The Backlog Missed by 70bp and the Stock Lost 15%; At 23x With a €10B Buyback, We Upgrade

Results vs. Consensus

SAP reports in euros; US desks consensus the NYSE ADR in dollars. Estimates below are translated at the 1.1971 USD/EUR spot that prevailed on the reaction day, a rate that reconciles the dollar-quoted actuals to the euro-reported income statement to the cent, so estimate and actual sit on one FX basis.

Fourth Quarter Year-over-Year

Full Year 2025

FY2025 Delivered Against Its Own Outlook

The most useful single table in the release is management's scorecard against the guidance it set in January and revised in October. It is the reason this quarter reads worse on the tape than it reads on the page.

Five of six lines met or beat. Two of them, operating profit and free cash flow, finished outside the guided range to the upside. The sixth, current cloud backlog, is the one the market traded, and it is fair to note that "slightly decelerate" turned out to mean a four-point decline from 29% to 25%. A guidance word that permits a four-point move is not a guidance word.

Quality of Beat/Miss

• Revenue: Q4 total revenue of €9,684M grew 3% reported and 9% at constant currency, with a 600bp translation drag that reflects a euro that appreciated through 2025. Cloud revenue grew 26% at constant currency, a one-point deceleration from Q3's 27%, which is consistent with the phasing effects flagged in October rather than with any change in demand. Software licences fell 34% to €0.45B and services fell 4% to €1.07B, flat at constant currency. The decisive detail is that the shortfall the market punished does not appear in this line at all.
• Margins: Non-IFRS operating margin expanded 320bp to 29.2%, and did so while absorbing approximately €0.1B of workforce transformation expense that management had pre-announced at Q3 and which landed at exactly the flagged size. IFRS operating profit additionally absorbed approximately €0.2B of Teradata litigation expense, which is why IFRS margin expanded 490bp on a base depressed by the prior-year restructuring. Non-IFRS cloud gross margin reached 74.6%, up 110bp, and 75.0% for the full year, up 170bp. The margin story is intact and ahead of where we modeled it.
• EPS: The Q4 non-IFRS EPS beat is genuine but flattered at the tax line. Non-IFRS EPS grew 16% to €1.62 against non-IFRS operating profit growth of 16% reported, so there is no below-the-line rescue in the quarter itself. The full-year picture is different and worth isolating: FY non-IFRS EPS grew 36% against non-IFRS operating profit growth of 28%, and roughly a third of that eight-point gap is the tax rate coming in at 30.4% against a guided 32%. Management attributed the improvement to an increasing ability to offset foreign withholding taxes in Germany, and then guided the midterm rate down to 28–30%, which makes the benefit structural rather than a one-quarter windfall.

Segment Performance

SAP does not report segment operating profit, so the analysis below works from the revenue-line disclosure. The Q4 composition is the clearest illustration yet of the structural trade running through this business: a cloud franchise compounding at 26 to 30% against a licence and support base decaying fast enough to hold total revenue growth at 9%.

Derivation note: support revenue is not separately disclosed. The figures above are cloud and software revenue less cloud revenue less software licences. The full-year derivation produces €10,525M, which management independently confirmed on the call as a support revenue base of "EUR 10.5 billion." That agreement is the reason we are comfortable carrying the derived quarterly figure.

The Two Backlogs — Where the Quarter Was Decided

SAP discloses two backlog measures and they diverged sharply this quarter. Current cloud backlog, the contracted cloud revenue expected to be recognized in the next twelve months, reached €21.05B, up 16% reported and 25% at constant currency. Total cloud backlog, the full contracted value across all future periods, reached €77.29B, up 22% reported and 30% at constant currency.

"Current cloud backlog reached EUR 21 billion, up 25%. This is a more pronounced slowdown than what we had anticipated and more than the slight deceleration we guided to at the beginning of last year." — Dominik Asam, CFO

The five-point gap between the two measures is the entire analytical story. Contracted business grew faster than the twelve-month recognition window, which is arithmetically what happens when a quarter is weighted toward large transformations whose revenue ramps in years two through four.

Assessment: A widening total-versus-current spread is bullish for the medium term and bearish for the metric the market prices off. Both statements are true, and the market resolved the tension in favor of the near-term measure. We think that is the wrong resolution at this price, because the twelve-month window is a reporting convention while the contracted value is the economics. The check on this view is renewal behavior in 2027 and 2028 when those ramps land, which is not testable for two years.

Cloud ERP Suite — Sixteen Quarters Above 30%

Cloud ERP Suite grew 30% at constant currency in the quarter and 32% for the full year, reaching €18,119M and 86% of full-year cloud revenue. That extends the streak of quarters above 30% constant-currency growth to sixteen. Management again pressed the point that a euro-reporting company's constant-currency figure understates its position against dollar-reporting competitors.

"Cloud ERP Suite had another notable year, reinforcing its position as a key engine of growth with an increase of 32% in 2025. And I want to make the remark, this is a constant currency number. In US dollar, the number would be 2 percentage points higher if you want to compare to competition." — Dominik Asam, CFO

Assessment: The core franchise did not slow. At a €18.1B annual run rate growing above 30%, this remains the most consistent large-scale growth asset in enterprise software. The quarter's disappointment was entirely about the timing of recognition on new contracts, not about the performance of the installed cloud base.

