Deferral Confirmed: Software Snaps Back to 9% and the Cash Guide Rises, but a 22% Rally Already Paid for the Recovery

Results vs. Consensus

The bar for this print was set by management in July. It cut the full-year revenue guide to 4–5% at constant currency and called the low end an "anchor" that assumed June's buying behavior persisted, while the high end assumed the slipped deals closed. The Street's ~$16.82B revenue and ~$2.86 EPS estimates sat close to that anchor. IBM printed the high end of the July range for the quarter, and management said so: "Our results represent the high end of the range we outlined in July." The beat was real, but it was concentrated in the lines tied to the deferred deals and to hardware.

Q3 2026 Scorecard

Year-over-Year Comparison (Q3 2026 vs. Q3 2025)

Sequential Comparison (Q3 2026 vs. Q2 2026)

Nine Months 2026 (for context)

Revenue

Six percent constant-currency growth is a sharp turn from Q2's 1%, and all three operating segments accelerated. The composition matters more than the headline. Software added four points of growth. Management linked the recaptured deals to about a quarter of that, but on our arithmetic its $250M-plus figure equals about 3.5 points of growth on the year-ago base, and Transaction Processing's +10% ran against a -3% comparison a year ago. Infrastructure's +5% came from Distributed Infrastructure (+32%), with record storage revenue growth, while IBM Z fell 19% against the +59% launch-year comparison. Consulting improved to +2% but missed its estimate. Averaging Q2 and Q3 removes the timing effect: revenue grew about 3.5% across the two quarters at constant currency, which is close to the 4% base case management still anchors to.

Margins and Cash

The productivity engine did the work again. Operating pre-tax margin expanded 70bps to 19.3% and adjusted EBITDA margin 60bps to 28.7%, with segment margins up 210bps in Software (35.0%) and 120bps in Infrastructure (19.3%). The weak spots were gross margin, down 20bps year over year and 90bps sequentially as Infrastructure's mix moved from IBM Z toward lower-margin Distributed hardware and Consulting gross margin fell, and Consulting, where the segment margin fell to 12.2%. Free cash flow of $3.0B rose $0.7B, and nine-month free cash flow reached $7.8B (+8%), which management called the highest nine-month free-cash-flow margin in IBM's reported history. That is what earned the guide raise, and it settles the July worry that the full-year cash target had become back-half-loaded.

EPS

Operating EPS grew 11% to $2.95, almost three points faster than operating pre-tax income because the tax rate fell to 14.7%, partly offset by 0.5% more diluted shares. Much of the beat traces to the recaptured license deals, which carry high incremental margins. That is a clean, operational beat, but it is a one-time recovery of Q2's shortfall rather than a higher run-rate, and it does not change our view of the full year by more than the quarter's own surprise.

Segment Performance

Segment Results — Q3 2026

Total revenue includes $66M of Other revenue ($38M in Q3 2025).

Sub-Segment Growth Trend (constant currency)

Software (+9% cc): the transactional book came back, the recurring book held

Software revenue of $7,804M grew 8.2% reported and 9% at constant currency, with three of four sub-segments growing double digits. The July framing still organizes the segment: roughly 80% recurring revenue (subscriptions, SaaS, support) and 20% transactional, mostly enterprise license agreements (ELAs) tied to the mainframe. In Q2 the transactional book fell high single digits as reported in July (the CFO now describes it as double digits) and cost four to five points of software growth. In Q3 it grew double digits, with ELAs up more than 20% and net revenue retention above 130%. The recurring book grew at a high-single-digit rate, and ARR reached $24.9B, up 8%, unchanged from Q2's growth rate. The segment margin expanded 210bps to 35.0% on productivity.

"Software revenue growth accelerated 4 points sequentially to 9% this quarter, including over 4 points of organic growth." — James Kavanaugh, CFO

Assessment: The quarter answers the July question about the transactional book: the deals were deferred, not lost. It does not yet answer the question about the recurring book, which needs to reach double digits for the long-term software ambition to hold without help from acquisitions. ARR growth stuck at 8% for a second quarter is the number we watch most closely into Q4, when management says recurring growth should be "approaching double digits."

