted assets rose $32B to $1.82T. The causes were three: $34B of ending loan growth, the penalty, and a $1.8B AOCI hit on available-for-sale securities as rates rose. Total AOCI fell $3.2B; a little less than half of that was cash-flow hedges, which do not count against CET1. The tangible common equity ratio held at 6.0%. The bank still bought back $6.0B and raised the dividend, returning $8.3B against $6.9B of reported net income.

“If you look at this quarter, for example, Chris, we supported $34 billion in loan growth. We obviously had a regulatory penalty. We saw rates go up 85 basis points. We paid an increased dividend, and we bought back $6 billion in shares. So we've got a lot of flexibility with the earnings that we generate to balance exactly what we've been doing through the course of this year and over the course of time.”
— Alastair Borthwick, CFO

Two constraints now frame the next few quarters. The regulatory minimum is 10%, and an analyst noted that a G-SIB surcharge step-up would lift it to 10.5% on January 1 unless the pending reform changes how the surcharge is calculated. The bank also manages to a tangible common equity ratio around 6%, which the CEO said can drift. Management expects final capital rules in Q4 and declined to say anything about buybacks until it sees them.

Assessment: The July worry was a buyback that slowed without explanation. October brought an explanation and a flat buyback through a quarter that contained a penalty, a rate shock and record loan growth. That is a genuinely good capital outcome. The pillar stays at risk for a different reason: if the January step-up applies, the cushion above the minimum shrinks to about 40bps, below the ~50bps buffer management described in April as its target. From here the buyback is set by the final rules, not by earnings.

9. Investment Banking: Top of the Range, and Pipelines Called Strong

Fees of $1.80B were down 11% year over year and 16% sequentially, at the top of the September range. Advisory was the weakest line at $471M (−19%), and debt underwriting fell 14% to $958M. Year to date, fees are $5.78B and up 16%. Management attributed the slowdown to the rate move and said the bank ranked third in announced M&A in the quarter.

“Obviously, there's a period in volatility like that where buyers and sellers just need to readjust valuations and expectations. That always happens. But if you look at our businesses generally, we can see from the loan growth activity across the commercial side that the CapEx cycle continues. We know from the investment banking pipelines that they remain strong, and our team added more in the way of pipeline in the quarter.”
— Alastair Borthwick, CFO

Assessment: Announced-M&A rank and pipeline growth are the right leading indicators, and both point to recovery. The CFO warned in July that the Q3 2025 comparison ($2.0B) would be hard, and it was. We hold IB at $1.7–2.0B a quarter. That is a lower plateau than the July model and a higher one than the FY25 base.

10. Loans: $34 Billion in a Quarter, Commercial in Every Segment

Ending loans rose $34B sequentially to $1.25T, and average loans rose 6% to $1.23T, the tenth consecutive quarter of sequential average growth. Commercial loans grew 8% with growth in every line: small business +2%, business banking +9%, global commercial +6%, global corporate +10%, and Global Banking international commercial +15%. GWIM custom lending rose 22% and securities-based lending 21%. Card grew nearly 5%; auto balances fell.

“The really core aspects, meaning middle market grew across the board in most of its segments. The business banking franchise, which the team's done a good job, Sharon Miller team done a good job getting to grow again, it's balanced. The small business grows, the credit card grew. They may come in ebbs and flows, but generally they're stick to your ribs type growth.”
— Brian Moynihan, CEO

Assessment: Loan growth is the input to the 2027 NII guide we can verify quarter by quarter, and it accelerated into the end of the quarter. The quarter-end jump flatters the average less than it flatters the spot balance, and part of it is volatile warehouse and bridge lending. Even so, the middle-market and business-banking growth is the durable kind. Combined with about $36B a year of mortgage-related runoff from the securities book, it is the mechanism by which the yield climbs toward 2.30%.

11. Deposits: The Sequential Streak Ends, and the Rate Paid Ticks Up

Average deposits were $2.022T, flat sequentially and up $31B (2%) year over year. The release, which in Q2 cited a twelfth consecutive quarter of sequential average-deposit growth, made no streak claim this time. Part of the flat quarter is deliberate: $11B of higher-cost deposits in All Other ran off over the year. Noninterest-bearing deposits grew 4.7% year over year and made up more than 75% of the year’s growth. The total rate paid rose 3bps to 1.49% after the September hike, with consumer at 0.51%. Ending deposits rose $14B.