Licences and Support — The Decay Now Has a Guide

Software licences fell 31% at constant currency in the quarter and 27% for the year to €0.99B. Support, derived at roughly €10.5B for the year and confirmed at that level by management, is the larger and more consequential line. The change this quarter is that the decline is no longer only described retrospectively; it is now formally guided.

Management framed the offsetting opportunity in conversion terms: customers representing 40% of the support revenue base have now initiated a move to cloud ERP, and the stated conversion multiplier on that base is two to three times.

"Second, we will continue converting our installed base to the cloud with a multiplier of 2-3x. Considering our support revenue base of EUR 10.5 billion, this represents a multi-billion euro cloud revenue opportunity for us." — Christian Klein, CEO

Assessment: Guiding the support decline to accelerate is honest and, in the same breath, is the clearest statement yet of why total revenue growth cannot accelerate quickly. A €10.5B base declining at a rate management expects to steepen is a growing subtraction from the top line. The conversion multiplier is the bull answer and it is credible, but it converts support euros into cloud euros over a multi-year ramp, which is exactly the outer-year phasing that just cost the stock 15%.

Regional Performance

Fourth-quarter cloud revenue performance was described as particularly strong in APJ and EMEA and solid in the Americas. Brazil, Canada, Germany, India, Italy, South Korea, Spain and the United Kingdom were called out as outstanding; Australia, Japan, Mexico, Saudi Arabia, Singapore and the United States as particularly strong. For the full year the outstanding list was Brazil, France, Germany, India, Italy, South Korea and Spain, with China, Japan, Saudi Arabia, the United Kingdom and the United States particularly strong.

Assessment: The Americas retain the softest descriptor for a second consecutive quarter, though the United States moved into the "particularly strong" list on both the quarterly and full-year views, which is consistent with the public sector channel reopening that began with the Army contract vehicle in Q3. Saudi Arabia's appearance in both lists is worth noting alongside the sovereign cloud commentary: the regions where sovereignty is a purchasing criterion are the regions posting the strongest growth and the longest deal cycles simultaneously.

Key Topics & Management Commentary

Overall Management Tone: Direct and pre-emptive on the one number that mattered, and noticeably more quantified than the October call. The CEO addressed the backlog shortfall in prepared remarks before any analyst could raise it, named both causes, and sized their combined effect. Where the posture was weakest was on the forward definition: having told the market in October that 25% would disappoint and then delivered 25%, management asked for trust in a second qualitative guide, "slightly decelerate," for 2026. The most convincing passages were financial rather than commercial, and the capital return decision did more to signal confidence than any language on the call.

1. Anatomy of the Backlog Shortfall

The CEO opened the substantive portion of the call by going straight at the number, which is the right instinct when a metric has missed a bar the company itself set.

"Let me directly turn to Q4 number, where I expect some questions. Our current cloud backlog. In Q4, it grew 25%. Back in Q3, we expected to reach 26%." — Christian Klein, CEO

Two causes were named. The first is deal size and its effect on ramp timing.

"First, we closed a higher share of significantly large deals in Q4 compared to our forecast in October, which is great for the total cloud backlog. But large customers often don't move their mission-critical ERP in the first year. These deals always have more back-end loaded ramps, and as a consequence, a limited impact on current cloud backlog in the first 12 months." — Christian Klein, CEO

The second is a definitional exclusion rather than a commercial event.

"Second, we closed a higher share of government deals in Q4 that included a termination for convenience by law. Such deals are not reflected in the CCB." — Christian Klein, CEO

Management sized the two effects together at approximately one percentage point of constant-currency current cloud backlog growth, and was explicit that the underlying commercial performance ran ahead of plan on bookings, with lower churn than expected and stable discount rates.

Assessment: Both explanations are structurally credible and, unusually, both are externally corroborated by a different disclosed metric. If large back-end-ramping deals displaced smaller front-loaded ones, total cloud backlog should outgrow current cloud backlog, and it did by five points. If government contracts with statutory termination clauses are excluded by definition, that is a measurement rule rather than a demand signal. The honest caveat is that a company gets limited credit for explaining a miss with the same metric-composition argument two quarters running, and this is now the second consecutive quarter in which the twelve-month cloud measure came in below where management pointed it.

2. Total Cloud Backlog at €77B, Up 30%

Total cloud backlog reached €77.29B, up 22% reported and 30% at constant currency, on what management described as the best bookings quarter of the year. Deal concentration reached a record: transactions above €5M contributed 71% of total cloud order entry in the quarter.

"This combined momentum continues to materialize in large cloud transactions, with deal volumes greater than EUR 5 million, contributing a record 71% to our total cloud order entry in the fourth quarter." — Dominik Asam, CFO

Management also made the point that contract duration did not lengthen, which matters because a rising total backlog achieved by extending contract terms would be a lower-quality outcome than one achieved by adding business.