Infrastructure (+5% cc): hardware strength where IBM is not the bottleneck

Infrastructure revenue of $3,707M grew 4.2% reported and 5% at constant currency. Distributed Infrastructure grew 32% on record storage revenue growth across FlashSystem, Storage Fusion, Tape and IBM Z storage, plus Power. IBM Z fell 19% against the +59% launch-year comparison, a smaller decline than Q2's -42%, helped by the new single-frame and rack-mount Z systems that produced the strongest mid-range launch in nearly two decades. The segment margin rose 120bps year over year to 19.3% but fell 250bps sequentially as the mix moved away from Z.

"Storage delivered record revenue growth and continued to gain market share, with broad-based demand across FlashSystem, Storage Fusion, Tape, and IBM Z Storage." — James Kavanaugh, CFO

Assessment: Distributed Infrastructure has become a genuine second engine, and the exit backlog supports Q4. The open question is how much of the storage growth is price rather than volume while memory and component costs are rising across the industry. Management did not split the two, and some on the Street attribute part of the strength to higher component-driven selling prices. Mainframe economics remain intact: installed z17 capacity is up in the high teens six quarters in.

Consulting (+2% cc): revenue up a notch, signings down

Consulting revenue of $5,335M was flat reported and up 2% at constant currency, its fastest rate since Q3 2025, led by Strategy and Technology (+4%). Intelligent Operations was flat, with cybersecurity growing nearly double digits. AI work was about 30% of Consulting revenue and 35% of backlog. The less flattering numbers came from the slides: signings of $5.0B fell 2%, gross margin fell 190bps to 27.4%, and segment profit fell 5% to $653M as IBM reinvested productivity savings into digital assets for a "services as software" model.

"AI is increasingly embedded in broader transformations our clients are undertaking, representing about 30% of our consulting revenue in the quarter and 35% of our backlog." — James Kavanaugh, CFO

Assessment: The revenue line finally moved in the direction the signings promised for two quarters, but the signings themselves turned down, which removes the lead indicator we were leaning on. A trailing book-to-bill of 1.04 is adequate, not strong. Consulting remains low-growth ballast, and the margin decline means it is not even contributing operating leverage this year.

Financing (-3% cc): small and steady

Financing revenue of $192M fell 3.9% reported, and the segment margin fell to 56.1% from 61.6%. The receivables portfolio is almost 80% investment grade, and IBM Financing debt fell to $12.7B at quarter-end from $15.1B at year-end as the portfolio shrank.

Assessment: No bearing on the thesis. The receivables run-off adds to reported operating cash flow but is correctly excluded from free cash flow.

Key KPIs

Guidance & Outlook

Full-Year 2026 Outlook

What the Guide Implies for Q4 2026

Implied figures are our estimates from the full-year guide, nine-month actuals and IBM's Q4 currency assumption. Q4 2025 operating pre-tax margin and free cash flow are derived from full-year 2025 less nine-month 2025.

The shape of the guide repeats July's: management anchors profit, cash and segment expectations to the low end and lets results earn the rest. What changed is the confidence in software, where the low end rose a point and Transaction Processing went from a decline to flat, and in cash, where the target rose to more than $1B of growth. Revenue did not move, and that restraint drew most of the immediate disappointment. The 4% base case implies Q4 constant-currency growth of roughly 3.5%, a deceleration from Q3's 6% even though Q4 carries a heavy 2025 comparison in Infrastructure (IBM Z grew 61% in Q4 2025).

"90 days ago, we provided a full-year revenue growth range with profit and free cash flow expectations anchored to a 4% base case. We continue to take the same approach for the full year, anchoring profit, free cash flow, and segment expectations to that base case." — James Kavanaugh, CFO

Implied Q4 ramp: Software needs only ~7% at constant currency to hit the base case, below Q3's 9%, and Transaction Processing is expected to moderate. Revenue needs ~3.5% for the base case and ~7% for the top end. The demanding line is margin: holding +100bps for the year requires a Q4 operating pre-tax margin near 25.8%, about 180bps above Q4 2025, after 70bps in Q3. That is achievable on IBM's seasonal software mix, but it is the place a miss would show.

Street at: Pre-print consensus sat at Q4 revenue of $20.46B and EPS of $4.62, and at FY2026 EPS of $12.34 on $70.44B. Street revenue sits between the base case and the top end, and Street Q4 EPS sits at about our base-case ~$4.65, so the guide does not force Q4 estimates up. Full-year EPS rises mainly by the Q3 beat, and our FY2026 operating EPS estimate moves to ~$12.45.

Guidance style: Deliberately conservative again, and this time the conservatism has earned some credit back. A quarter ago the cut damaged credibility. This quarter IBM delivered the top of its own range and raised cash, which is the right repair sequence, but the decision to leave revenue unchanged shows management still wants Q4 to prove the run-rate.