“Our deposit mix remained attractive, with non-interest-bearing deposits accounting for more than 75% of year-over-year deposit growth. We also remain disciplined on pricing with the total deposit rate rising just 3 basis points to 1.49%.”
— Alastair Borthwick, CFO

Management also addressed agentic AI as a channel for deposit flight. Its view is that operating balances serve payments and are not idle cash an agent can optimize away, and that investment cash already sits in money-market alternatives the bank also offers.

Assessment: The pillar holds, but this is the softest deposit quarter of the six we have covered: consumer deposits were flat year over year, the average was flat sequentially, and the rate paid moved up after the hike. None of that breaks the mix argument, which is the one that matters. It does mean the “compounds through cycle” line in our thesis needs a footnote for a hiking cycle.

12. Credit: Criticized Loans at the Lowest Rate Since 2019

Provision was $1.27B, net charge-offs $1.35B (0.44%), and the reserve release $80M. The card charge-off rate fell to 3.22% from 3.55%, and card delinquencies improved year over year for the sixth consecutive quarter. Commercial net charge-offs were $352M, and only $220M excluding small business, a 0.12% rate. Reservable criticized commercial exposure fell $2.2B to $19.9B, a 2.49% rate, the lowest since 4Q19. Nonperforming loans fell $207M to $5.54B, about $920M below consensus. The CRE office cleanup is, in the CFO’s words, at “the very tail end.”

“In terms of 2027, I would say this res crit number is pretty low relative to even pre-COVID numbers. I think this is the lowest we have had since COVID. So can it improve? Yes. But if you look at the charge-offs today in commercial, they are really, really low at $220 million for the quarter.”
— Alastair Borthwick, CFO

Assessment: Credit is the cleanest part of the quarter and the reason the miss stayed small. The one number to watch is that consumer 30-day-plus delinquencies on non-insured loans rose about 3% year over year, to $4.17B, which is modest and mostly seasonal. The “isolated corporate and commercial lending losses” language from July did not return, and commercial charge-offs fell. The CFO’s own framing, that it is hard to improve much from here, is the honest one: credit is a source of stability, not of further upside.

13. ROTCE: Back on the Original Glide Path, Not Ahead of It

ROTCE was 12.6% reported and 14.9% excluding the charge, against 17.0% in Q2 and 15.3% a year ago. Year to date it is 15.2% reported, or roughly 16% excluding the charge. The glide path set in January was 16% within eight to twelve quarters and 18% by year three.

“That's a really important dynamic because if you go back to the 6% TCE ratio, that gives you a chance to really push the returns on a relative size that gets you the 16%, et cetera. So that's what we've been working at, is that efficiency balance sheet because of the dynamics we are in as liquidity.”
— Brian Moynihan, CEO

Assessment: In July we said ROTCE was running about two years ahead of schedule. That was a markets-quarter reading, and we withdraw it. A normal markets quarter puts adjusted ROTCE near 15%, which is where the original glide path would have it at this point. The pillar stays on track. What changed is the starting point for 2027: we model 15.5–16.5%, not 17%+.

Guidance & Outlook

Implied Q4. An FY26 NII growth rate of 8% on FY25 FTE NII of $60.7B implies a Q4 of roughly $16.9B, up ~$270M sequentially and ~6% year over year. That is consistent with the “improve sequentially” language and with the fixed-rate repricing run-rate. On operating leverage, a positive Q4 would leave the full year at roughly 150–200bps excluding the charge, about half the July guide.

Street positioning. The questioning on the call was not about whether NII would arrive. It was about whether expenses, compliance and capital rules would let it reach the bottom line. Our FY27 estimate is built the same way: the NII line is close to management’s floor, and the haircut is in operating leverage and the multiple.

Guidance style. The NII guide continues a pattern of being set conservatively and walked up: 5–7%, then 6–8%, then the top of 6–8%, and now 8%+ for the following year. The operating-leverage guide broke that pattern. Management raised it on a record trading half and let it lapse when trading normalized. We would now treat NII guidance as a floor and fee-dependent guidance as a midpoint at best.