Assessment: This is the strongest disclosure in the release and it is being discounted almost entirely. A 30% increase in contracted value on a €77B base, with stable duration and lower churn, is a better forward indicator than a twelve-month recognition window that is sensitive to deal mix. The record 71% large-deal concentration cuts both ways: it validates SAP's position with the largest enterprises and it makes the near-term metric structurally more volatile, which is precisely the volatility that produced this quarter's miss.

3. The 2025 Scorecard: Both Profitability Guides Cleared

Full-year non-IFRS operating profit reached €10.66B at constant currency against a €10.3–10.6B range, and free cash flow reached €8.24B against a revised €8.0–8.2B range. Both finished above the top of their guided bands. Free cash flow nearly doubled, up 95% to €8,239M, and operating cash flow rose 76% to €9,156M.

The cash improvement is partly a base effect that will not repeat. Restructuring payouts fell from €2.5B in 2024 to €0.8B in 2025 as the 2024 transformation program concluded, and share-based compensation payouts also declined. Q4 alone swung from negative €908M of free cash flow a year ago to positive €1,034M.

Assessment: We modeled FY25 free cash flow at €8.1B and non-IFRS operating profit at €10.5–10.6B at initiation; both came in at or above the top of those estimates. The composition matters for 2026: with the restructuring drag now fully lapped, next year's guided step to approximately €10B has to come from operations and from the widening gap between share-based compensation expense and its cash cost, not from a disappearing headwind. Management addressed exactly that in Q&A and gave a usable formula.

4. The 2026 Guide and What It Does Not Contain

The 2026 outlook guides cloud revenue to €25.8–26.2B at constant currency, up 23 to 25%; cloud and software revenue to €36.3–36.8B, up 12 to 13%; non-IFRS operating profit to €11.9–12.3B, up 14 to 18%; free cash flow to approximately €10B; and the non-IFRS tax rate down to approximately 29%.

What the outlook does not contain is a total revenue figure for 2026. SAP guides cloud and cloud-plus-software but stops short of the top line, while separately committing that constant-currency total revenue growth will accelerate through 2027. Working the arithmetic explains why those two facts sit together.

Assessment: This is the most important thing in the release that management did not say out loud, and it explains a meaningful share of the sell-off beyond the backlog headline. Investors who read "accelerate total revenue growth through 2027" as a 2026 commitment were mis-calibrated, and the guide corrected them. The acceleration case now rests on the outer-year ramps inside that €77B total backlog arriving in 2027, which is the same argument management used to explain the current backlog miss. The argument is coherent. It also asks the market to underwrite two more years before the thesis is testable.

5. A €10B Buyback, and What It Signals

SAP authorized a new share repurchase program of up to €10B, scheduled to begin in February 2026 and to complete by the end of 2027. For context, SAP's 2020, 2022 and 2023–2025 programs together repurchased approximately 56 million shares for about €8.0B. The new program is larger than all three predecessors combined and runs over two years rather than three.

"As a result, SAP has decided to further step up its capital returns with a new two-year share repurchase program of up to EUR 10 billion, scheduled to start in February." — Dominik Asam, CFO

Year-end net liquidity was €3.38B, and 2026 free cash flow is guided to approximately €10B, so the program is fundable from two years of cash generation without leverage.

Assessment: This resolves the open capital-allocation question we flagged at initiation, where the prior €5B program had completed in August with no successor announced. At the post-reaction market capitalization the authorization is worth roughly 5% of shares outstanding over two years, which is a material per-share tailwind at a mid-20s multiple. Buybacks announced into a de-rating are also the cheapest form of management conviction to verify: execution is disclosed, and we will grade it quarterly.

6. Business AI Reaches Commercial Scale

The AI narrative moved from qualitative to quantified this quarter, and the attach-rate disclosure we flagged as withheld in October returned in a stronger form.

"More than two-thirds of our Q4 cloud order entry includes business AI, increasing by more than 20 percentage points compared with Q3. Looking at the 50 largest deals in Q4, 90% of them included AI or SAP Business Data Cloud. We also saw the number of customers using our AI copilot tool growing ninefold over the course of the year." — Christian Klein, CEO

Business Data Cloud secured approximately €2B of total contract value in under a year since launch. In Q4 alone, almost two-thirds of deals above €1M involved four or more lines of business, an increase of 25 percentage points, which is the cross-sell metric that supports the conversion-multiplier claim.

Assessment: This is the disclosure we said was missing at initiation, and it arrived with better numbers than expected. A move from roughly 45% to more than two-thirds of cloud order entry including AI in a single quarter is a genuine step change in attach. The remaining measurement gap is monetization: attach tells us AI is present in the deal, not what it added to the price. Until SAP discloses AI-attributable revenue or an uplift figure, the metric proves AI is a deal-shaping lever rather than a revenue line, which is what we assumed at initiation and what the pillar was scored on.

7. The €2B Internal Efficiency Program

Management put a number on the internal AI transformation for the first time, targeting a run rate of approximately €2B in cost efficiencies by the end of 2028, equal to 15 to 20% of addressable costs. In R&D, 35% of code is already automated.

