ROTCE 17% and an Op-Leverage Guide Raised to 300–400bps Carry a Fee-Led Beat; NII Was the Quarter’s Lone Miss — Maintaining Outperform, FV $62–72

Results vs. Consensus

A beat on every reportable line except one, and the exception is the one that matters most to the standing thesis. Consensus figures below reconcile four independent sources; the revenue bar clustered in a $30.49B to $30.78B range and we use ~$30.6B, while $1.13 was the EPS bar cited by the clear majority.

Year-over-Year Comparison

Sequential Comparison (vs. Q1 2026)

The sequential view is where the composition shift is visible. NII added $253M quarter over quarter, a respectable but unremarkable step. Nearly everything else that moved, moved in Global Markets.

Revenue

Revenue of $31.6B (+15%) beat by roughly $960M, and the composition is 83% weighted to the fee side. Noninterest income grew 22% to $15.6B while NII grew 9% to $16.0B on a GAAP basis. For the first time in the four quarters we have covered, the fee engine and the spread engine are within $400M of each other in absolute size, which is a structural statement about what this company now is. Management characterized the growth as broad-based and led by NII, investment banking, wealth management fees, and sales and trading, which is accurate but flattering to the ordering: NII contributed roughly $1.3B of the $4.2B year-over-year revenue increase, and the fee lines contributed $2.8B.

Assessment: The revenue beat is high-quality in the sense that it is client-driven, diversified across three fee engines, and unaccompanied by any one-time item management flagged. It is lower-quality in the sense that its marginal dollar came from the most mean-reverting business in the portfolio during a quarter that management itself described as benefiting from "favorable capital markets conditions." We take the beat, and we do not extrapolate the equities line.

Margins and Operating Leverage

The efficiency ratio at 58.7% improved 359bps year over year and 230bps sequentially, and it has now crossed into the 55–59% target band. Two quarters ago that band was a destination; it is now the current print. Operating leverage of 660bps is more than double the 290bps of Q1 and more than triple the Investor Day "sustainable" commitment of 200–300bps. Expenses grew 8% to $18.6B, which sounds undisciplined until you note that revenue grew 15% and that a meaningful share of the expense growth is the variable cost of the revenue itself: brokerage, clearing and exchange fees on the trading volumes, and financial-advisor incentive compensation on the wealth flows. Headcount has been flat to slightly down for six consecutive quarters.

Assessment: This is the cleanest operating-leverage quarter of the sequence, and the guide raise to 300–400bps is the single most important forward-looking disclosure on the call. But note the mechanism honestly: the leverage is being manufactured largely by revenue outrunning a cost base that grows with it, not by a cost base that is shrinking. That works beautifully while revenue compounds at 15% and inverts quickly if it does not. Management said as much, and we credit them for saying it.

EPS

EPS of $1.21 (+34%) against $1.13 outgrew net income (+27%) by seven percentage points, and the wedge is buybacks: $6.0B of repurchases this quarter on top of $7.2B last, against a ~7.1B diluted share count. Below the line, the tax rate printed 21.5%, which management called "consistent with our full-year guidance." We flag that this sits roughly 150bps above the ~20% FY26 effective-rate framing given on the Q1 call, a gap worth approximately $0.07 to $0.08 of annual EPS, and it was neither reconciled nor asked about.

Assessment: The EPS beat is operationally earned rather than financially engineered: a 27% net-income increase is doing the work, with the buyback adding roughly seven points of accretion on top. There is no tax-rate benefit flattering this quarter, and if anything the tax line is a modest headwind versus the guide the market was carrying.

Segment Performance

Every segment grew revenue, grew net income, generated positive operating leverage, and improved its efficiency ratio. That has not happened in any prior quarter of our coverage, and it is the strongest single structural claim in the print.

Segment revenue sums to $32.4B against $31.6B consolidated; segment net income sums to $9.3B against $9.1B consolidated. The reconciling item both times is All Other.

Global Markets: Records in Both Halves of the House

Revenue of $8.0B (+34%) and net income of $2.6B (+72%, or $2.7B and +70% excluding DVA) make this the quarter's engine. Equities delivered a record $3.6B (+70%) on client financing activity plus derivatives and cash trading. FICC delivered $3.5B (+9%), its strongest quarter in more than a decade. Both halves of the trading floor firing simultaneously is unusual and is precisely the diversification argument management has been making since Q1. Geographically, the US grew 31% and international grew 38%, with Asia-Pacific the standout, which management tied back to Investor Day messaging on international build-out. Segment operating leverage was 16% and return on allocated capital 20%, up from 13%.

“Perhaps what stands out most is the consistency of our performance, because we’ve now delivered 17th consecutive quarters of year-over-year sales and trading revenue growth and 14th consecutive quarters of year-over-year net income growth.”
— Alastair Borthwick, CFO

The consistency claim is well-founded and is the strongest rebuttal to a peak-cycle reading. Seventeen quarters spans multiple rate regimes, a regional-banking crisis, and two distinct volatility environments. But consistency of direction is not the same as durability of level. A 70% year-over-year increase in equities is not a growth rate that any franchise sustains, and the segment now contributes roughly 25% of consolidated revenue and 28% of pre-All-Other segment net income.

Assessment: We mark the Markets revenue base structurally higher for the third consecutive quarter, and we simultaneously open a new bear pillar on concentration. Both are correct. The segment is better than the market gives it credit for and is also now large enough that a normal reversion in equities would visibly dent the consolidated operating-leverage guide. Management flagged the geopolitical dependency themselves.

Global Wealth & Investment Management: The Quietest Compounder in the Building

Record revenue of $6.9B (+16%) and net income of $1.4B (+42%), with pre-tax margin through 27% and return on allocated capital jumping to 26% from 20%. Client balances reached a record $4.9T (+12%), AUM $2.3T (+17%), with ~$14B of AUM flows in the quarter and $78B over the trailing four. Loans grew $13B or 5% sequentially to $277B on custom and securities-based lending. The franchise added 6,000 net new affluent households, which management called the best quarter in the last few, and advisor attrition sits near historic lows.

