The Guide-Raise IBM Withheld in Q2 Finally Arrives, and the Dip Gets Bought

Results vs. Consensus

Quality of Beat/Miss

• Revenue: High quality and, for the first time this year, accompanied by a top-line guide raise. The +7% constant-currency growth is the fastest in several years, and management was explicit that all segments accelerated sequentially. Software's three-point organic acceleration came with declining inorganic contribution, which is the cleanest possible composition.
• Margins: Operating gross margin expanded 120bps and operating pre-tax margin 200bps, the ninth consecutive quarter of operating pre-tax margin expansion. Adjusted EBITDA grew 22%. Consulting margin hit a three-year high on the software-led delivery model. None of this is one-time.
• EPS: The $2.65 operating figure is the clean read. GAAP EPS of $1.84 is depressed by a one-time, non-cash income-tax charge tied to the July 2025 enactment of H.R.1, and the year-ago GAAP base carried a $2.7B pension-settlement charge, so GAAP year-over-year comparisons are not meaningful this quarter. Use operating EPS.

Year-over-Year and Sequential Detail

The acceleration is clearest at the IBM level: constant-currency growth moved from +5% in Q2 to +7% in Q3. Sequentially the headline revenue declined (from $16.98B to $16.33B) because Q2 carried the z17 launch surge in Infrastructure, but that is seasonality, not deceleration. The signal is the year-over-year line, where every segment grew faster than it did a quarter ago, and Software's improvement was organic.

Segment Performance

Software — organic acceleration, with Red Hat cooling

Software revenue of $7.209B grew 10% (9% at constant currency), and management was unusually specific that the roughly three-point organic acceleration was genuine: inorganic contribution actually fell as prior acquisitions anniversaried. The standout was Automation, up 22% at constant currency, with HashiCorp delivering its highest bookings quarter in history. Data grew 7%. The two soft spots: Red Hat decelerated to 12% cc (from 14%) as exceptional prior-year RHEL consumption wraps, and Transaction Processing fell 3% cc for a second quarter.

"Software accelerated about three points organically quarter to quarter. This wasn't an inorganic contribution. In fact, our inorganic contribution came down as we wrapped on some of these. It's being driven by the strong contribution of our high-value recurring revenue, now a book of business, $23 billion, up 9%." — James Kavanaugh, CFO

The $23.2B ARR base (80% of software) growing 9% and expected to exit Q4 at double digits is the structural support. Red Hat's deceleration is presented as a comp issue, with ~20% bookings growth and a mid-teens revenue-under-contract base argued to re-accelerate it toward mid-teens in 2026.

Assessment: This is materially better than Q2 on the one axis that matters most for the re-rating: organic. The clean organic acceleration with falling inorganic contribution is exactly what the bull thesis needed. The watch items narrow to Red Hat (is the deceleration just a comp, or a trend?) and TP (will it actually inflect?).

Infrastructure — IBM Z's best Q3 in nearly two decades

Infrastructure revenue of $3.559B grew 17% (15% cc), with Hybrid Infrastructure up 26% cc and IBM Z up 59% cc, its highest third-quarter revenue in nearly twenty years. Distributed Infrastructure turned positive (+8% cc) on broad storage strength as clients scale capacity for AI, reversing the Q2 trough. Segment profit margin expanded 420bps. Management raised the full-year infrastructure contribution to "over 1.5 points" of IBM growth.

"IBM Z delivered its highest third-quarter revenue in nearly two decades, up 59% year to year, fueled by the early success of our z17 platform... we're at a 130% program... higher capacity opportunity creates higher monetization opportunity creates higher price opportunity." — James Kavanaugh, CFO

Assessment: The z17 cycle is now demonstrably stronger than z16 (130% of the prior program's MIPS, the most successful in IBM history), and the storage recovery confirms AI as a tailwind for the legacy hardware stack. The stack-multiplier thesis from our initiation is intact and, if TP inflects in Q4 as guided, about to start paying off in software.

Consulting — the inflection arrives

Consulting revenue of $5.324B returned to growth, up 3% reported and 2% at constant currency, which management called a positive inflection point after several flat-to-down quarters. Both lines of business showed sequential momentum (Intelligent Operations +4% cc; Strategy & Technology stabilizing). Segment margin expanded 200bps to a three-year high on the software-driven delivery model, and the consulting GenAI book of business crossed $1.5B in the quarter, now over 12% of segment revenue at a 2-3 point margin premium.