"Altogether, our goal is to achieve a run rate of around EUR 2 billion in real cost efficiencies by the end of 2028, thanks to the internal usage of AI. This equates to efficiency gains of 15%-20% of addressable costs, which also helps us to reinvest into our AI roadmap." — Christian Klein, CEO

Assessment: A €2B run-rate target against FY25 non-IFRS operating profit of €10.4B is a large number, and the framing matters as much as the size: management explicitly said part of it funds reinvestment rather than dropping to the margin line. The credibility question is whether €2B of efficiency can be extracted without a restructuring charge, which management addressed directly in Q&A and answered with a qualified no-for-now. Watch the operating expense to revenue growth ratio, which management guided toward the lower end of its 80 to 90% band, as the observable proof.

8. Sovereignty: A Moat That Slows the Sales Cycle

Sovereign demand emerged this quarter as a genuine two-sided factor, and management was unusually candid that the same trend helping SAP win is hurting the metric the market watches.

"While SAP is extremely well positioned in this segment, and we have a significant pipeline of opportunities due to the trust Germany and SAP continue to enjoy on a global scale, it takes longer to negotiate these more complex transactions and also longer to deploy and ramp as compared with plain vanilla offerings of U.S. infrastructure as a service vendors." — Dominik Asam, CFO

The strategic case was stated more assertively later in the prepared remarks, alongside a defense of SAP's decision to retain its own infrastructure capability.

"As the largest non-US software, SaaS, and PaaS vendor, there's no company better positioned than SAP to satisfy this rapidly growing demand. Our strategy to design a stack which is not locked into any particular IaaS vendor is particularly valuable in that respect." — Dominik Asam, CFO

Concrete Q4 evidence included a OneGov agreement with the US General Services Administration and one with HMRC in the United Kingdom, following the November launch of an EU AI Cloud offering and partnerships with Bleu, Capgemini and Mistral AI in France.

Assessment: Sovereignty is the most underrated element of the SAP case and the most awkward for the near-term metric. Being European is a durable competitive asset in a segment that is growing fast and where US hyperscalers are structurally disadvantaged, and the retained SAP Converged Cloud infrastructure that looked like a legacy drag now reads as optionality. The cost is longer negotiation and longer ramp, which suppresses current cloud backlog while raising total cloud backlog. That is the same mechanical trade as the large-deal effect, arriving through a second channel.

9. The Tax Rate Steps Down Structurally

The full-year non-IFRS effective tax rate came in at 30.4% against a guided 32%, and management guided the midterm rate to 28 to 30%, the lower half of the previously communicated 28 to 32% band. The 2026 outlook assumes approximately 29%. The driver named was an increasing ability to offset foreign withholding taxes in Germany.

Assessment: Quietly one of the more valuable disclosures in the release, and it received no analyst questions. Moving the sustainable tax rate from 32% to 29% adds roughly four to five percent to net income at any given level of operating profit, permanently. Applied to our 2026 operating profit estimate, the rate change alone is worth roughly €0.30 of EPS. In a quarter where the market removed roughly a sixth of the market capitalization over a 70bp backlog measure, a permanent 300bp reduction in the tax rate went unremarked.

10. Customer NPS Misses Its Target Range

The one non-financial target SAP missed deserves more attention than it received. Customer Net Promoter Score fell three points to 9, below the 12 to 16 target range. Management attributed the decline primarily to lower scores from on-premise customers who have not yet transitioned to the cloud, noting that cloud-oriented customer scores held steady and rose in the enterprise segment.

Assessment: The attribution is plausible and it is also uncomfortable, because the dissatisfied cohort is the same population SAP is trying to convert at a two-to-three-times revenue multiplier. Customers whose satisfaction is deteriorating while they sit on a support contract SAP is repricing and de-prioritizing are, at minimum, a harder sale, and at worst a competitive opening for a third-party support provider or a rival ERP. This also sits directly adjacent to the European Commission's open proceedings into SAP's on-premise maintenance and support policies, and the two facts describe the same pressure from different angles.

11. Teradata Litigation and the EU Review

IFRS operating profit absorbed approximately €0.2B of Teradata litigation expense in the quarter, which the release cross-references to a dedicated disclosure section. The expense is excluded from the non-IFRS figures. Separately, on November 14 SAP reaffirmed its commitment to fair competition amid the EU review of its maintenance and support practices, the proceedings opened on September 25.

Assessment: The Teradata charge is a real cash-relevant cost being adjusted out, and it deserves to be tracked rather than ignored, but at €0.2B against €9.8B of IFRS operating profit it is not thesis-relevant on its own. The EU review remains the more consequential item and remains almost entirely undiscussed. For a second consecutive quarter it drew no analyst questions on the call. The exposure is not the potential fine; it is that a remedy constraining maintenance and support policies would blunt the economic pressure that drives the cloud conversion the entire bull case depends on.