Against the Investor Day target of 5% net new asset growth over three to five years, management framed the seven-month scorecard as "off to a good start" while acknowledging the flows question directly when pressed.

Assessment: Margin has compounded from roughly 20% a year ago to 26–27% now, against a 30% Investor Day target that is starting to look conservative rather than aspirational. This is the segment we have flagged as the cleanest leverage point in the franchise since Q1, and nothing this quarter changes that. If anything, 42% earnings growth on 16% revenue growth is the most impressive scale demonstration in the print, and it received almost no airtime on the call.

Global Banking: Investment Banking Fees +50%

Revenue of $6.2B (+10%) and net income of $2.0B (+20%), with return on allocated capital at 15%. The headline is corporate investment banking fees excluding self-led transactions at more than $2.1B, up 50% year over year, with strength across debt underwriting, advisory, and equity underwriting. Against the $1.87B consensus bar, that is roughly a $270M beat. Average loans grew 7% to $413B and average deposits 8% to $652B, the strongest deposit growth of any segment.

Management was careful to contextualize the comp: Q2 2025 was, in the CFO's words, "a slower quarter for investment banking for the entire industry," which is a material part of why the growth rate reads at 50%.

Assessment: The absolute level matters more than the growth rate here. $2.1B of quarterly IB fees against roughly $2.0B in Q3 2025 tells you the run-rate is durable even if the percentage decelerates sharply from here, which management explicitly warned it would. Deposit growth of 8% in the corporate book is the underappreciated line: it is the highest-quality funding in the company and it is compounding faster than consumer.

Consumer Banking: Steady, and the Card Flywheel Is Turning

Revenue of $11.3B (+5%) and net income of $3.3B (+10%), with a 51% efficiency ratio, a 29% return on allocated capital, and a fifth consecutive quarter of positive operating leverage. Average deposits of $957B marked the fifth consecutive quarter of year-over-year growth at a rate paid of 48bps. The franchise added 162,000 net new checking accounts, card spending grew 9% to $266B, and consumer investment assets hit a record $640B (+18%). Digital remains the structural cost story: roughly 50 million active digital users, more than 24 million Erica users, and 70% of sales now digital. Management relaunched the rewards program in late May and has taken more than 2 million enrollments since, alongside a FIFA World Cup marketing campaign.

Card balances grew 4% year over year, tracking the path management laid out toward 5%.

“Just keep half an eye also on card where Holly laid out a plan to say we want to get back towards 5% type card growth. We were at 1%, then two, then three. You can see this quarter we’re at 4%.”
— Alastair Borthwick, CFO

Assessment: The 1→2→3→4 card progression is the kind of unglamorous, four-quarter, self-reported scorecard we like, because it is falsifiable and management keeps volunteering it. A 5% card book at BAC's spread is a meaningful NII contributor in 2027 and beyond. The 5% revenue growth here is the slowest in the company and it is fine: this segment's job is to fund the balance sheet at 48bps and throw off a 29% return, and it did both.

All Other: A $292M Hole Nobody Explained

A net loss of $292M, wider than the year-ago quarter, disclosed with the phrase "no significant drivers to note." At roughly 3% of consolidated net income it is not thesis-relevant, and we note it only because a widening unexplained loss line is the kind of thing that is immaterial every quarter until it is not.

Key KPIs

Key Topics & Management Commentary

Overall Management Tone: Management was the most relaxed it has been across the four calls we have covered, and the posture was quantitative rather than promotional: guides were raised with arithmetic attached and the tougher second-half comparisons were volunteered before analysts could raise them. Where the tone thinned was on capital, where a decelerating buyback and a flat CET1 ratio went unmentioned, and on the durability of a record trading quarter, which was defended with a consistency streak rather than a forward framework. The call ran short, with one analyst dropped to a muted line and the queue exhausted after eight questioners, which itself is a signal that the Street found little to contest.

1. NII: Guided to the Top of the Range, and the Quarter's Only Miss

NII on an FTE basis was approximately $16.2B, up 9% year over year and $253M sequentially, with net interest yield at 2.08% (+1bp sequentially, +14bps year over year). Against a consensus bar of $16.24B, that is a ~$80M miss, the only one on the board. Management raised the full-year growth guide for the third consecutive quarter, from 5–7% in January, to 6–8% in April, to the upper end of 6–8% now.

“Looking ahead on NII expectations, in January, we told you to expect 5%-7% full year NII growth. In April, we raised that full year range to be 6%-8%. We now expect full year 2026 NII growth to be at the upper end of that 6%-8% range, supported by anticipated loan and deposit growth, fixed rate asset repricing, and balance sheet optimization.”
— Alastair Borthwick, CFO

The trajectory framing management offered is that NII has grown from $13.9B in Q2 2024 to $16.2B now, a $2.3B climb over eight quarters achieved while short rates fell. Under questioning the CFO went further than the formal guide, offering "more like 8% for the full year just based on the comps."

Assessment: A $80M miss on a $16.2B line is 0.5% and is not a trajectory break, particularly against a full-year guide that went up on the same call. But it is the first NII miss of our covered sequence, and NII is the pillar we upgraded this stock on at Q4 2025. The honest read is that the spread book is performing exactly as guided while the market's quarterly bar drifted slightly ahead of it. We treat this as phasing, not deterioration, and we watch it.

2. The Rate Curve Flipped to a Hike, and the Original Bear Case Inverted

The single largest change in the macro assumption underneath this company is easy to miss because management delivered it in one clause. The FY26 NII guide is built on a forward curve carrying one 25bps rate hike in September. At Q4 2025 the guide assumed two cuts. At Q1 2026 it assumed none. The path from "two cuts" to "one hike" in six months is a 75bps swing in the direction that helps an asset-sensitive bank most.