"Consulting returned to growth in the third quarter with revenue up 2%, improving sequentially and marking a positive inflection point in performance... reflected in the 220 basis points of segment profit margin expansion year to date." — James Kavanaugh, CFO

Assessment: This directly resolves the second of our three initiation watch items. The "green shoots" from Q2 became actual constant-currency growth, and the margin story (highest in three years) is genuinely strong. The durability question (does AI keep cannibalizing legacy work faster than GenAI projects ramp?) remains, but the inflection is real.

Key KPIs

Key Topics & Management Commentary

Overall Management Tone: The most confident of our short coverage history, and for once the numbers earned it: management led with "revenue growth accelerated to 7%, our highest growth in several years" and backed it with a rare in-year revenue-guide raise. Where they leaned forward most was 2026, framing this year's organic acceleration as a stepping stone rather than a peak. The only place the message stayed careful was Transaction Processing, where the recovery is still a forward call.

1. The Revenue Guide Raise

The single most important change from Q2 was structural, not tonal: IBM raised its full-year constant-currency revenue-growth guide to "more than 5%" and lifted free cash flow to about $14B. In Q2, management had explicitly declined to raise revenue despite raising cash and margin, which we flagged as the loudest thing left unsaid. This quarter they said it.

"Given the strength of this performance, we are raising our expectations for revenue, adjusted EBITDA, and free cash flow. We now expect to deliver revenue growth of more than 5%... and free cash flow of about $14 billion for 2025." — James Kavanaugh, CFO

Assessment: This is the confirmation our initiation called for. Raising the top-line number, not just cash and margin, is management putting its name behind the acceleration. It is the most important single data point in the quarter.

2. Organic vs. Inorganic Software

Management pre-empted the Q2 criticism by quantifying the organic acceleration directly: about three points of sequential improvement, with inorganic contribution falling. For 2025, they framed organic software growth at roughly 6%+, with three-to-four points of inorganic from the HashiCorp-heavy year.

"When you take a look at 2025, you go do the math... probably gonna be approaching six plus percent organic growth overall. We're gonna have somewhere three to four points this year because we took advantage of a very strategic opportunity with HashiCorp." — James Kavanaugh, CFO

Assessment: Directly addresses the Bear #1 risk from our scorecard. Organic software at ~6% growing and inorganic declining is a healthier mix than the headline 9-10% reported growth implied. The HashiCorp wrap in spring 2026 is the next test of whether organic can carry the line.

3. Transaction Processing: Guided to Inflect

TP fell 3% cc again, the one segment still going backward, but management drew a tight analogy to the z16 cycle (flat in 2022, high-single in 2023, double-digit in 2024) and argued z17's stronger start pulls the recovery forward, with TP returning to growth in Q4.

"If you map it back to the z16 cycle... in '22 our TP revenue was flat. In '23 we grew high single digit. In '24 we grew double digit. Look at '25 right now, we're calling back to growth probably a quarter early compared to a historical cycle." — James Kavanaugh, CFO

Assessment: The cycle precedent is real and the 130%-of-z16 MIPS figure supports it, but this is still the one thing investors have to take on faith. A Q4 TP that actually inflects positive would remove the last reservation in our thesis.

4. AI Book of Business Crosses $9.5B

The GenAI book of business jumped to more than $9.5B inception-to-date, up roughly $2B sequentially, with the consulting piece adding $1.5B in the quarter alone and the number of consulting GenAI projects more than doubling year over year. Management quantified that GenAI contributed about two points of the ~9% software growth.

"Our GenAI book of business continues to show momentum, at over $9.5 billion inception to date... About two points of that [software] growth is coming out from our GenAI book of business." — Arvind Krishna, CEO / James Kavanaugh, CFO

Assessment: The acceleration in the AI book, and the fact that it is now a quantified two-point contributor to software growth, is the clearest evidence yet that IBM is monetizing AI in the P&L, not just signing it. This is the strongest leg of the bull case.