Guidance & Outlook

Disclosed Currency Translation Impact

The 2026 outlook is set at an assumed 1.13 USD/EUR average, while the sensitivity table above assumes 1.18. The Q1 drag of 8.0 percentage points on cloud revenue growth is the steepest quarterly translation headwind SAP has disclosed in this cycle, and it means reported Q1 cloud growth will land in the mid-teens even if the constant-currency rate holds near the guided range. Investors reading a reported number in April without adjusting will see an apparent collapse that is entirely translation.

Implied Q1 setup: With cloud revenue guided to 23 to 25% at constant currency for the year and an 8.0 point translation drag in the first quarter specifically, reported Q1 cloud revenue growth should print around 15 to 17%. Against Q1 2025 cloud revenue, that is the single most likely source of a misleading headline in the next three months.

Street at: Consensus entering the print had current cloud backlog growth at approximately 25.7% and had priced 2026 cloud revenue above the guided €25.8–26.2B range. Both were missed, by small margins. The free cash flow guide of approximately €10B was described in the Q&A as clearly ahead of expectations, and the €10B buyback was not anticipated.

Guidance style: SAP's pattern of guiding profit conservatively and revenue optimistically held for a second consecutive year, with both profitability lines finishing above their ranges and the cloud line at the low end. The change worth noting is a credibility cost on qualitative guidance specifically. "Slightly decelerate" meant four points in 2025, and the same phrase is now doing the same job for 2026. Management was pressed on this directly and gave a bound, which is progress, but the bound is a negative definition rather than a number.

Analyst Q&A Highlights

What "Slightly Decelerate" Means After It Meant Four Points

The opening question went directly at the credibility problem, pairing the specific October framing with a request to bound the 2026 language. The questioner also asked whether the reported 25% was rounded up or down, which management did not address. The answer contained the most useful forward constraint on the call.

Q: "You know, you spoke obviously at the end of the year of 25% being disappointing, which is unfortunately where we ended up… And then maybe just finally, you talked about slight deceleration. I guess everybody's going to say, well, slight deceleration last year was 4 points, I guess we could say 3 points organic. Is that the same type of range that people should be thinking about? Because I think that's, you know, initially where people would go to."
— Adam Wood, Morgan Stanley

A: "Now, for 2026, what means marginal decline or slight decline over the year? Definitely not a 4% decline. That is not what we are seeking for. It will be not such a decline like what you have seen in 2025. Also, again, given the strong pipeline, what we have now at the beginning of the year."
— Christian Klein, CEO

Assessment: The most tradeable sentence on the call, and the direct successor to the equivalent moment last quarter. "Definitely not a 4% decline" bounds 2026 current cloud backlog growth above 21%, and the framing implies management is targeting something closer to 23 to 24%. That is a falsifiable claim and it is the number to grade in twelve months. The unanswered rounding question is a small but real omission: the difference between 25.4% and 24.6% is material to a model, and management chose not to close it.

Whether the Large-Deal Mix Permanently Resets the Recognition Profile

A question probed whether the shift toward large transformations with long ramps is a one-quarter mix event or a structural change in how contracted business converts to revenue, and what it implies for the medium-term cloud growth trajectory. Management's answer conceded a forecasting failure and then declined to change the incentive that produced it.

Q: "How should we think about that maybe midterm trajectory of, of total cloud revenue growth, you know, into 2027 and perhaps beyond? Just help us with, you know, how you think about that pace of cloud revenue growth, either deceleration or scope for stabilization."
— Ben Castillo-Bernaus, BNP Paribas

A: "I mean, good question, and for sure, when I'm looking back to October, there is definitely, I mean, honestly, also a lessons learned at Q4, and we didn't see this in the forecast… And the phasing, actually, I mean, should we now incentivize our people to keep the first 12 months up? That would be the wrong thing to do, because, again, it's against the nature of how these transformations work."
— Christian Klein, CEO

Assessment: The right answer, and an expensive one to give on a day the stock is down 15%. Refusing to compensate the sales force for optimizing a twelve-month reporting metric is correct capital stewardship and directly costly to the number the market prices. It also implicitly concedes that current cloud backlog will stay noisy for as long as the large-deal mix persists, which management said it expects to continue. Investors should treat the metric as lower-quality going forward and weight total cloud backlog more heavily.

How Accurate the Cloud Revenue Guide Has Proven

Within the same exchange, the finance seat made a precision argument to distinguish the reliability of the revenue guide from the reliability of the backlog guide, which is the distinction that matters for anyone modeling 2026.

Q: "There's lots of positives in here, lots of large deals in the mix, either sovereign cloud opportunity, you know, the high volume of AI and the backlog record TCB… I guess, could you help us just think about, you know, the changing landscape here, that growing mix of large deals in, in the pipeline that are converting the longer deal ramps?"
— Ben Castillo-Bernaus, BNP Paribas

A: "On the cloud revenue in terms of what we see going forward, first of all, we guided cloud revenues now for 26, and maybe just looking back at 25, I venture to say that the accuracy of forecasts on cloud revenues is actually by now extremely high. This is also due to the extremely high share of recurring, more predictable, revenues… we were a kind of only 75 million away from the midpoint of the guidance at the end of 2025."
— Dominik Asam, CFO

Assessment: A strong and verifiable claim. Landing within €75M of a €21.75B midpoint is a 0.3% variance across a year that included a tariff shock and a sovereignty-driven lengthening of deal cycles. It supports treating the 2026 cloud revenue range as high-confidence and the backlog language as low-confidence, which is the correct asymmetry for a model. It does not address the deeper issue, which is that the market prices SAP off the metric management forecasts worst.