The disclosed sensitivity moved accordingly: a 100bps parallel shift above the forward curve is now expected to add roughly $1B of NII over the next twelve months on a dynamic-deposit basis. At Q1, the equivalent upward sensitivity disclosed was approximately $500M.

“I think generally speaking, because the banking book is liability insensitive, that’s the predominant benefit. The markets business is slightly liability sensitive, so that’s a slight offset. Net, net, it’s a positive for us, and that’s what we’re trying to communicate.”
— Alastair Borthwick, CFO

Assessment: This retires the original bear case entirely and converts it into a tailwind. We initiated at Hold in Q2 2025 explicitly because the asset-sensitive book was exposed to a 2026 easing cycle nobody had sized. That easing cycle is not happening. The asymmetry now runs the other way, and it is not in the guide beyond the September hike. The caveat is what a hiking curve implies about the world: management named inflation and tighter monetary policy as the key risks in the same breath as the constructive outlook, and a Fed hiking into 2027 is not obviously a friendly environment for either credit or equity multiples generally.

3. Operating Leverage: The Guide Nearly Doubled

Full-year operating leverage guidance moved from "more than 200 basis points" at the Q1 call to 300–400bps, after a first half that delivered more than 450bps and a second quarter that delivered 660bps. For reference, the Investor Day commitment for the sustainable through-cycle level was 200–300bps.

“On our first quarter earnings call in April, we told you we expected full-year operating leverage of more than 200 basis points. Operating leverage for the first half of 2026 has now exceeded 450 basis points. With that first half performance and our continued expectations for a strong second half, we now expect full-year operating leverage to be in the range of 300-400 basis points.”
— Alastair Borthwick, CFO

The arithmetic embedded in that range is worth spelling out, because management did not. A first half at 450bps and a full year at 300–400bps implies a second half in the roughly 150–350bps band. The guide raise is therefore also, quietly, a deceleration guide.

Assessment: This is the most important forward disclosure on the call and it is a genuine upgrade to the FY26 earnings framework. It is also less aggressive than it first reads: the company is guiding the second half below both the first half and, at the low end, below the Investor Day sustainable range. Management attributed that to two things it named plainly, tougher NII comps and an investment-banking base that gets much harder after Q3 2025's ~$2B print. We take the raise and model the second half at the middle of the implied band.

4. ROTCE 17% and ROA Through 1%

Return on tangible common equity printed 17.0% against 13.6% a year ago, and return on assets printed 1.03% against 0.84%, crossing 1.00% for the first time in our covered sequence (FY25 full-year ROA was 89bps). Recall the framing: at Q4 2025 management articulated a glide path to 16–18% ROTCE over eight to twelve quarters. Q1 cleared 16% in the first quarter of that path. Q2 has now cleared 17%.

“The return on tangible common equity was 17%. We thought it would take us longer to get there. That has in part due to the strong operating of the general businesses that are operating very well, getting good operating leverage, Consumer Banking, commercial banking, and benefiting by the NII lift as well as the strong performance markets.”
— Brian Moynihan, CEO

The CEO's own admission that "we thought it would take us longer" is the most candid line on the call, and it is doing double duty: it concedes the glide path was conservative and it implicitly concedes that markets performance is part of why the number is where it is.

Assessment: Two consecutive quarters inside the medium-term target range, on a path that was supposed to take two to three years, is the single strongest validation of the Q4 upgrade. The upper bound (18%) is now within one good quarter's reach and management still declined to put a date on it. We model FY26 ROTCE at 16–17% and treat 18% as plausible in 2027 rather than committed.

5. The Markets Concentration Question Nobody Asked

Global Markets produced 25% of consolidated revenue and roughly 28% of pre-All-Other segment net income this quarter, on a record equities result up 70% year over year and a FICC result management called the best in more than a decade. Not one of the eight analysts who reached the queue asked whether a record equities quarter driven by "client financing activity and strong trading performance in derivatives and cash" constitutes a run-rate.

Management's own framing of the risk came unprompted, in a different answer:

“Right now, we see very strong performance this quarter, and we expect it to continue based on the market conditions. The Iran war is intractable, and we can’t predict what will happen next in it, and that could affect the market’s perception, IPOs, et cetera. Right now, the pipelines are feeling very good about that.”
— Brian Moynihan, CEO

Assessment: We are adding markets concentration as a new bear pillar this quarter. The point is not that Global Markets is low quality; 17 straight quarters of growth across multiple regimes says otherwise, and the diversification between equities and FICC is real. The point is arithmetic: the segment is now large enough and is printing far enough above trend that a normal reversion in equities alone would consume most of the operating-leverage guide raise. The CEO named the geopolitical trigger himself. That the Street did not press on it is a gap in the market's diligence, not evidence the risk is absent.

6. Investment Banking Fees +50%, With the Comp Caveat Volunteered

Corporate investment banking fees excluding self-led transactions exceeded $2.1B, up 50% year over year, with strength across debt underwriting (revenue $1.11B against a $958.7M bar), advisory, and equity underwriting. Against the $1.87B consensus estimate this was a ~$270M beat, the largest single line-item surprise in the print.

Management immediately and voluntarily discounted its own growth rate:

“Remember, second quarter last year was a slower quarter for investment banking for the entire industry. You think about it, if we put up $2.1 billion of investment banking this year, up 50%, if you kind of sustain that relative to what we did in the third quarter of last year, which was $2 billion or so, you just don’t get the same kind of uplift in the second half.”
— Alastair Borthwick, CFO

Assessment: Credit where due: a CFO who pre-emptively explains why his own 50% growth rate overstates the trend is behaving like a steward rather than a promoter. The signal to extract is the level, not the rate. $2.1B per quarter against roughly $2.0B in Q3 2025 says the IB franchise has settled at a structurally higher plateau than the FY25 base, and that plateau is what belongs in the model. The 50% is noise; the $2.1B is the number.