Analyst Q&A Highlights

Whether the Consent Order Constrains the Business

The first compliance question was the one that matters for valuation: does the order limit what the bank can do? The answer was unambiguous. There is no asset cap, nothing will be crowded out, and the people and technology required are already in place.

Q: “Are there any constraints on the business? Does it impede your normal business operations at all?”
— David Chiaverini, Jefferies

A: “No, there is no asset cap here, so no, there is nothing there. In terms of it will not crowd out anything either because we have invested the technology and we have invested the people that we feel like we need to get the work done. So that is in the run rate. We have made the decisions we need to. We have added the resources we feel like we need. So it should not impact our clients in any way.”
— Alastair Borthwick, CFO

Assessment: This is the answer that separates this order from the kind of growth restriction that can take years to lift. No asset cap means the loan growth and the NII guide are unaffected. The risk that remains is cost and reputation, not capacity.

Whether the Remediation Cost Is Really in the Run Rate

The sharpest exchange of the call put management’s January 2025 statement, that the AML work was already in the run-rate, back to it after a $1.2B charge. The CEO answered with the expense line and the resolution, not with a timeline for the lookback.

Q: “Brian, it is obviously a surprise when you talked on the January 2025 call, you talked about how everything was already in the run rate for fixing this issue, the AML issue. Yet clearly you had a huge charge this morning. It is not that you cannot absorb $1.2 billion, but obviously it is a reputational and confidence issue. What should we think about? How should we get confidence that this is really in the run rate?”
— Multiple analysts incl. Vivek Juneja, JPMorgan; John McDonald, Truist Securities; Erika Najarian, UBS

A: “We just resolved the timely aspect of that, which is in the order. If you look at the cost across all the quarters, basically, $18.5-ish billion is the last three quarters, $18.6 billion down to $18.5 billion, and we're telling you it's relatively flat. It's in the run rate. It's thousands of people. We're repositioning people from other units as we get the efficiencies to this, and then we'll let the headcount drain down. It's in the run rate.”
— Brian Moynihan, CEO

Assessment: On operating cost, the evidence supports management: three flat quarters of adjusted expense with the remediation inside them. On the question actually asked, how investors know this is behind them, there was no milestone, no lookback timeline and no range. The distinction between a penalty, now paid, and a remediation, now ongoing, is the one the market is pricing.

Whether Operating Leverage Survives the Expense Line

A recurring line of questioning pressed on whether 2027 operating leverage is still credible after a negative Q3. Management held the 200–300bps medium-term range and deferred specifics to the January guide. No one asked directly about the FY26 300–400bps guide, and management did not mention it.

Q: “Given that expenses have come in higher than expected, how would you instill confidence in investors that you could do the high end of 300 basis points of operating leverage next year, much less 200-300 basis points?”
— Multiple analysts incl. Erika Najarian, UBS; Chris McGratty, KBW; Manan Gosalia, Morgan Stanley

A: “The expense side, I think, still holds together there. The operating leverage, we've said we intend to deliver 200-300 basis points through the medium term. We haven't changed from that. I think as we get into Q4, I'll give you a better sense for what that looks like in 2027.”
— Alastair Borthwick, CFO

Assessment: A holding answer. The range is defensible because NII drops largely to the bottom line, as management noted in a separate answer. But the market has just watched a raised range lapse in one quarter, so the January guide now carries more weight than usual. We give no credit for “could be higher” until it is in writing.

What “8% or More” Means for the Core Banking Book

The most constructive exchange of the call. Analysts worked out that if markets NII falls in a hiking cycle, the 8%+ guide implies faster growth in the banking book, and asked management to confirm it. Management would not split the guide but confirmed the direction.

Q: “Number one, just a clarification question on NII ex markets. Alastair Borthwick, given your comments on markets NII being down, I would essentially assume that core NII would be up around 10% year-over-year in 2027, in line with what JPMorgan said yesterday. Is that about the right geography?”
— Multiple analysts incl. Erika Najarian, UBS; Jim Mitchell, Seaport Global Securities; David Chiaverini, Jefferies

A: “Well, look, we do not guide the two differently and independently for a variety of different reasons. ... So that 8% is including markets. So that would tell you that the number is likely higher for the core banking book. But we do not break it out because things move around, and they go between, and we manage it for the entire company. So I think directionally, that is how you should think about it.”
— Alastair Borthwick, CFO

Assessment: “Directionally” is as close to a core-NII guide as this team will get, and it is a strong one. The core book grew about 5% in Q3, so management is effectively guiding it to well above 8% in 2027. That is ambitious, and it is the right place to focus: it is the single assumption on which our FY27 estimate depends most.