5. The 2026 Setup

Management spent unusual airtime on 2026, sketching a path to sustained-or-better growth: Red Hat re-accelerating toward mid-teens on 20% bookings, TP turning to low-single-digit growth, Automation staying double-digit even as HashiCorp wraps, Data mid-to-high single digit, plus the usual two-to-three points of inorganic. The ARR base exiting Q4 at double digits is the forward indicator they leaned on hardest.

"Over a $23 billion ARR book of business, [we] feel we're gonna exit the fourth quarter at double digits. That's a great indicator for 2026 because that is 80% of our software portfolio." — James Kavanaugh, CFO

Assessment: The 2026 bridge is detailed and internally consistent. If even most of it lands, IBM is a structurally faster grower next year than the market models, which is the core of any upgrade case. The risk is that it again leans on a TP recovery and an M&A contribution that have to show up.

6. AI Partnerships and the Hyperscaler Question

IBM positioned itself as a beneficiary, not a victim, of hyperscaler AI growth: it partners with all the major clouds (and is a large CoreWeave client), deploys its software stack on others' infrastructure including Groq/xAI, and frames its mainframe and storage as direct AI beneficiaries (a fully-populated z17 does 450 billion inferences per day). New partnerships with Anthropic (Claude into IBM products) and a Granite 4.0 model launch reinforce a model-agnostic posture.

"We are actually the direct beneficiary of the hyperscaler growth of AI capability and capacity as enterprises use this capability... our biggest beneficiary of AI infrastructure is our mainframe, and our storage portfolio." — Arvind Krishna, CEO

Assessment: A credible reframing that turns the usual "IBM is not a hyperscaler" bear point into a tailwind. Whether enterprise inferencing on Z is a material revenue driver or a talking point is unproven, but the storage AI-attach is already showing up in Distributed Infrastructure's return to growth.

7. M&A Discipline and Capacity

Management reiterated mid-$20B-plus of M&A capacity over a three-year window and a disciplined three-part screen (strategic fit, growth synergy above standalone, cash-accretive by year two). It would not rule out a larger transformative deal but insisted size alone is never the rationale.

"Anything that is of size has to fit three criteria. It has to fit with the strategy... there has to be synergy... if it is of size, then we are very disciplined also that we like it to become accretive to cash by the end of the second year." — Arvind Krishna, CEO

Assessment: The discipline is reassuring, but the standing M&A appetite is the double-edged part of the story: it funds growth synergies yet keeps reported software growth ahead of organic and keeps cash flowing to deals rather than buybacks. A large deal would be a swing factor in either direction.

Guidance & Outlook

This is the inverse of Q2's guidance shape. Last quarter everything moved up except revenue; this quarter revenue moved up too, and the only sub-line still carrying a "watch" is TP, which is now guided to inflect in the very next quarter. Software remains "approaching double digits" for the year with a Q4 reported acceleration toward double-digit, and Consulting Q4 is guided similar to Q3.

Implied Q4 ramp: To hit the raised full-year marks, Q4 needs Software reported growth to step up toward double digits (helped by easier RHEL comps wrapping), TP to print positive, and Consulting to hold its inflection. Street at: management is "comfortable with consensus" for Q4 cc revenue and profitability, capping blow-out risk but de-risking a miss. Guidance style: a clear shift from the conservative Q2 stance to a more forward-leaning one, consistent with a management team that now has the acceleration in hand rather than in forecast.

Analyst Q&A Highlights

Free cash flow durability into 2026

The opening question pressed on whether the ~$14B free-cash-flow guide and ~125% conversion are sustainable into 2026 or flattered by one-offs. Management leaned on the operating-leverage engine and reiterated a mid-to-high-120s realization rate it has hit for four straight years.

Q: "This sort of implies free cash flow is up double digits in '25 and your conversion rates are around 125%... is there anything that could preclude free cash flow from growing a few points higher than sales growth?"
— Amit Daryanani, Evercore ISI

A: "Those underlying fundamentals... deliver a sustainable realization number... in the mid to high one twenties... By the way, we've been there for four years in a row already. So we can handle that."
— James Kavanaugh, CFO

Assessment: A confident, evidence-backed answer. The free-cash-flow engine is the most reliable part of the IBM story, and the four-year track record on conversion makes the ~$14B credible rather than aspirational.