Whether AI Is Winning or Threatening the Installed Base

A line of questioning raised the sector-wide concern that AI tooling could disintermediate enterprise application vendors, and asked for adoption penetration within the existing cloud base rather than the attach rate on new bookings. The answer supplied the penetration figure and then argued the disintermediation case backwards.

Q: "We've seen growing concerns about risks from AI impacting the enterprise SaaS… It would be good to share, in terms of AI traction, what percentage of existing cloud customers currently using your AI offering… do you see any of your customers starting to use, you know, different tools to respond to needs that they were previous-previously addressed by SAP, either in core ERP or across, HCM, et cetera?"
— Frederic Boulan, Bank of America

A: "Round about 60% of our customers are already using our AI actively. 20% are on the way to it… So to make this very clear, we are winning deals because of AI. We are not losing deals because of AI, and definitely, you know, these deals are actually now leveraging AI to increase the win rate in Q4."
— Christian Klein, CEO

Assessment: A 60% active-usage rate across the installed cloud base, with a further 20% in flight, is a materially better number than the market assumes for enterprise AI adoption, and it is the first time SAP has framed penetration this way. The supporting anecdotes about customers abandoning self-built large-language-model agents for lack of business context are the strongest argument in SAP's favor on the disintermediation debate. What remains absent is any monetization figure, and the question explicitly asked for a revenue uplift range that was not provided.

Funding the €2B Efficiency Target Without a Restructuring Charge

A question tested whether the internal AI savings target implies another restructuring program, which matters because SAP has taken workforce charges in each of the past two years. The answer was a qualified denial with an unusually frank caveat attached.

Q: "I think you referred to a EUR 2 billion saving at, at SAP over the next couple of years. Can you talk to how that's going to flow through the business? Are you gonna extract that through, natural churn in the organization, or, or do we have to think more about, say, another restructuring program?"
— Charles Brennan, Jefferies

A: "So today, as of today, there is no restructuring plan, restructuring plan. Obviously, can you hold this out forever? No. But today I can tell you we're gonna achieve that by just scaling our business way more than in the past with AI."
— Christian Klein, CEO

Assessment: "Can you hold this out forever? No" is a more honest answer than the question required, and it should be read as guidance that a third workforce program is a live possibility rather than an excluded one. For modeling purposes, assume the €2B is delivered primarily through headcount growing well below revenue growth, with a non-trivial probability of a charge in 2027 or 2028. The finance seat reinforced the point elsewhere in the same exchange by committing to the lower end of the 80 to 90% operating leverage band while explicitly refusing to underinvest in the top line.

Reconciling a €10B Free Cash Flow Guide With Known Headwinds

The final question asked how the step to approximately €10B squares with the cash tax, currency and migration credit headwinds discussed through 2025. The answer produced the most directly usable modeling formula on the call.

Q: "Just on the free cash flow guidance, Dominik, of EUR 10 billion, clearly well ahead of expectations and looks to imply a pickup in cash conversion. I know we talked through the year about cash tax, FX, and the migration credit headwinds. Could you just help us reconcile these headwinds with the better free cash flow outlook… Improved cash conversion you're now expecting?"
— Toby Ogg, JPMorgan

A: "I would say, the upside is from two sources, partially from operational further improvements, which we have matured to a point that we feel comfortable guiding it now, but also from the fact that the delta on stock-based compensation between the P&L and cash is increasing… We've always said, take the effective tax rate of the non-IFRS operating profit, add back round about EUR 1 billion. So now I say round about EUR 1 billion plus, so that gives us a part of the upside. And the good news is that's sustainable, so that's the new base, basically, to jump off."
— Dominik Asam, CFO

Assessment: The most valuable exchange of the call for anyone building a model. The formula given is explicit: apply the non-IFRS effective tax rate to non-IFRS operating profit, then add back somewhat more than €1B. On the 2026 guide midpoint that reconciles to approximately €10B directly, and management characterized the enlarged add-back as the sustainable new base rather than a one-year benefit. Free cash flow is now the cleanest line in the SAP model and, at the post-reaction price, the most important one.