The G-SIB Step-Up and the $6 Billion Buyback

The capital question that matters for the next two quarters: whether a scheduled surcharge increase in January, together with a tangible-equity constraint, forces the buyback below $6B. Management declined to guide and tied everything to the final rules.

Q: “Finally, on the capital on January 1st, 2027, you're set to step up a GSIB surcharge ... Or are we waiting for the finalization of GSIB to see whether or not they'll treat your GSIB surcharge calculation only prospectively or retroactively to prevent that step up?”
— Erika Najarian, UBS

A: “We don't guide on share buyback for lots of different reasons. It's definitely business as usual in terms of how we think about that equation with one exception. We need to wait for the final rules. I know we're all desperate for the final rules. I'm desperate for the final rules. But we don't have them yet.”
— Alastair Borthwick, CFO

Assessment: “Business as usual” with “one exception” is honest, and the exception is large. A step-up to a 10.5% minimum would leave about 40bps of CET1 headroom at the current 10.9%. We model the buyback at $5–6B a quarter until the rules are final, which makes this the biggest single swing factor in the FY27 share count.

Whether a 6% TCE Target Is Shaping the Balance Sheet

The inter-quarter speculation was that the bank had shrunk the markets balance sheet to protect a 6% tangible common equity ratio, costing trading revenue. The CEO rejected the link outright and, separately, described the 6% as a zone rather than a line: it “may drift down with outsized growth in the quarter.”

Q: “There was a lot of speculation inter-quarter that efforts to manage to a 6% TCE may have weighed on markets performance in 3Q, as evidenced by the balance sheet contraction in the segment. I was hoping you could speak to that linkage of how managing to a 6% TCE target informs your willingness to deploy more balance sheet in the markets business or frankly, any other areas that are just more leverage versus RWA intensive.”
— Multiple analysts incl. Steven Chubak, Wolfe Research; Vivek Juneja, JPMorgan

A: “It has no bearing at all. The simple proof in that is year-over-year, the markets balance sheet is up and the revenue's down. ... When customers stop cold like they did last quarter, that happens.”
— Brian Moynihan, CEO

Assessment: The data supports the answer. Average trading-related assets rose 8% year over year while revenue fell, so the business had the balance sheet and lacked client flow. That matters, because a capital-constrained trading desk would be a structural problem; a client-activity air pocket is a cyclical one.

Deposit Pricing When Everyone Needs Funding

A thoughtful challenge: this hiking cycle differs from the last one because loan growth is strong and banks need deposits to fund it, so betas could run higher than the industry assumes. Management’s answer was the balance sheet itself: $2T of deposits against $1.2T of loans.

Q: “I guess the long-winded question is, isn't there a risk that deposit pricing picks up more than maybe you and others have assumed just because of more loan growth and therefore more deposit competition?”
— Multiple analysts incl. Matt O'Connor, Deutsche Bank; Ben Gerlinger, Citi

A: “We have $2 trillion in deposits and $1.2 trillion in loans. There is a lot of room to grow loans without thinking about needing to raise money to fund them. ... We don't see the dynamic because we have a lot of room to grow loans without having any pressure to raise funds for them.”
— Brian Moynihan, CEO

Assessment: The loan-to-deposit ratio is BAC’s structural answer, and it is a good one for BAC specifically. It is less of an answer for the industry, and if competitors pay up for deposits, BAC’s wealth and corporate depositors, who already receive pass-through pricing, will notice. The 3bps rise in rate paid this quarter is the first data point. We watch it each quarter against the commitment that betas on the way up will match those on the way down.

Whether FICC Snaps Back

The question for Q4 was whether the FICC damage was marks or volume. Management said volume: clients stopped trading when rates moved, and October has not repeated September.