Software organic growth and the Transaction Processing inflection

The dominant line of questioning sought to pin down the real organic software rate and whether TP will actually turn. Management quantified the organic acceleration and committed, again, to a Q4 TP recovery on the z17 cycle.

Q: "If you could just clarify the organic growth in software in the third quarter and expectations for transaction processing going into the end of the year?"
— Wamsi Mohan, Bank of America

A: "We accelerated about three points organically quarter to quarter. This wasn't an inorganic contribution... calling a return to growth in TP in the fourth quarter with the strong pipeline we got."
— James Kavanaugh, CFO

Assessment: This is the exchange that should have reassured the early sellers. Organic acceleration with declining inorganic, plus a specific TP recovery call, is the substance behind the dip-buying. It is also a checkable commitment for next quarter.

Can software accelerate into 2026 as HashiCorp wraps?

An analyst asked whether software can sustain or accelerate from double-digit reported growth once the HashiCorp acquisition anniversaries in spring 2026, given Red Hat's deceleration. Management walked the four software lines and argued for acceleration, not deceleration.

Q: "You do wrap the Hashi acquisition in the spring... Are there signs that it can accelerate from here? Obviously with Red Hat decelerating a little... can you keep double-digit next year or even accelerate?"
— Ben Reitzes, Melius Research

A: "As that becomes the bulk of the Red Hat book of business entering 2026, we do expect to see Red Hat returning to mid-teens or close to mid-teens growth. So that would be an acceleration... I expect to see TP grow... low single digits for sure into next year."
— Arvind Krishna, CEO

Assessment: A forward-leaning answer that frames 2025 as a trough year for organic software. The Red Hat re-acceleration rests on the 20% bookings number converting to revenue, which is the single most important 2026 driver to track.

The hyperscaler AI opportunity

An analyst probed whether the explosive growth in cloud AI infrastructure is a threat or an opportunity for IBM. The CEO reframed it as a tailwind, positioning IBM as a software-and-services layer riding on others' infrastructure plus a direct hardware beneficiary.

Q: "We've seen cloud providers experience exceptional growth recently, particularly in infrastructure services and large-scale AI workloads. How does IBM view that trend? And do you see a similar opportunity for IBM Cloud?"
— Eric Woodring, Morgan Stanley

A: "We are actually the direct beneficiary of the hyperscaler growth of AI capability and capacity as enterprises use this capability... our biggest beneficiary of AI infrastructure is our mainframe, and our storage portfolio at this time."
— Arvind Krishna, CEO

Assessment: The reframing is strategically sound and already partly visible in the storage recovery. It neutralizes a standard bear point, though enterprise inferencing on Z remains more thesis than line item today.

Enterprise AI adoption: how early are we?

A closing question on enterprise AI mix (cloud-native point solutions versus self-hosted integration) and where IT budgets are flowing drew a "first innings" framing from the CEO and a useful read on budget dynamics.

Q: "Could you generalize what you're seeing with regard to mix as enterprises focus on AI readiness?... within the IT budgets, where is the spending coming from? What's at risk of getting trimmed?"
— Brian Essex, JPMorgan

A: "We [are] in the first innings of enterprise AI rollout... IT budgets are growing typically two to three points ahead of GDP growth... the ratio used to be seventy thirty. 70 are running what is, 30 are new. I think that is shifting more towards the sixty forty spread."
— Arvind Krishna, CEO

Assessment: The budget-mix shift toward "new" spend is the structural tailwind under IBM's automation and hybrid portfolios. The "first innings" framing is self-serving but consistent with the AI-book acceleration the numbers show.

What They're NOT Saying

1. A Red Hat re-acceleration number with conviction: Management says "mid-teens or close to mid-teens" for 2026, hedged with "close to" and "low end." Red Hat decelerating to 12% cc while bookings run 20% is presented as a comp issue; if it is something more, it undercuts the software bridge.
2. Buybacks, still: Year-to-date the company returned $4.7B in dividends and zero meaningful buyback; the share count keeps drifting up. With M&A capacity reiterated at mid-$20Bs, capital allocation remains tilted to deals and dividends, and management again did not address repurchases.
3. A clean GAAP bridge: The one-time H.R.1 tax charge is mentioned but not quantified line-by-line in the release narrative, leaving GAAP EPS of $1.84 to look optically weak against screens. Management steered entirely to operating EPS.
4. What happens if TP does not inflect: The Q4 TP-recovery call is stated with confidence and a cycle analogy, but there is no contingency framing for what a second-half-of-cycle slip would mean for the software guide.