What They're NOT Saying

1. A 2026 total revenue number: SAP guided cloud revenue, cloud and software revenue, operating profit, free cash flow and the tax rate, and did not guide total revenue. The arithmetic implied by the guided lines puts 2026 constant-currency total revenue growth at roughly 10 to 12% against 11% in 2025, meaning the acceleration promised "through 2027" does not appear in 2026. Not guiding the line is how that fact stayed out of the headline.
2. Whether 25% rounded up or down: Asked directly, twice in effect, and not answered. The gap between 25.4% and 24.6% changes the 2026 base and therefore what "definitely not a 4% decline" actually bounds. Management gave a qualitative floor while withholding the precise starting point it is measured from.
3. Any AI revenue or pricing uplift: Attach rates improved dramatically and were disclosed generously, from more than two-thirds of cloud order entry to 90% of the top 50 deals to 60% active usage in the installed base. A question explicitly asked for "a range of revenue uplift you're seeing" and no figure was given. SAP is telling the market AI is in the deals without telling it what AI is worth in the deals.
4. The European Commission proceedings: A second consecutive quarter with zero call airtime and zero analyst questions, four months after formal proceedings opened, and in a quarter where Customer NPS from on-premise customers deteriorated enough to break the target range. Those two facts belong in the same conversation and neither side put them there.
5. What the Teradata litigation could still cost: Approximately €0.2B ran through Q4 IFRS operating profit with a cross-reference to a disclosure section, no forward exposure range, and no questions.
6. Extension Suite, for a second straight quarter: Growing 10% at constant currency in the quarter and 8% for the year against Cloud ERP Suite's 30 to 32%, and again not mentioned in prepared remarks or Q&A. It remains the natural landing place for acquired assets, which makes its growth rate a standing comment on acquisition returns that nobody is asked to defend.
7. Sovereign cloud economics in aggregate: Deal margins were described as "almost the same" when asked, which is useful, but no pipeline value, expected revenue contribution or ramp profile was quantified for a category management credits with both material wins and materially longer cycles.

Market Reaction

SAP released Q4 and full-year results on the morning of January 29 in Europe, with the analyst call at 07:00 CET and the press conference at 10:00 CET, both concluding before the US open. The reaction session for the ADR was therefore the print day itself rather than the following day.

• Pre-print setup: The NYSE-listed shares closed at $236.11 on January 28, down 2.8% year to date against the S&P 500's +1.9%, and down 14.7% over the trailing twelve months. The 52-week closing range entering the print was $225.10 to $311.93, placing the stock roughly 24% below its high before the release.
• Reaction-day open: Shares gapped to $199.04, a 15.7% decline.
• Reaction-day range and close: The session traded between $195.12 and $201.44, a low 17.4% beneath the prior close, and finished at $200.21, down 15.2% or $35.90.
• Volume: 9.5M shares against a 1.7M thirty-day average, 5.6 times normal.
• Benchmark: The S&P 500 fell 0.1% on the session, so essentially the entire move was company-specific.
• Context: This was SAP's steepest one-day decline since October 2020, and erased more than €40B of market capitalization.

The decline was not about the reported quarter. Revenue met forecasts, non-IFRS EPS beat by roughly 9 to 12%, both full-year profitability guides finished above their ranges, and a €10B buyback was announced. The market sold two forward data points: a current cloud backlog exit rate of 25% against a 25.7% consensus, and a 2026 cloud revenue guide the market had positioned for above its top end.

The magnitude relative to the miss is the notable feature. A 70 basis point shortfall on one forward metric removed roughly a sixth of the equity value in a single session when measured from the January 28 close through the 17.4% intraday low. That ratio of price response to information content is characteristic of a crowded position unwinding rather than a re-underwriting, and the sector context supports that reading: capital has been rotating out of enterprise software broadly, which is why the stock entered the print already 24% below its 52-week high with the index up on the year.

The close at $200.21, near the top of the day's range and above both the open and the low, is the one constructive detail in the tape. A stock that gaps 15.7% lower, trades to 17.4% down, and then closes at 15.2% down on 5.6 times volume found buyers into the decline rather than sellers into the close.

Street Perspective

Debate: Is the Backlog Miss Mechanical or Is Demand Slowing?

Bull view: Total cloud backlog grew 30% at constant currency to €77.29B, five points faster than the twelve-month measure, on the best bookings quarter of the year with lower churn and stable discount rates. If demand were weakening, contracted value would not be outgrowing recognized backlog. Government contracts excluded by statutory clause and large deals with outer-year ramps are measurement effects, and management sized them at roughly one point.

Bear view: This is the second consecutive quarter in which the cloud metric landed below where management pointed it, and the explanation has been metric composition both times. A company that keeps missing the number it forecasts, using a different composition argument each quarter, is describing a forecasting problem at best. "Slightly decelerate" meaning four points, then being reissued for 2026, compounds the credibility cost.

Our take: The bull case is better evidenced this quarter than last. The divergence between the two backlogs is externally verifiable, arithmetically consistent with the stated cause, and was accompanied by disclosure of record large-deal concentration at 71% of order entry. That is a coherent, cross-checked story rather than an assertion. The bear case wins on credibility rather than on substance, and the credibility cost is real and appropriately priced into a lower multiple. It is not worth 15% in a day on top of a 24% de-rating that preceded it.

Debate: Does the 2026 Guide Actually Deliver Acceleration?

Bull view: Cloud revenue guided to 23 to 25% at constant currency, cloud and software to 12 to 13% against 12% delivered in 2025, free cash flow to approximately €10B against €8.24B, and the tax rate down 140 basis points. Operating profit compounds at 14 to 18% on a base that just grew 31%. The €77B total backlog with its outer-year ramps is the mechanism that delivers the 2027 acceleration, and it is already contracted.