Key Topics & Management Commentary

Overall Management Tone: Management was relieved and assertive, clearly more confident than on the contrite July call, and it pressed the point that the deferral debate is over. The posture was still disciplined: the revenue guide stayed anchored at the low end, 2027 was deferred to January, and the CEO volunteered that double-digit Transaction Processing growth will not persist. Management was least forthcoming on Consulting signings and on pricing, the two places where the quarter's numbers were weaker or harder to read.

1. Grading the Q2 Commitments

In July we set eight items to grade this quarter. Five were delivered, two were mixed, and one was not addressed.

Assessment: This is a strong scorecard against the items that mattered most in July, conversion and cash. The two mixed grades, organic software and Consulting, are exactly where the next leg of the thesis has to come from.

2. Deferral, Not Destruction: The Slipped Deals Closed

The central July debate was whether tens of large mainframe-attached software deals had slipped or disappeared. In Q3 about two-thirds of them closed, and management put the slipped total at roughly $400M, the size of the Q2 software miss. The recaptured portion fell mostly into Transaction Processing, with little in Data and Automation, which management said should partly land in Q4.

"The slip deals were tens of slip deals, so let's kind of put that in a ballpark from 2Q of, what, $400 million, which is what we missed software by. We closed two-thirds of those. That's $250-plus million." — James Kavanaugh, CFO

In July management had said one-third had closed in the first three weeks of the quarter and that it expected two-thirds to three-quarters to close over six months. It reached the low end of that range in three months.

Assessment: The bear case of July, that "deferred" was a euphemism, is refuted. This was the single most important question for the stock, and management answered it with numbers. It also confirms that the June disruption was a budget-timing event, which lowers the odds that a similar air pocket repeats.

3. How Organic Is the Software Acceleration?

Software grew 9% at constant currency. Management said organic growth was "over 4 points," which leaves roughly 4.5 points from acquisitions, chiefly Confluent. Of the organic growth, the recaptured deals account for about 3.5 points ($250M against the $7.21B year-ago base). The underlying organic rate, excluding both the catch-up and the acquisitions, is therefore near 1%, consistent with bearish Street estimates. Management offered a different decomposition on the call, attributing about a quarter of the software acceleration to the slipped deals and three-quarters to new business. That split is hard to reconcile with its own dollar sizing.

"I would tell you before all of you ask, about a quarter of it is the slip deals with regards to overall software acceleration, and about three-quarters of it is net new business that we are capturing overall" — James Kavanaugh, CFO

A fairer test averages Q2 and Q3, which nets the slipped deals out. On that basis organic software grew roughly 2% across the two quarters. On a reported basis, which includes Confluent, the recurring book grew at a high-single-digit rate and the transactional book was roughly flat across the two quarters.

Assessment: The acceleration is real at the reported level and flattering at the organic level. That matters because Confluent laps in March 2027. Without a step-up in recurring growth, 2027 software growth will slow toward the organic rate, and that is the central question for the January guide.

4. Transaction Processing: +10%, and Management Says It Will Not Last

Transaction Processing grew 10% at constant currency after falling 9% in Q2, against a -3% comparison a year ago. The CEO explained the spike as the deferred Q2 deals landing on top of a normal Q3 mainframe buying cycle, and he volunteered that double-digit growth will moderate in Q4. TP guidance rose to flat for the full year, from a low-to-mid-single-digit decline that July framed for the second half. The CFO tied future monetization to installed capacity: z17 installed MIPS are up in the high teens and shipped MIPS up more than 100% six quarters into the program.

"A lot of discussions a quarter ago about demand destruction versus timing. I think this puts this to rest right now" — James Kavanaugh, CFO

Assessment: The two-year stack, roughly +7%, is the honest growth rate, and a flat full year is the right expectation. The capacity build is the real asset. TP has become a 2027 growth story again on management's own telling, but this time from a stronger installed base rather than a delay.

5. Red Hat and OpenShift: The Recurring Engine Re-Accelerates at the Bookings Line

Red Hat grew 11% for a second quarter, and bookings growth returned to the high teens. OpenShift ARR reached $2.4B with growth above 30%, which management tied to clients moving AI applications and agents into production. Virtualization contracts reached nearly $800M since early 2024 with about $750M in pipeline. The CEO said IBM is comfortable with double-digit Red Hat growth in Q4 and declined to frame 2027.