7. Deposits: Twelve Straight Quarters, and the Mix Is the Story

Average deposits of $2.02T grew $49B or 2.5% year over year, marking the twelfth consecutive quarter of average deposit growth, with noninterest-bearing balances up $19B or 4% in the seventh consecutive quarter of NIB growth. Consumer rate paid fell to 48bps on $957B of balances. Global Banking led with 8% growth. Sequential growth was muted on seasonal tax-related outflows, which management called out rather than let it pass.

“The lower rate paid is really about mix, Chris. We’re competing out there for deposits like everyone else. We compete tooth and nail to get deposits where we can. At the end of the day, our strategy is about relationship value, all the things around digital and security and rewards that we talk about. It’s that favorable mix of growing the non-interest bearing that makes a difference.”
— Alastair Borthwick, CFO

Assessment: This is the pillar that has never wobbled across four quarters of coverage and it did not wobble here. The mechanism is worth restating because it is the whole franchise: BAC is not winning deposits on price, it is winning operating accounts on relationship, and operating accounts are disproportionately noninterest-bearing. Seven consecutive quarters of NIB growth in a high-rate environment is the hardest thing on this scorecard to replicate and the easiest to overlook.

8. Loans: Nine Straight Quarters, Commercial Leading at 11%

Average loans and leases of $1.22T grew $88B or 8%, the ninth consecutive quarter of both average and ending loan growth. Commercial led at $733B (+$75B, +11%), broadening both domestically and internationally and, importantly, broadening away from the Global Markets-related activity that drove much of last year's growth. Consumer loans grew 3%, led by securities-based lending and card (+4%). Home equity extended a ninth consecutive quarter of average growth.

“On the commercial loan growth, yes, it’s broader than just an AI theme. The AI theme has helped because there’s so much in the way of capital investment going on globally now. When we look at our Global Banking segments, each of the lines of business, whether it’s business banking, the commercial bank or the corporate bank, they’re all contributing.”
— Alastair Borthwick, CFO

Assessment: The explicit disclaimer that commercial growth is broader than the AI capex theme is the answer to the question we would have asked. Loan growth concentrated in data-center and AI-infrastructure lending would be a different risk profile than broad middle-market and corporate demand. Management says it is the latter, across all three Global Banking lines, and the 11% growth rate against 3% consumer growth tells you where the balance sheet is being deployed.

9. Credit: Criticized Down $2.3B, and a New Phrase Worth Noting

Provision of ~$1.4B and net charge-offs of ~$1.4B were both largely unchanged sequentially, with an NCO ratio of 0.47% against 0.55% a year ago and 0.48% last quarter. Consumer card charge-offs and delinquencies improved both year over year and sequentially. Reservable criticized commercial exposures declined ~$2.3B to roughly $22B, driven primarily by CRE improvement. Nonperforming loans held stable at ~$5.8B and the firm recorded a modest reserve release.

One phrase in the credit commentary is new: commercial credit was described as solid "with CRE improvement offset by some isolated corporate and commercial lending losses." Isolated losses in the corporate book have not appeared in this language in prior quarters.

“Importantly, the issues of the moment, whether it’s real estate four or five years ago or whether it was private capital lending and all this stuff, just aren’t surfacing the way people thought they would.”
— Brian Moynihan, CEO

Assessment: The CRE office arc that we tracked from Q2 2025 through its formal resolution at Q1 2026 continues to unwind favorably, and a $2.3B decline in criticized exposure is a real improvement, not a rounding item. The NCO ratio at 47bps remains below the through-cycle 50–55bps reference, which keeps the normalization skew negative by construction. We flag the "isolated corporate and commercial lending losses" language as a watch item rather than a concern: one quarter of a new phrase is not a trend, but it is the sort of thing that reads differently in hindsight.

10. Balance-Sheet Optimization: $800B of Dry Powder and a Repo Paydown

Ending assets were steady at $3.5T, with lower securities balances replaced by loan growth and Global Markets activity. The CFO quantified the liquidity position bluntly: roughly $800B of excess between cash and securities over loans, or in his words, "we're not loaned up at this point." The optimization program has two remaining levers: continued paydown of repo and institutional CDs, which management expects to accelerate in the second half, and the ongoing fixed-rate asset repricing.

“We still feel like we carry some repo, some institutional CDs, that over time, we’re just continuing to pay down. If those are invested at the Fed, we’re not capturing a lot of spread. It doesn’t do anything for NII. It actually hurts NIY. It also ties up a little bit of capital. As we continue to pay that down, and I think you’ll see more of that happening in the second half of this year, that’ll free up more capital.”
— Alastair Borthwick, CFO

On the longer arc, the net interest yield target of 2.30% was reaffirmed and the timeline pulled in: originally framed as two to three years, management now says "we're probably inside a couple of years now to get there," from 2.08% today.

Assessment: The $800B of unlevered liquidity is the most underrated line in the print. It means the loan growth of the last nine quarters has cost the company nothing in funding stress, and it means the deposit franchise can fund several more years of 8% commercial loan growth without touching wholesale markets. The repo and CD paydown is a capital-release story as much as a margin story, and management explicitly tied it to ROTCE. The NIY path from 2.08% to 2.30% is roughly $3B of annualized NII on the current earning-asset base.

11. AI: 300 Approved Use Cases, and a Productivity Claim With Numbers Attached

Management added a dedicated AI slide and populated it with specifics: more than 200,000 teammates actively using AI-enabled capabilities, more than 400,000 prompts per day, more than 300 approved use cases "all of which have good economics," 114 live generative AI use cases, and 34 fully implemented. Applications span relationship-manager meeting prep, banker research and presentation automation, and developer coding support.