Market Reaction

• Pre-print setup: IBM closed at $287.51 on 2025-10-22, up 30.8% year to date and 5.9% over the trailing 30 days, near the top of its 52-week range ($204.90 - $294.78). The stock had held its post-Q2 level despite two more quarters of execution, so the multiple had not expanded further.
• Reaction session (2025-10-23): Shares gapped down to open at $264.95 (-7.8%) and traded as low as -8.3%, then recovered through the session to close at $285.00, down just 0.9% ($2.51), on volume of 16.7M (3.4x the 30-day average). The S&P 500 closed +0.6%.

The round-trip is the signal. The open priced the same reflex that drove July's selloff (software not beating, Red Hat decelerating, TP still negative), but the close reflected the substance that emerged on the call: accelerating revenue, a raised top-line guide, a $9.5B AI book, and a Consulting inflection. A stock that absorbs a 7.8% gap-down on news that would have sent it down 7% a quarter earlier is telling you the marginal holder has changed. It does not, however, make the stock cheap.

Street Perspective

Debate: Did the revenue-guide raise confirm the growth re-rating?

Bull view: Raising the top-line guide (not just cash and margin) with broad sequential acceleration and a $9.5B AI book is management validating that IBM is now a 5%+ grower; the re-rating to the mid-20s multiple is earned.

Bear view: "More than 5%" is hardly a hyper-growth number, organic software is still only ~6%, and the reported growth leans on HashiCorp that wraps in spring; the multiple already prices a durability that 2026 has to prove.

Our take: The bulls won this quarter on the facts, the guide raise was the missing confirmation, but the bears still own the valuation. We move closer to their growth case while holding their valuation discipline, which is precisely a Hold with rising conviction.

Debate: Is Transaction Processing about to inflect or about to disappoint?

Bull view: z17 is running at 130% of the most successful prior cycle; the z16 precedent (flat then high-single then double-digit) plus a strong pipeline makes a Q4 TP return to growth nearly mechanical.

Bear view: TP has now been negative for two straight quarters and the recovery has been promised before; if Q4 TP stays negative, the whole software-acceleration bridge wobbles.

Our take: The cycle mechanics favor the bulls and the MIPS data is hard to argue with, but until TP actually prints positive it is the single open question in the thesis. Q4 resolves it.

Debate: Is the AI book of business real revenue or a vanity metric?

Bull view: $9.5B inception-to-date, accelerating $2B a quarter, now a quantified two-point contributor to software growth and 12% of consulting revenue at premium margins; this is monetization, not marketing.

Bear view: "Book of business" blends software revenue, SaaS ACV, and consulting signings, so it overstates near-term revenue, and consulting GenAI may simply be cannibalizing legacy consulting.

Our take: The metric is fuzzy but the trend and the quantified P&L contribution are not. We weight the bull case here; the AI book is the strongest single leg of the thesis.

Model Update Needed

Valuation impact: At $285.00, IBM trades around 26x consensus FY2025 operating EPS near $11.00, essentially unchanged from the post-Q2 level despite a quarter of confirming execution. The multiple is full and at a multi-year high, but the earnings and free-cash-flow base it capitalizes are now growing faster and de-risked by the guide raise. We see fair value modestly above the current price and would turn outright constructive on either a TP inflection in Q4 or any multiple give-back, which would shift the risk/reward decisively.

Thesis Scorecard Post-Earnings

Overall: Thesis strengthened. The revenue-guide raise and the Consulting inflection resolved two of the three open questions from our initiation, and the AI book accelerated. Only valuation and the not-yet-delivered TP inflection keep this from being an upgrade.

Action: Maintain Hold, conviction rising. We move to Outperform on (1) Q4 Transaction Processing printing positive as guided, confirming the stack-multiplier flywheel, or (2) any multiple give-back that improves the entry. We would reconsider if Red Hat's deceleration proves structural rather than a comp, or if the software guide slips as HashiCorp wraps.