"In software, Red Hat again grew 11% with bookings growth accelerating back to the high teens. OpenShift ARR growth accelerated to over 30% as organizations scale super intelligence applications and agents into production" — Arvind Krishna, CEO

Assessment: High-teens bookings are the best forward indicator in the software book, and they support the case that recurring software can approach double digits. Red Hat is no longer a bear point. It is the part of the portfolio most likely to carry growth after the Confluent lap.

6. Data and Confluent: 20% Growth With an Expiry Date

Data grew 20% at constant currency on Confluent, whose bookings grew about 20%. Management positioned Confluent as the real-time data layer for enterprise AI and previewed a cross-sell product, Streamhouse, to be announced within weeks. The CFO said 2026 was about integrating Confluent and 2027 is about growth synergies.

"2026 was about integrating and getting Confluent into our mode of operation and go to market and scale. 2027's around the growth synergies." — James Kavanaugh, CFO

Assessment: Confluent is performing, but its contribution to reported Data growth ends when the deal anniversaries in mid-March 2027. In Q2, organic Data was down mid-single digits. Synergies must replace the acquisition contribution for Data to stay a growth line, and management has not yet quantified them.

7. Infrastructure: Storage Records and the Pricing Question

Distributed Infrastructure grew 32% and storage posted record growth with share gains, which management linked to AI-driven data growth. IBM Z declined 19% but delivered the strongest mid-range launch in nearly two decades with new single-frame and rack-mount systems. The z17 program-to-date comparison against z16 was 125%, slightly below the "nearly 130%" cited in July, still the highest on record.

"Program-to-date performance remains well ahead of the prior generation, up over $1 billion, and this quarter, we had the strongest mid-range launch in IBM Z in nearly 2 decades." — James Kavanaugh, CFO

Assessment: The infrastructure business is healthier than we assumed in July, even after management raised the segment guide then. The caveat is the price component in a market where memory and component costs are rising. If that is a meaningful part of the storage growth, it supports revenue more than units and is less durable than the headline suggests.

8. Consulting: Better Revenue, Weaker Signings, Lower Margin

Consulting grew 2% at constant currency, its fastest rate since Q3 2025, and AI work reached about 30% of revenue and 35% of backlog. IBM added a strategic partnership with OpenAI and expanded its Google Cloud relationship. But signings fell 2% to $5.0B after two quarters of +6%, and the segment margin fell to 12.2% from 12.9% as IBM reinvested productivity into digital assets.

"Consulting segment profit margin declined modestly as we reinvested productivity gains into digital assets" — James Kavanaugh, CFO

Assessment: The reinvestment is strategically sensible: an asset-based services model is the defense against AI-driven labor deflation. But a margin decline plus a signings decline in the same quarter moves the Consulting bear point from contained to emerging. We need signings back above flat in Q4.

9. Productivity and the Q4 Margin Ask

IBM drove 500bps of gross productivity in the quarter, well above its model, and cumulative productivity now exceeds $5.5B. Management's rule is that roughly a third protects margin and two-thirds funds reinvestment. The expense detail shows the trade: base SG&A fell about 8% while base R&D rose about 8%. The full-year +100bps margin target was maintained.

"We continue to expect 100 basis points of operating pre-tax margin expansion for the full year, reflecting the flow-through from the third quarter earnings beat and continued operational discipline." — James Kavanaugh, CFO

Assessment: The productivity pillar is the most reliable part of the IBM model and it held again. The arithmetic on Q4 is demanding, roughly 180bps of expansion after 70bps in Q3, but Q4 is IBM's highest-margin software quarter, and the beat flows through. We treat the margin guide as achievable with little room for error.

10. Free Cash Flow Raised, Balance Sheet Re-Levering Slowly

Free cash flow of $3.0B rose $0.7B, and nine-month free cash flow reached $7.8B, up $0.6B, driven by about $1B of higher adjusted EBITDA offset by higher capital spending, cash taxes and interest. Management raised the full-year target to growth of more than $1B and cited free-cash-flow realization of about 130%. Cash ended at $11.1B, down $3.4B from year-end after $11.4B of acquisitions this year, and total debt was $63.8B, up $2.6B year to date, with non-financing debt at $51.2B.