Asked whether the revenue upside creates a risk of under-investing, the CEO connected AI to the cost curve rather than to headcount reduction:

“Even the coding process has become more and more efficient using these tools. The same amount of money in 2027 will get us more code, for lack of a better term, in 2028. We’re driving everything as hard as we can. ... I don’t think you’ll see a major change in our methodology of how we think about spending.”
— Brian Moynihan, CEO

Assessment: The disclosure is a step up in specificity from Q1's more anecdotal framing, and the funnel shape (300 approved, 114 live, 34 fully implemented) is honest about how early this is. What it is not yet is quantified in dollars. Six quarters of flat-to-down headcount is the only hard evidence in the print that AI is doing economic work, and headcount discipline has plenty of other explanations. We continue to treat AI as supporting the operating-leverage case rather than driving it, and we would treat a specific efficiency-savings disclosure as the milestone that changes that.

12. Capital Return Decelerated While CET1 Held Flat

The firm returned $8.0B this quarter, $6.0B of buybacks plus $2.0B of dividends, against $9.2B in Q1 ($7.2B plus $2.0B). CET1 finished at 11.2%, flat sequentially, with CET1 capital at $201.6B and RWA at $1.8T. The ratio remains well above the ~10% minimum.

Recall the Q1 framing: management described a deliberate drift lower in CET1 toward a target buffer of roughly 50bps over the regulatory minimum, and Q1's $9.2B return deliberately exceeded earnings generation to effect that drift. This quarter the drift stopped. Capital return of $8.0B against $9.1B of net income means the firm retained capital rather than releasing it, and RWA growth from loan and markets activity absorbed the rest.

Assessment: This is the one pillar that went backwards, and the silence around it is what makes it notable. There is a perfectly benign explanation available, that 8% loan growth and a larger markets balance sheet consumed the capital that would otherwise have funded buybacks, which is capital deployed at attractive returns rather than capital hoarded. We think that is probably what happened. But management did not say so, no analyst asked, and the Q1 narrative of a deliberate glide toward a 50bps buffer was left dangling. We downgrade this pillar's status and want it addressed at Q3.

Guidance & Outlook

Implied second-half ramp. First-half operating leverage exceeded 450bps against a full-year guide of 300–400bps, which implies roughly 150–350bps in the second half. On NII, first-half growth ran 9% against a full-year guide of "about 8%," implying second-half growth of roughly 7%. Both guides are, read literally, deceleration guides dressed as raises. Management was explicit about why: the second half of 2025 is where last year's NII build happened, and Q3 2025's ~$2B investment banking print removes the easy comp.

Street positioning. The consensus view entering the call, surfaced directly in the Q&A, was that the NII guide was conservative. Management's response was to nudge it up verbally without formally moving the range, which is the behavior of a team that expects to beat it. The average sell-side price target sits in the $62–64 range with peak targets around $71.

Guidance style. This management team has now raised the FY26 NII guide three consecutive times (5–7%, then 6–8%, then the top of 6–8%) and raised the operating-leverage guide once, substantially. That is a consistent pattern of setting a beatable bar and walking it up, which argues the "about 8%" NII framing and the 300–400bps operating-leverage range are floors rather than midpoints. The offsetting consideration is that both raises have been powered increasingly by a fee engine running at record levels.

Analyst Q&A Highlights

Whether the NII Guide Is Deliberately Conservative

The dominant line of questioning on the call was skepticism that the raised NII guide was still too low. The framing put to management was that first-half NII grew 9% while the full-year guide implies something lower, and that the components management had described (modest deposit growth, good loan growth, improving card yields) did not obviously produce that deceleration. The response was the most valuable disclosure of the call, because management effectively pre-announced the top of its own range in conversational form rather than formally moving it.

Q: “The investor feedback so far, is that they feel that the net interest income guide is conservative. ... The first half of the year, NII growth is up nine. Clearly the second half of the year is tougher comps, right? Which you’re saying would get you within the range. ... Are we getting earning asset growth for the second half of the year, but not much NIM expansion? I guess we’re just trying to think about the jumping off point that will slow your NII growth from the 9% that you’ve printed for the first half of the year.”
— Erika Najarian, UBS

A: “Well, first thing I should say is it’s not slowing it much. ... What you see is most all of the NII build last year was in the second half of the year. We’re just up against tougher comps, that’s all. It’s not more complicated than that. ... We’re talking about 8% or 9%. I don’t know. They’re both very good. We kind of feel like right now it looks to us like more like 8% for the full year just based on the comps.”
— Alastair Borthwick, CFO

Assessment: "It's not slowing it much" and "more like 8%" are a guide raise delivered without the formality of one. The exchange establishes 8% as the working number rather than the ceiling of a range, and the comp explanation is arithmetic rather than hedging. This is the exchange we would point to if asked why we stayed at Outperform: the spread engine is not decelerating, the base it is measured against is rising.

Boxing the Second-Half Operating Leverage

Having raised the full-year operating-leverage guide to 300–400bps after a 450bps first half, management was pressed to explain what the range implies for the back half, given that the arithmetic makes the raise look like a deceleration. The answer conceded the framing openly and attributed the step-down to two specific comps rather than to any change in cost trajectory.

Q: “Obviously, with 300 and 660 basis points of leverage and now talking to a full year of 300-400, just how do we kind of box the operating leverage potential for the second half as we get into this kind of tougher comps, albeit with the good top-line revenue growth continuing?”
— Ken Usdin, Autonomous Research

A: “We’re offering the 300-400, recognizing that we’ve already booked 450 for the first half, so that’s good. We’re just trying to give you a range, Ken, that allows you to kind of work backwards, because as you point out, in the second half of last year, the NII went up more than the first half. ... Remember, second quarter last year was a slower quarter for investment banking for the entire industry. ... Otherwise, as you can imagine, business conditions are very good.”
— Alastair Borthwick, CFO

Assessment: "A range that allows you to kind of work backwards" is an unusually transparent admission that the guide is constructed to be beaten rather than met. Management is not signalling deterioration; it is signalling that the comparison base rises. The distinction matters for the model: we hold the cost trajectory and let the comps do the decelerating.