Q: “Maybe just on FICC trading, you noted higher rates hurt the micro businesses. Is that most negative marks or was the slowdown of volume? Just trying to see if this pressure could linger or could we see a snapback if rates stabilize.”
— Multiple analysts incl. Jim Mitchell, Seaport Global Securities; Ken Usdin, Autonomous Research; Glenn Schorr, Evercore ISI

A: “It just happens when rates go up 85 basis points in a quarter, that tends not to be good for those businesses. Clients tend to slow their activity. It's not good for asset prices. So it's largely a client activity thing. We haven't seen the same kind of a performance, meaning we haven't seen rates behave the same way in Q4.”
— Alastair Borthwick, CFO

Assessment: “Largely a client activity thing” is the better of the two answers for the stock, because activity returns and marks do not. Combined with the CFO’s comment that seasonality may be lighter than usual, Q4 sales and trading near the Q3 level is plausible. We model $4.8–5.2B and treat any recovery in mortgages and credit as upside.

Where AI Takes Productivity

Asked how far AI can push revenue per employee, the CEO gave the most detailed expense bridge we have heard from this team, then put dollar figures on AI benefits, which the July call did not.

Q: “In this period of AI and AI deployment, where can that take you guys? I see revenues per employee year-to-date are up 10%, and you would think with all the tools at your disposal that you would be able to take that much higher.”
— Mike Mayo, Wells Fargo Securities

A: “A key one of those is in the development of technology, which is $4 billion in change a year. We are 10% more efficient, so next year we get 10% more work for the same cost. We are going to continue to invest in driving the growth in the franchise, and AI is a major part of that. Say, $300 million going to $700 million, headcount down by 3,000, 1.5% year-over-year.”
— Brian Moynihan, CEO

Assessment: Specific, small and believable. A 10% efficiency gain on a $4B change budget is $400M of capacity, and $700M of 2027 benefits is consistent with the 1–2% core expense growth the CEO described. AI is now a quantified contributor to the cost line, but not yet one large enough to change the operating-leverage range on its own.

What They’re NOT Saying

1. The FY26 operating-leverage guide. Raised to 300–400bps in July and not mentioned once on this call. The year-to-date figure is roughly 190bps excluding the charge. Management neither confirmed nor withdrew the range; it talked about Q4 and the medium term instead.
2. Why trading missed the September guide. The CEO guided sales and trading to roughly flat with two weeks left in the quarter, and the result was about $350M lower. The call explained the rate move but did not reconcile the result to that guide, and no analyst asked.
3. The size or timing of the lookback. The consent orders require a third-party review of unreported suspicious transactions. Asked whether that review brings new costs, management repeated that resources are in the run-rate. There was no timeline for the review and no range for what it might find.
4. The cost of the Mythos remediation on its own. The CEO combined it with AML remediation as “a couple hundred million dollars plus” and said it should run down. The program’s own cost and end date were not given.
5. The downside rate sensitivity. The CFO cited +$1.4B of NII for +100bps. The deck also shows ($1.9B) for −100bps, which matters more for a guide that assumes three more hikes. It was not discussed.
6. A core-NII guide. Asked twice, management again declined to split the guide between the banking book and markets, even though the deck already sizes markets NII each quarter.
7. The deposit streak. Q2’s release counted a twelfth consecutive quarter of sequential average-deposit growth. This release counted loan growth instead. Average deposits fell $0.9B sequentially, and the change was not mentioned.
8. The G-SIB step-up, sized. An analyst stated that the minimum rises to 10.5% on January 1. Management neither confirmed nor sized the effect on buybacks, and deferred to final rules.
9. Succession. Asked how long the CEO intends to stay in his 17th year, the answer was that it is the board’s decision. There was no timeline and no process description.
10. Berkshire. Unmentioned by management and unasked by analysts for a fifth consecutive quarter.