"Our free cash flow margin for the first 9 months of the year is the highest in reported history, highlighting our strong execution against productivity initiatives and laser focus on operational discipline." — James Kavanaugh, CFO

Assessment: The cash pillar is intact and stronger than in July. Quarterly free cash flow covered the $1.6B dividend about 1.9 times. Leverage is elevated after Confluent, which caps the room for another large deal in 2027 without debt.

11. AI Posture: ROI Realism, Sovereignty and Open-Weight Economics

The CEO used the Q&A to describe a more disciplined phase of enterprise AI. Large clients are scrutinizing ROI and worrying about uncontrolled spend, and they are concentrating on fewer, larger workflow redesigns with integration partners. He described sovereign AI as a fast-rising topic in 2026 and made the economic case for open-weight models on a few racks of on-premise infrastructure at a fraction of frontier-model token costs. The market read the ROI remark as a caution, while management presented it as a tailwind for IBM's integration and infrastructure role.

"To get to sovereign AI, sovereign AI is definitely becoming a much larger conversation." — Arvind Krishna, CEO

Assessment: We read the posture as constructive for IBM. A shift from experimentation to governed, cost-controlled production plays to OpenShift, Confluent, HashiCorp and the mainframe. The risk is that ROI scrutiny lengthens Consulting sales cycles, which the signings decline does nothing to rule out.

12. Optionality: Lightwell and Quantum

Lightwell, the open-source remediation offering introduced last quarter, is now generally available with a pipeline approaching $700M and more than 40,000 remediated packages released. In quantum, IBM finalized a $1B Commerce Department award for its quantum wafer foundry, was selected for Stage C of DARPA's Quantum Benchmarking Initiative, and has signed 16 enterprise engagements this year. A Quantum Day is scheduled for December 10.

"Since then, some of the world's largest enterprises have signed up as clients, and we have built a pipeline approaching $700 million in a short period of time." — Arvind Krishna, CEO

Assessment: Lightwell's pipeline build is fast for a product that reached general availability only this quarter, and it is a good monetization of the Red Hat franchise, though pipeline is not bookings. Quantum remains a long-dated option with rising external validation. Neither moves 2026 or 2027 estimates.

Analyst Q&A Highlights

Recovered Deals or New Demand Behind the Software Rebound

The opening question asked whether the software rebound was just the recapture of deferred transactions or the start of broader demand, and whether high-teens Red Hat bookings signal faster software growth in 2027. Management split the acceleration into roughly a quarter from slipped deals and three-quarters from new business, and committed only to double-digit Red Hat growth in Q4.

Q: "Are you primarily just recovering deferred transactions at this point, or are you actually starting to see infrastructure investments starting to create more incremental demand for IBM software? With Red Hat bookings actually growing at high teens, is that an early indication that software can actually accelerate as you go into 2027?"
— Amit Daryanani, Evercore ISI

A: "I would tell you before all of you ask, about a quarter of it is the slip deals with regards to overall software acceleration, and about three-quarters of it is net new business that we are capturing overall"
— James Kavanaugh, CFO

Assessment: The quarter/three-quarters split is the most bullish framing available, and it is hard to square with the CFO's own dollar sizing of the recaptured deals at $250M-plus. On the $7.21B year-ago base that is about 3.5 points of growth on our arithmetic, though the CFO described it as about a quarter of the four-point acceleration. We put more weight on the dollar figure. The refusal to frame 2027 software growth is consistent with a January reset.

Whether Double-Digit Transaction Processing Can Persist

A detailed question noted that management had guided Transaction Processing down for the second half and called it a 2027 growth vector, yet it grew 10%. Management attributed the spike to deferred deals stacking on a normal Q3 cycle and explicitly guided to moderation.

Q: "Last quarter, you were thinking that TP might be down low to mid-single digits for the second half, and you called it a 27 growth vector. It grew 10% instead."
— Wamsi Mohan, Bank of America

A: "I think what really happened was we got not only the deferral from 2Q, we got a normal 3Q buying cycle on mainframe, so it kind of got it both together. I would be candid and tell you, I do not expect double-digit performance in TP to keep going, but I would expect to see now more normalized performance, especially going into next year, and perhaps a moderation from 3Q going into the fourth quarter."
— Arvind Krishna, CEO

Assessment: This was the most credible answer on the call. Management volunteered the caveat rather than extrapolating the spike, which is exactly the discipline that rebuilds trust after a guide cut.

The Long-Term Double-Digit Software Target and the Role of M&A

A question on the long-term software target asked whether IBM still expects to approach double digits and how much it will rely on acquisitions to get there. Management deferred quantification to January but listed its indicators: recurring software approaching double digits in Q4, new products, 2027 Confluent synergies and mainframe capacity.