What Actually Changed to Justify the Operating-Leverage Raise

A separate line of questioning probed whether the raise reflected structural scalability finally arriving, the long-awaited moment when incremental revenue lands at low marginal cost, or simply better revenue. Management's answer was candid that it is mostly the latter, with two named contributors and no claim of a step-change in the cost structure.

Q: “Could you elaborate more on the change in your operating leverage guide? That’s quite a big lift there. ... Your marginal margin or the scalability of your model, I think you’ve been kind of waiting for this moment when you can layer on more revenues at lower marginal cost. Can you highlight the areas that’s impacting that the most?”
— Mike Mayo, Wells Fargo

A: “The two reasons for that are, first, NII just keeps grinding higher, and as Brian pointed out, all of that’s dropping to the bottom line. That’s really powerful. Second, we had really terrific fee-based performance over the course of the first six months. ... At some point, you’re halfway through the year with 450 basis points of operating leverage, and it’s pretty clear we’re going to be above 300.”
— Alastair Borthwick, CFO

Assessment: The honest answer, and the one that supports our composition caution. The operating-leverage raise is revenue-driven, not cost-driven. NII dropping through to the bottom line is durable; six months of "terrific fee-based performance" is not the same kind of durable. Management did not claim a marginal-margin breakthrough when handed the opportunity to, which we read as integrity rather than absence.

The Deliberate Absence of Expense Guidance

Asked directly for a full-year or second-half expense number, management declined, and the reasoning is worth surfacing because it changes what the operating-leverage guide can be held to. With revenue-linked costs (brokerage, clearing, exchange fees, advisor incentives) moving with the fee lines, management argues an expense guide would be revised every quarter and would mislead. It offered headcount as the substitute discipline metric.

Q: “Did you provide any expense guidance for the year or the second half?”
— Mike Mayo, Wells Fargo

A: “No, we’ve largely gone away from that, Mike, for the very simple reason that when you have revenues increase this quickly, and some of them come with brokerage clearing and exchange costs, some come with FA incentive cost. It’s really hard for us to just keep updating the expense through the case of the quarter. ... I think if you look at the core, the best core measure is probably headcount. Our headcount discipline over the last six quarters has been excellent. It’s flat to slightly down. ... The expense at this point is really going to be based on what happens with revenue. If the revenue isn’t there, the expense will come down.”
— Alastair Borthwick, CFO

Assessment: The reasoning is defensible and the headcount substitute is a real metric. The consequence is that the 300–400bps operating-leverage guide cannot be independently checked against a cost commitment; if revenue disappoints, the guide adjusts by construction because "the expense will come down." That is a self-healing guide, which is comfortable for management and less useful for underwriters. We would rather have both.

Deposit Pricing in a Higher-for-Longer Environment

The call opened on deposit pricing, with the observation that BAC's rate paid outperformed peers this quarter and a request for the near-term outlook. Management redirected from price to mix, which is the correct answer and also the whole franchise argument in one paragraph.

Q: “Notice the deposit discipline in the quarter. Alastair, I’m interested in your thoughts about pricing in a higher for longer environment. I know you didn’t change the full-year NII guide, but your deposit pricing outperformed some of your peers this quarter. Any comments on the near-term outlook would be great.”
— Chris McGratty, KBW

A: “I think you know we’ve got a lot of liquidity. We’re not loaned up at this point. We’ve got $800 billion of excess between our cash and securities over our loans. That really allows us to concentrate on our strategy. ... When consumer puts up 162,000 of net new checking accounts, or when we grow non-interest bearing for seven consecutive quarters, that’s what helps us to drive the non-interest bearing up 4%. The lower rate paid is really about mix, Chris.”
— Alastair Borthwick, CFO

Assessment: The $800B excess-liquidity figure is the answer beneath the answer. A bank that does not need the marginal deposit can decline to pay for it, which is why the rate paid fell while balances grew. This is the structural advantage that makes the NII guide credible in a hiking curve, and it is why we have never seriously doubted this pillar.

Net Interest Income Excluding Global Markets

The one substantive question management could not answer live. With Global Markets carrying an increasing share of the balance sheet and of NII, the question of how much of the ~8% full-year NII growth comes from the banking book versus from balance sheet allocated to trading is central to judging the quality of the guide. Management could not produce the figure, deferred it offline, and then declined the disclosure on principle.

Q: “Just a quick follow-up on net interest income. The guidance for this year is essentially to be up 8%. What is that ex markets?”
— Matt O’Connor, Deutsche Bank

A: “I need to take a look. Give me one second. I need to work it back. ... Matt, I can help with that offline once we finish it up. ... We don’t normally provide the guidance ex Global Markets because there’s lots of moving pieces that go backwards and forwards, and there’s a lot of loans in markets. It gets a little confusing when we strip that out sometimes. ... Bottom line is markets NII I think will be flattish, could be slightly down.”
— Alastair Borthwick, CFO

Assessment: The qualitative answer is actually favorable and largely defuses the question: if markets NII is flat to slightly down, then essentially all of the second-half NII growth is the banking book, which is the higher-quality source. But the number was not given, publicly or in the guide, and it is the second time in two quarters that the markets balance sheet has been described as a growing NII contributor without being sized. We would like the disclosure. Management's stated reason for withholding it, that it "gets a little confusing," is the weakest sentence on the call.

Second-Derivative Credit Exposure to the AI Capex Cycle

The most forward-thinking question of the call went past direct data-center lending exposure to the knock-on effects: if AI investment decelerates in two to four years, what breaks in the loan book that is not obviously AI-related today. Management's answer described process rather than quantification, and connected the underwriting question back to the firm's own AI adoption.