Market Reaction

• Pre-print setup: BAC closed at $53.96 on October 13, the day the penalty was announced, down 1.1% on 1.2x volume in a weak session for bank stocks. It entered the print −1.9% year to date (S&P 500 +14.0%), −9.3% over 30 days and +7.7% over twelve months. The stock had fallen 11.0% since our Q2 report (S&P 500 +3.5%, KBE bank index −8.5%) and sat 16.7% below its August 12 closing high of $64.81. The two largest steps down were September 14 (−5.1%, the day of the conference guide) and September 16 (−2.7%, the Fed hike).
• Reaction session (October 14, before-the-open report): The stock opened at $49.83, a 7.7% gap down, and touched $49.26 (−8.7%) before recovering 3.1% from the open to close at $51.39, down 4.8% (−$2.57).
• Volume: 118.0M shares against a 37.0M 30-day average, or 3.2x. In July the reaction ran at 1.3x.
• Relative: The S&P 500 fell 1.1% and the KBE bank index 0.7%. Morgan Stanley, which reported the same morning, fell 2.8%. The banks that had reported the day before were mixed: JPMorgan +1.2%, Wells Fargo +0.2%, Citigroup −1.8%, Goldman Sachs −1.5%.
• Follow-through: +0.6%, −0.9% and −1.6% over the next three sessions, to $50.39 on October 19. Cumulatively −6.6% from the pre-print close against −1.5% for the S&P 500, +0.3% for KBE, +0.6% for JPMorgan and +1.3% for Wells Fargo.

No bounce, which is the signal. Wells Fargo, which also missed on revenue this season, recovered its entire post-print loss within three sessions. BAC recovered none of it: after a 0.6% bounce on the second day, it fell in each of the next two sessions. A reaction that keeps going after the facts are known is not a reaction to the facts. It is a re-rating, and the call’s own framing, a “confidence issue,” describes what is being re-rated: management’s forecasts on markets and expenses, and the compliance function, rather than the earnings power of the bank.

The open priced the whole package; the session priced the answers. The call began at 8:30 a.m., so the 7.7% gap at the open came after the prepared remarks, including the first look at 2027 NII. The gap is therefore the market’s verdict on the full set: a miss, a penalty, a trading shortfall against a guide the market already knew, and an above-range NII guide. The 3.1% recovery from the open is consistent with the Q&A answers, which ruled out an asset cap, placed the remediation in the run-rate and confirmed that core NII should grow faster than 8%. The market accepted those answers. It did not pay for anything beyond them.

The setup was not crowded long. In July we described a fully priced stock near its 52-week high. This time the stock entered 16.7% below its high, down 11% in three months and well behind both the index and its peers. A heavy, persistent sell-off from that starting point is a statement about confidence in management’s numbers, and confidence discounts can last. They also tend to create the entry points that a valuation-based Outperform rating is meant to use.

Street Perspective

Debate: Is the Penalty the End of the Story or the Start of a Remediation Tail?

Bull view: The bull case is that the penalty resolves the matter carried since the December 2024 order. It is fully expensed, does not restrict growth, and the remediation cost is visibly inside a flat expense base. Without an asset cap, this camp argues, the penalty is a one-quarter event, and $1.2B is small against the roughly $33B a year the bank now returns to shareholders.

Bear view: The bear camp contends that admitted willful violations, more than a year into a cease-and-desist, describe a control culture rather than a project. The lookback can surface new findings, the independent reviews can add cost, and management already told investors once that the work was in the run-rate. Each of those carries reputational weight with regulators that the expense line does not capture.

Our take: The bulls are right about the financial effect and the bears are right that it is not over. We add a compliance bear pillar at EMERGING rather than MATERIALIZING because the penalty is behind the bank and the lookback outcome is unknown. A business restriction or a second charge would move it, and we would treat either as rating-relevant.

Debate: Was Q3 Trading a Rates Accident or the End of the Markets Premium?

Bull view: The bull case is that an 85bps rate move in a quarter is an unusual shock, that it hit exactly the businesses where BAC leads, that equities posted a record third quarter, and that year-to-date sales and trading is still a record at +13%. Activity is already returning in October, and the trading balance sheet is larger than a year ago.

Bear view: The bear camp notes that management guided flat with two weeks left and missed by ~$350M. The camp adds that the FICC franchise is concentrated in the most rate-sensitive products, and that the market has now seen the earnings mix swing from 28% markets to 15% in one quarter. On that view the multiple should not pay for a Q2-type quarter again.

Our take: Both are right, and the bears have the better point on valuation. The franchise is real and the quarter was unusual. But the swing confirms the concentration risk we opened in July, and we no longer give the stock credit for a markets run-rate above $5.0–5.5B a quarter. That reset is in our estimates, so further good trading quarters are upside rather than base case.

Debate: Can 8%+ NII Carry 2027 if the Hikes Do Not Arrive?