Q: "at your latest Analyst Day, you laid out that you would approach double digits in software as a long-term target. Based on what you're seeing, do you still feel confident in that? And how is M&A going to play a role in terms of achieving that?"
— Ben Reitzes, Melius Research

A: "Obviously, and it's not going to surprise you, we'll spend a lot more time on this question here in January. ... We should be approaching double digits here in the fourth quarter on that 80% high value."
— James Kavanaugh, CFO

Assessment: The answer did not address M&A directly. The Q4 recurring-growth marker is now the explicit test for our re-upgrade case, and it is a falsifiable one.

What Drove Margins When Gross Margin Fell

One question asked what drove the margin performance. Gross margin had actually declined 20bps, and the answer moved to operating leverage: 500bps of gross productivity, split between margin protection and reinvestment.

Q: "Gross margins came up with investors. Can you just give us a comment there, what drove the performance there?"
— Raimo Lenschow, Barclays

A: "It's delivered by, we are driving, in the third quarter, 500 basis points of gross productivity in our model. We've always said in our model, about a third is for margin protection, about two-thirds enables us to drive that productivity flywheel where we reinvest back into the business"
— James Kavanaugh, CFO

Assessment: The productivity story is genuine, but the response sidestepped the gross-margin decline, which reflects Consulting gross-margin pressure and a shift within Infrastructure from IBM Z toward Distributed hardware. Margin expansion this year is an expense story, not a pricing or mix story.

Whether AI Hype Is Pushing Buyers Back to Incumbents

A question asked whether clients that had paused to evaluate AI claims were returning after finding the claims overstated. The CEO reframed it as large enterprises becoming more disciplined about AI ROI and spend, choosing fewer, larger projects with trusted partners.

Q: "I am curious if you are maybe seeing some element of this in some of the deals recouping where the claims were big, but perhaps the reality was not matching the claim, and you are able to deliver on that."
— Brent Thill, Jefferies

A: "I think what we are seeing is that at least larger enterprises are becoming a lot more cognizant and a lot more aware of what is the true ROI of all of their AI spend. They are also getting worried if some of the AI spending completely out of control with no real return."
— Arvind Krishna, CEO

Assessment: The market heard caution about AI demand, and the remark coincided with the after-hours decline. In context it was a positioning argument for IBM's integration and governance role. It is consistent with Software strength but could slow Consulting decisions.

Price Versus Volume in the Transaction Processing Recovery

A pointed question asked for a split of the Transaction Processing recovery between quantity and price, and whether IBM had used pricing power on the recaptured deals. The answer explained the capacity-based monetization model and the MIPS growth but did not separate price from volume.

Q: "To the extent you have been able to leverage a little bit of pricing power in this situation, because I know that is not historically a lever that you have used."
— Fatima Boolani, Citi

A: "We got different dynamics playing out right now, so we are about prudently cautious, calling flat for the year. With 100% shipped MIPS and 18% installed MIPS, I would say that is a pretty good future monetization opportunity in 2027 going forward."
— James Kavanaugh, CFO

Assessment: A non-answer on price. In July the CEO said IBM had held price on the recaptured deals. This quarter no one repeated it, which leaves the price contribution to the TP rebound unknown.

Cost Versus Control in Enterprise AI Architecture

The final question asked how enterprises balance AI cost against control. The CEO argued that open-weight models now deliver about 90% of frontier performance at a fraction of the token cost, and that a few racks of infrastructure are enough to start.

Q: "Could you just talk a little bit about how people are addressing the trade-offs, if there are trade-offs between those 2 segments, and how they are architecting their networks in order to maximize for both those things"
— Matt Swanson, RBC Capital Markets

A: "It's a simple question for any enterprise. Would you like something that's 10% of the cost if it's 90% as good as the best, or do you want to pay that 10 times higher cost?"
— Arvind Krishna, CEO

Assessment: This is IBM's strategic bet in one line: a world of cheaper open models run on governed hybrid infrastructure favors Red Hat, watsonx and the mainframe over frontier-model dependence. It is a coherent thesis, not yet a revenue line.