Q: “AI is such a powerful economic force in this country. Have you guys been able to frame out, not exposure to data center build-out, but the second derivative? Because you wonder two, three, four years from now if AI ever kind of slows down or rolls over. What’s the second derivative credit aspect from that? Have you guys given that much thought?”
— Gerard Cassidy, RBC

A: “When we look at the work we do with clients and others, we factor in across all our companies as part of the underwriting work that the team does. They have to factor in the question of what’s the impact of AI in the industry and the company and what will happen. ... As we look at underlying deals, we’re always looking at the credit, the capability and the earnings power of the underlying tenants, so to speak, that are driving the revenue to the build-out.”
— Brian Moynihan, CEO

Assessment: The best question of the call drew a process answer, not a sized answer. "We look at the earnings power of the underlying tenants" is the right framework for data-center credit and says nothing about the second-derivative exposure that was actually asked about: the borrowers whose demand is indirectly propped up by an AI capex wave. No number, no scenario, no portfolio slice. That is a fair reflection of how early the industry is on this question, and it is also an unquantified tail we now carry.

What They’re NOT Saying

1. Basel III endgame and G-SIB indexing, entirely. The comment period closed mid-June. At the Q1 call management said the combination would produce "some reduction in overall capital requirements relative to the current regime," and we carried that as an active driver of the capital-return pillar. This quarter: no mention in prepared remarks, no analyst question, no update on final-rule timing. A regulatory tailwind the company itself introduced two quarters ago has gone silent, and silence is not confirmation.
2. Why the buyback decelerated. $7.2B to $6.0B, total return $9.2B to $8.0B, with CET1 flat at 11.2% instead of continuing the deliberate drift toward the ~50bps-over-minimum buffer described at Q1. The benign explanation (loan and markets RWA growth consumed the capital) is probably correct and would be a good answer. Management did not give it, and nobody asked.
3. NII excluding Global Markets. Asked directly, could not be computed live, deferred offline, and then declined as a matter of standing practice. Markets NII is now material enough to move the consolidated growth rate and has been described as a contributor for two consecutive quarters without ever being sized.
4. Whether a record equities quarter is a run-rate. Equities at $3.6B (+70%) was the single largest driver of the beat and drew zero questions. Management volunteered no forward framing beyond "we expect it to continue based on the market conditions," with the CEO naming the Iran war as the thing that could change market conditions.
5. The tax-rate reconciliation. 21.5% printed and was described as "consistent with our full-year guidance," against a ~20% FY26 effective-rate framing from the Q1 call. Roughly 150bps of tax rate is roughly $0.07 to $0.08 of annual EPS. Neither reconciled nor challenged.
6. Any expense number at all. Explicitly and deliberately withdrawn as a disclosure, with headcount offered as the proxy. Defensible, and it makes the raised operating-leverage guide unfalsifiable on the cost line.
7. The All Other loss. $292M, wider year over year, "no significant drivers to note." A widening loss line described as having no drivers is a non-explanation.
8. A date for 18% ROTCE. 17% printed against a 16–18% medium-term range that was supposed to take eight to twelve quarters to enter. Asked about return targets and willingness to accept lower-returning business, the CEO answered with the general framework and no timeline for the upper bound.
9. Berkshire. Still unmentioned by management, still unasked by analysts, four quarters running. We continue to treat the marginal-bid overhang as fully normalized and no longer a thesis input.
10. What "isolated corporate and commercial lending losses" refers to. New language in the credit commentary, offsetting CRE improvement, unquantified and unexplored.

Market Reaction

• Pre-print setup: BAC closed at $59.50 on July 13, entering the print +8.2% year-to-date (versus the S&P 500's +9.8%), +6.2% over the trailing 30 days, and +26.4% over the trailing twelve months. The 52-week closing range was $44.92 to $59.90, which put the stock roughly 0.7% below its 52-week closing high going in. This was a fully-priced setup, and notably one in which the stock had underperformed the index year-to-date despite two consecutive framework upgrades.
• Reaction session (July 14, before-open reporter): Premarket indications pointed higher, but the stock opened at $59.00, gapping down 0.8%, and traded as low as $58.67 (−1.4%) before reversing. It ran to an intraday high of $61.21 (+2.9%) and closed at $60.62, up 1.9% (+$1.12), a new 52-week closing high.
• Volume: 49.4M shares against a 37.4M 30-day average, or 1.3x normal. Elevated but not extreme.
• Relative: The S&P 500 closed +0.4% on the session, so BAC outperformed the tape by roughly 150bps.

The shape of the session is more informative than the close. A stock that beats on EPS by 7%, beats revenue by 3%, raises two guides, and prints a 17% ROTCE does not normally open down. The most plausible reading of the early fade is the NII line: $16.16B against a $16.24B bar was the one number in the release that a fast-reading algorithm or a rates-focused desk would flag, and it is the number that this stock's 2025 rerating was built on. Add profit-taking against a 52-week closing high, and a soft open is coherent.

The recovery is where the call did its work. The operating-leverage guide raise from 200bps+ to 300–400bps and the CFO's verbal "more like 8%" on NII were both call events, not release events. The stock bottomed early and climbed through the session, which maps to the market processing the guidance rather than the print. That is an unusually clean example of a call adding value to a release.

What the close does not say is that the market got excited. Paying 1.9% for the best quarter in the franchise's recent history, on 1.3x volume, is a measured response. Our read: positioning was already long into the print, the beat was recognized as fee-led and discounted accordingly, and the stock is being repriced on earnings delivery rather than on multiple expansion. That is precisely the pattern that has left BAC trailing the S&P year-to-date while its earnings framework has been revised up twice, and it is the core of why we stay Outperform.

Street Perspective

Debate: Is the Markets Engine a New Base or a Cycle Peak?