Bear view: Total revenue is not guided for 2026, and the implied figure is roughly 10 to 12% constant-currency growth against 11% delivered in 2025. That is flat. The acceleration has been promised since 2024, has now been deferred to 2027, and depends on ramps that will not be observable for two years. Meanwhile support decay is guided to accelerate, which works directly against the top line.

Our take: The bear case is factually correct about 2026 and this is the most legitimate criticism in the release. The acceleration is deferred and the market was right to reprice for that. Where we part company is on what the deferral is worth. The guided lines that do compound, operating profit at 14 to 18% and free cash flow at approximately 21%, are the lines that determine value at a mid-20s multiple, and both are guided ahead of the top line. An investor buying SAP for 2026 total revenue acceleration should not own it. An investor buying €10B of free cash flow, a 5% buyback and a structurally lower tax rate should.

Debate: Is European Sovereignty a Durable Moat or a Sales-Cycle Tax?

Bull view: Sovereignty is a structural advantage no US vendor can replicate, arriving exactly as European and other governments make it a purchasing requirement. SAP retained its own infrastructure capability when the market called that a legacy drag, and can now deploy across its own data centers, trusted European infrastructure, or on-site. Q4 delivered OneGov agreements with the US General Services Administration and HMRC, and an EU AI Cloud offering launched in November.

Bear view: Sovereign deals negotiate slower, deploy slower and ramp slower, which is a direct tax on the metric the market prices. Management named sovereignty as a contributor to the backlog shortfall. A moat that suppresses your headline growth metric for several years is worth less than one that does not, and running proprietary infrastructure at sub-scale against hyperscalers carries its own margin risk.

Our take: Both sides describe the same fact and disagree on the discount rate. Sovereignty is real, durable, and currently being paid for in slower recognition. Management stated the deal margins are "almost the same," which if it holds removes the profitability objection and leaves only timing. We weight this positively, with the caveat that it is the third distinct mechanism this quarter pushing revenue into outer years, alongside large-deal ramps and government termination clauses. Three separate reasons for the same delay is a pattern, and the pattern is that SAP's business is lengthening. That is acceptable at 23x forward earnings and was not at 39x.

Model Update Needed

Valuation: At the reaction-day close of $200.21 and the FMP-verified 1.1971 USD/EUR spot on January 29, the implied price is approximately €167 per ordinary share, the ADR being a one-for-one claim. Share count of approximately 1,167M derives from full-year non-IFRS profit after tax of €7,176M against basic non-IFRS EPS of €6.15, putting market capitalization near €195B and, net of €3.38B of year-end net liquidity, enterprise value near €192B.

On those figures SAP trades at approximately 27x trailing FY25 non-IFRS EPS of €6.15 and approximately 23x our FY26 estimate of €7.05 to €7.15. Free cash flow yield on the approximately €10B 2026 guide is roughly 5.1%, and enterprise value to 2026 free cash flow is roughly 19x. The €10B repurchase authorization equals roughly 5% of market capitalization over two years.

The comparison that drives the upgrade: at our initiation three months ago the stock was roughly 39x forward non-IFRS EPS on a 2.9% free cash flow yield. It is now roughly 23x on a 5.1% yield. Over the same period FY25 non-IFRS EPS came in at €6.15 against our €6.05 to €6.10 estimate, both profitability guides were exceeded, the sustainable tax rate fell 300 basis points, and a buyback larger than the prior three programs combined was authorized. Earnings moved up and the multiple fell 40%. That is a valuation event, not a business event.

Thesis Scorecard Post-Earnings

Graded against the six pillars established at initiation, unchanged in definition.

Grading Last Quarter's Commitments

Overall: Thesis strengthened on the operating and capital-allocation pillars, weakened on the growth-metric pillar, and resolved on the valuation pillar. Five of eight commitments were delivered, two partially, and the single most-watched one was missed.

Action: Upgrading to Outperform from Hold. We stated at initiation that the upgrade trigger was a backlog exit rate at or above 26% together with a formalized 2026 acceleration guide, particularly on a de-rated price. The backlog leg was missed and the acceleration guide was only partly delivered. We are upgrading anyway, on the third condition, and the reasoning should be explicit: the de-rating over-delivered so far beyond what the shortfall justifies that the risk/reward has inverted. The stock is 28% below where we rated it Hold, on earnings that came in above the estimates that Hold was based on. The business is compounding operating profit at a guided 14 to 18% and free cash flow at 21%, and is returning 5% of its market capitalization over two years. What was wrong at initiation was the price, and the price has been more than corrected.

What would change our mind: a downgrade back to Hold if 2026 current cloud backlog growth tracks below 22%, which would breach management's "definitely not a 4% decline" bound and turn three quarters of composition explanations into a demand trend. A downgrade to Underperform if that occurs alongside cloud revenue falling below the guided 23% floor, or if the European Commission proceedings produce a remedy touching maintenance and support pricing. The near-term risk to monitor is presentational rather than fundamental: the disclosed 8.0 percentage point translation drag means reported Q1 cloud growth should print in the mid-teens, and that headline will look far worse than the underlying rate.