What They're NOT Saying

1. The Consulting signings decline: Signings fell 2% to $5.0B after two quarters of +6%, and neither the prepared remarks nor the Q&A mentioned it. The slides framed it only as a trailing book-to-bill that remains above 1.0. The lead indicator for Consulting turned down, and management chose not to discuss it.
2. Price versus volume on the recaptured deals: Asked directly whether IBM used pricing power, management answered with MIPS growth. In July the CEO said IBM had held price. The silence this quarter leaves the quality of the TP rebound partly unknown.
3. A clean underlying organic software rate: Management gave "over 4 points" of organic growth and $250M-plus of recaptured deals, but no organic rate excluding the catch-up, and its quarter/three-quarters split does not reconcile with the dollar sizing. Our derivation puts the underlying figure near 1%.
4. Whether the hardware crowd-out has ended: The opening question asked what had changed since clients prioritized infrastructure over software purchases in June. Management answered with AI application demand for Red Hat and Confluent and did not say whether the pressure from securing servers, storage and memory ahead of price increases has eased. Hardware demand clearly stayed strong, and IBM itself benefited in storage, but software budgets may still be exposed to it.
5. Automation outside HashiCorp: Automation fell 3% while management said HashiCorp revenue growth accelerated sequentially, which implies the rest of the sub-segment declined. There was no split, and the full-year guide is only low-single-digit growth.
6. The z17 ratio drift: The program-to-date comparison moved to 125% from "nearly 130%" in July without comment. It is still the best cycle on record, but the direction is down.
7. 2027, and life after the Confluent anniversary: Management declined to frame 2027 growth until January. With Confluent lapping in mid-March 2027, the question of what software growth looks like without the acquisition contribution is the biggest unaddressed issue for next year's estimates.

Market Reaction

• Pre-print setup: IBM closed at $251.46 on October 21, up 3.9% in the session before the after-close release and up 22.2% from the July 22 52-week closing low of $205.77. Entering the print the stock was down 15.1% year to date and 10.8% over twelve months, and up 8.4% over the trailing 30 days, against an S&P 500 up 13.3% year to date. The 52-week closing range was $205.77 to $329.23.
• After hours (October 21): Shares fell about 3.7% in extended trading, to roughly $242, as the unchanged revenue guide and the CEO's comments on AI ROI drew attention.
• Reaction session (October 22): The stock opened at $241.25 (-4.1%), traded between $240.00 and $256.30, and closed at $252.11, up 0.3% (+$0.65), on 11.6M shares, 2.0x the 30-day average. The S&P 500 fell 1.2%, and Microsoft (-2.2%) and Oracle (-2.8%) also declined, so IBM outperformed the index by about 1.5 points.
• Since our July note: From the July 23 reaction close of $206.65, IBM gained 22.0% to the October 22 close against +3.4% for the S&P 500. It closed at $253.56 on October 23.

A run-up that pre-paid the print. The stock rallied 22% from its July low into the release, including 3.9% on the day of the print itself, so the market had already priced a good part of the deferral recovery. An unchanged revenue guide gave momentum buyers no reason to chase after hours, and the 4% gap-down at the open was the sound of that positioning clearing.

The intraday reversal. The stock recovered 5.0% from its intraday low to close higher on a day when the index and large-cap software fell. Buyers looked through the unchanged guide to the free-cash-flow raise, the Transaction Processing recovery and the narrowed software range. That is the behavior of a holder base that now trusts the cash flow again, which also means the easy re-rating has happened. From here the stock needs earnings growth, not relief, to move higher.

Street Perspective

Debate: Is the Software Rebound Durable or a One-Quarter Catch-Up?

Bull view: The bull case holds that IBM resolved the deferral faster than promised, three of four software sub-segments grew double digits, ELAs grew more than 20% with net retention above 130%, and Red Hat bookings in the high teens point to accelerating recurring revenue into 2027.

Bear view: The bear camp argues that the recaptured deals added about 3.5 points and acquisitions about 4.5, leaving underlying organic software growth near 1%, and that ARR growth stuck at 8% shows no acceleration in the recurring base. With Confluent lapping in March, 2027 software growth could fall below the high-single-digit expectations now in consensus.

Our take: The bears have the better arithmetic for 2026 and the bulls have the better lead indicator for 2027. Red Hat bookings and OpenShift growth are real, but the organic number has not yet proven itself. Q4 recurring growth "approaching double digits" decides it.

Debate: Is Infrastructure Strength Volume or Price?

Bull view: Optimists see Distributed Infrastructure growth above 30% for two straight quarters, record storage revenue growth, share gains and a strong mid-range Z launch as evidence that AI data growth is pulling IBM hardware, with a sizable backlog into Q4.