Bull view: The bull case on the Street points to the streak, 17 consecutive quarters of year-over-year sales-and-trading growth and 14 of segment net-income growth, spanning multiple rate regimes and volatility environments, as proof this is a share-gain story rather than a volatility story. Equities and FICC both printing records simultaneously is portfolio diversification working, international growth of 38% is the Investor Day build-out converting, and the client-financing balances underpinning equities are stickier than flow trading.

Bear view: The bear camp contends that a 70% year-over-year increase in equities is definitionally not a run-rate, that Global Markets at 25% of revenue and 28% of segment earnings is the highest cyclical concentration this franchise has carried in years, and that the operating-leverage guide raise is therefore borrowing against a revenue line that mean-reverts. The CEO named the geopolitical trigger unprompted.

Our take: Both are right about different things, and the bulls are answering a question the bears are not asking. The streak establishes that the franchise has structurally gained share; it does not establish that this quarter's level repeats. We mark the Markets base higher for a third quarter and simultaneously open a concentration bear pillar. The practical test is Q3: management has already told us the IB comp gets much harder, so a Q3 that holds $2B in IB fees and $6.5B or better in sales and trading would confirm a plateau rather than a peak.

Debate: Is 8% NII the Ceiling or the Floor?

Bull view: The bull case is that this management team has raised the NII guide three consecutive times, that the CFO volunteered "more like 8%" rather than defending the range, that the curve now carries a hike rather than cuts, that the disclosed upside sensitivity has roughly doubled to $1B per 100bps, and that the balance-sheet optimization program still has repo and institutional CDs to retire in the second half. A team that has beaten its own guide three times running is not sandbagging by accident.

Bear view: The bear camp notes that NII actually missed the quarterly bar for the first time, that first-half growth of 9% against a full-year of 8% requires a genuine second-half deceleration, that the fixed-rate repricing tailwind is further through its cycle than a year ago, and that the growing markets balance sheet is flattering the consolidated NII number in a way management declines to quantify.

Our take: The bulls have this one, but by less than they think. The comp explanation is arithmetically sound and the miss is 0.5% on a raised full-year frame. The genuine unresolved point is the bears' last one: without an ex-markets NII figure, nobody outside the company can decompose the 8%. Management says markets NII will be flat to slightly down, which if true means the banking book is carrying essentially all of the growth and the quality is better than the headline. We believe them and would still rather see the number.

Debate: Has the Multiple Already Absorbed the Story?

Bull view: The bull case is arithmetic. At $60.62 against an FY26 framework of $4.75–4.90, BAC trades at roughly 12.6x, essentially the same multiple it carried before the Q4 upgrade, while ROTCE has moved from the low teens to 17% and the efficiency ratio has entered its target band. The stock has lagged the S&P year-to-date through two framework upgrades. Add a ~7.5% annual shareholder yield on the current return pace and tangible book compounding at 7%, and the total-return math clears the index without any rerating at all.

Bear view: The bear camp contends the stock sits at a 52-week closing high having risen 26% over twelve months, that the easy part of the rerate is done, that the beat's fee-led composition deserves a lower multiple rather than a higher one, and that a market paying only 1.9% for a quarter this good is telling you the good news is in the price.

Our take: The bull framing is stronger and it is the reason we are not downgrading. The multiple has not participated in the earnings upgrade, which means the market is treating the improvement as cyclical. If it is structural, and the deposit franchise, the efficiency ratio, and the ROTCE path say much of it is, then the multiple is the upside option we are not paying for. The bears' composition point is legitimate and is why our conviction steps down rather than up, and why fair value at the midpoint implies a high-single-digit rather than a spectacular return.

Model Update Needed

Valuation impact. Fair value moves to $62–72 from $52–60, on roughly 12–13.5x our FY27 EPS framework of $5.15–5.40, which is the earnings base the market will be discounting across our twelve-month horizon. At the $60.62 close, the midpoint of that range implies roughly 10% price appreciation plus a ~1.9% dividend yield. The floor under that math is the shareholder yield: at the current $8B quarterly pace against a ~$425B market capitalization, BAC is returning roughly 7.5% of its equity value annually while growing tangible book 7%.

Thesis Scorecard Post-Earnings

Scored against the standing thesis carried since Q1 2026, pillar for pillar, with one addition.

Overall: thesis strengthens on earnings power, weakens modestly on composition and capital. Four of five bull pillars are confirmed or confirmed-plus, and Bull #5 (ROTCE) plus Bear #1 (asset sensitivity) have moved decisively in our favor: the glide path is running roughly two years ahead of schedule and the macro assumption that originally kept us at Hold has inverted into a tailwind worth ~$1B per 100bps. Against that, Bull #4 (capital return) stalls and is downgraded to AT RISK, the Basel III driver we carried out of Q1 has gone silent and comes out of the model, and we open Bear #6 on markets concentration. The net is a company earning more than we modeled, in a mix we trust slightly less, with a capital story that needs an answer at Q3.

Action: Maintaining Outperform; fair value raised to $62–72 (roughly 12–13.5x our FY27 EPS framework of $5.15–5.40). We step conviction down from high to moderate (7/10) on the composition shift, and we are explicit that this is now a narrower Outperform than the one we carried out of Q1. The case does not depend on the trading engine repeating: it depends on NII delivering ~8%, on the efficiency ratio holding inside its target band, and on a ~12.6x multiple not compressing while ROTCE sits at 17%. The stock has lagged the S&P 500 year-to-date through two consecutive framework upgrades, which tells us the market is pricing this improvement as cyclical; we think a majority of it is structural, and the gap between those two readings is the return. We would revisit the rating if Q3 equities revert hard without an NII offset, if the buyback stays at or below $6B with CET1 flat and still unexplained, or if the NCO ratio breaks above the 50–55bps through-cycle band. We hold.