Bull view: The bull case is that the guide is mostly not about rates: fixed-rate repricing, loans growing about $20B a quarter, and noninterest-bearing deposits growing 4–5% do most of the work. The team has raised NII guidance at every opportunity in 2026, and the yield is already rising 8bps a quarter.

Bear view: The bear camp contends that the guide assumes three more hikes and that downside sensitivity ($1.9B per 100bps) now exceeds upside ($1.4B). Markets NII is expected to fall, so the core book must nearly double its growth rate from ~5%. The camp also argues that deposit competition in a loan-growth environment could push betas above management’s assumption that they will match the way down.

Our take: The bulls have the stronger argument, though by less than the guide suggests. We build FY27 on 8% exactly, not “or more,” and our low case strips out most of the hikes. Even there, NII growth of about 6% is enough to support EPS growth near 10% with a modestly lower share count. This is the debate that decides the rating, and it is the one where the evidence is strongest for the company.

Debate: Is BAC Now Cheap, or Cheap for a Reason?

Bull view: The bull case is arithmetic. At $50.39 BAC trades at about 9.8x our FY27 estimate and 1.73x tangible book, for a franchise earning ~15% ROTCE in a bad markets quarter and guiding NII above its medium-term range. It is returning ~9.5% of its market value a year and has underperformed the S&P 500 by more than 20 points year to date.

Bear view: The bear camp contends that the discount is deserved: two guides went wrong in one quarter (trading in September, operating leverage in October), a compliance tail is open, and capital return now depends on rules that have not been published. On that view the stock needs a clean January guide before it re-rates.

Our take: We think the stock is cheap for a reason that is mostly temporary. The confidence discount is reasonable, and we have cut our multiple for it. But the gap between a 2–3% cut in FY27 earnings and a ~17% fall in the share price since July is larger than the new information justifies. That gap is the return we are underwriting, and the January guide is the catalyst that tests it.

Model Update Needed

Valuation impact. Fair value moves to $55–66 from $62–72, on 11–12.5x our FY27 EPS framework of $5.00–5.30. The multiple comes down one turn from 12–13.5x for three reasons: the compliance tail, a markets line that has proven more volatile than we credited, and a capital-return path that now depends on rules not yet published. At $50.39 the midpoint implies roughly 20% upside plus a 2.5% dividend yield. Cross-checks: the range is 1.9–2.3x current tangible book of $29.14, reasonable for a 15–16% ROTCE franchise. At the current $8.3B quarterly pace, capital return is about 9.5% of a ~$349B market value.

Thesis Scorecard Post-Earnings

Scored against the standing thesis carried since Q2 2026, pillar for pillar, with one addition.

Overall: thesis weakened on fees, capital and governance; strengthened on NII. The pillar we upgraded the stock on, NII, delivered and was extended into 2027 above the medium-term range. Credit improved again. Against that, Bull #3 moves to AT RISK, Bear #6 materializes exactly as framed in July, Bear #3 re-opens, and we add Bear #7 for compliance. Bull #4 and Bull #5 hold their tags with weaker evidence. The thesis is now narrower than it was in July: a spread-and-credit franchise, run by a team whose fee and expense forecasts have to be discounted, trading at a price that already discounts more than that.

Action: Maintaining Outperform; fair value cut to $55–66 (11–12.5x FY27 EPS of $5.00–5.30), with conviction lowered to 6/10 from 7/10. We are explicit that our July Outperform has cost readers: the stock is down 16.9% since July against +2.0% for the index. Of the three conditions we said the case depended on, NII delivered, the efficiency ratio left its target band and the multiple compressed. We stay at Outperform because our earnings estimates fell far less than the price: FY27 EPS is down 2–3%, the stock about 17%, and the NII line that carries the estimate is the one management keeps beating. In July we said we would revisit the rating if equities reverted hard without an NII offset, if the buyback stalled unexplained, or if charge-offs broke 50–55bps. Equities reverted with an offset, the buyback held with an explanation, and charge-offs fell to 44bps. We would move to Hold on any of four signals: a January FY27 guide below 200bps of operating leverage or below 8% NII growth; a new charge or business restriction from the lookback; a buyback cut below ~$5B a quarter after the final capital rules; or charge-offs above 55bps. We hold, with a narrower margin for error than at any point since the upgrade.