Bear view: Skeptics argue that part of the storage and server growth reflects higher component-driven selling prices in a tight memory market, which flatters revenue, compresses gross margin and will reverse when component prices normalize.

Our take: Both are probably true. Infrastructure gross margin fell 170bps sequentially to 56.7% as the mix moved toward Distributed hardware, which is consistent with pass-through component costs. We credit the share gains but haircut the growth rate when modeling 2027.

Debate: After a 22% Rally, Is IBM Still Cheap?

Bull view: The bull camp argues that the July crash over-discounted a timing issue, that free cash flow is at a record margin and growing more than $1B, and that 19x forward earnings with a ~6.6% free-cash-flow yield is reasonable for a business compounding high-single-digit EPS with a 2.7% dividend.

Bear view: The bear camp notes the stock has re-rated from 17x to 20x this year's earnings in three months on a quarter that held the revenue guide, and that the decade-long pattern of IBM rallies fading on mid-single-digit growth argues against paying up before 2027 is framed.

Our take: The re-rating was justified, and we were too cautious in July. But at ~19x FY27 the stock now sits inside our fair-value range, and our expected return roughly matches the market's. Valuation is neutral, not a reason to buy.

Debate: Can Consulting Monetize AI, or Is AI Eating It?

Bull view: Supporters point to AI at about 30% of Consulting revenue and 35% of backlog, the fastest revenue growth since Q3 2025, new partnerships with OpenAI and Google Cloud, and nearly double-digit cybersecurity growth.

Bear view: Critics focus on the 2% signings decline, a lower margin and flat Intelligent Operations, arguing that AI is deflating traditional services work faster than new AI engagements replace it.

Our take: One quarter of signings does not make a trend, but the reversal from +6% removes the support we cited in July. Consulting is a drag on the multiple until signings recover.

Model Update & Valuation Framework

Valuation impact: At the October 23 close of $253.56, IBM trades at 20.4x our ~$12.45 FY26 operating EPS estimate, up from 17.2x at our July note on the same-year basis, and at 18.9x our ~$13.40 FY27 estimate. The free-cash-flow yield on ~$15.8B is about 6.6%, down from roughly 8% in July, and the dividend yields 2.7%. We roll the valuation to FY27 and apply 18–20x. July's 16–18x applied to FY26, and our new $255 midpoint equals about 20.5x FY26, so on a same-year basis the multiple we accept rose by about three and a half points. That reflects repaired guidance credibility, still capped by mid-single-digit revenue growth and the Confluent anniversary. That gives a fair value of $240–270. The $255 midpoint sits about 0.6% above the share price, and with the dividend the expected total return is roughly 3%, close to what we expect from the market. The upside case, 22x on ~$13.60 if 2027 software growth holds near double digits, reaches about $300. The downside case, 16x on ~$13.00 if the Q3 catch-up flattered the run-rate and 2027 slows, lands near $208. The skew is balanced, which is a Hold.

Thesis Scorecard Post-Earnings

Overall: The thesis strengthened on fundamentals and weakened on price. The pillars that broke in July recovered: the deferral resolved, Transaction Processing returned to growth, the July crowd-out did not recur, and management repaired its credibility with a quarter at the top of its own range and a cash raise. The pillars that carried us through July, productivity and free cash flow, held again. Two risks rose: Consulting signings turned down, and valuation moved from support to neutral after a 22% rally. The M&A-assisted-growth question is unresolved and becomes the key issue for 2027 as Confluent laps.

Action: Maintain Hold; fair value raised to $240–270 (18–20x FY27E operating EPS of ~$13.40), conviction 6/10. We were right about the business in July and wrong about the stock: deferral, not destruction, proved correct, and the shares rose 22% against +3% for the index while we waited for proof. That proof has now arrived, and the price reflects it. We would move to Outperform if Q4 recurring software reaches double digits and January frames 2027 software growth near high single digits despite the Confluent anniversary, or if the stock pulls back toward $215–225 (about 16–17x FY27E) with the thesis intact. We would move to Underperform if Q4 software lands below the ~7% base case, Consulting signings decline again with revenue turning negative, the margin or free-cash-flow guide is missed, or IBM funds a large acquisition with debt. Going forward we watch Q4 ARR and recurring growth, Consulting signings, the Q4 margin step-up, the price component in storage, the Streamhouse launch, the December 10 Quantum Day and the January 2027 outlook.