The Cliff Arrives: Novartis Decelerates From 21% to 7%, Then Tells the Street Its 2026 Numbers Are Too High

Results vs. Consensus

Novartis is a Swiss foreign private issuer reporting under IFRS. It furnishes a 6-K rather than an 8-K, its primary operating measure is "core" (non-IFRS) results, and it guides in constant currencies. Every actual below traces to the October 28 media release and the Condensed Interim Financial Report attached to that 6-K. Growth rates are as reported by the company, with the currency basis labelled wherever it changes the reading.

Q3 2025 Scorecard

Core EPS consensus spans $2.26 to $2.31 across the sources we track; we use the more widely cited $2.26, on which the shortfall is 0.4%, and note that on the higher mark it is 2.6%. Revenue comparisons require care with this issuer: several data feeds store Novartis "revenue" as $14,358M, which is net sales of $13,909M plus other revenues of $449M, and then compare it against a net-sales-basis estimate. That mismatch manufactures a $0.66B beat that does not exist. Where no reliable consensus mark exists for a line, the column reads "n/a" rather than a number we invented.

Q3 2025 vs. Q3 2024

The IFRS lines and the core lines tell opposite stories and both are true. Operating income grew 27% cc because the prior-year quarter carried $1,140M of other expense against $335M this year, largely impairment that did not repeat: core adjustments fell to roughly $1.0B from $1.5B. Strip that out and the underlying operating performance is the core line, which grew 7% cc on 7% cc sales with the margin flat. The 31% cc growth in IFRS EPS is not a business event.

Q3 2025 vs. Q2 2025 (sequential)

The sequential table is where the quarter is legible. Between June and September the core margin gave back 290 basis points and Entresto shed $480M of quarterly revenue, while the growth cohort added $506M. That is the patent cliff and its offset arriving in the same three months, and for one quarter they very nearly cancelled. Net sales fell 1.0% sequentially. The question the fourth quarter answers is whether they keep cancelling once Entresto has a full quarter of generic exposure behind it rather than one.

Nine months 2025 vs. nine months 2024

Nine months into 2025 this is an exceptional year: a 41.2% core margin, 250 basis points of constant-currency expansion, core EPS up 21% and free cash flow up 26% to $15,941M, already within $312M of 2024's full-year $16,253M. None of that is in dispute. All of it was produced by the three quarters before the largest patent expiry in the company's history took full effect.

Quality of Beat/Miss

Brand and Franchise Performance

The growth cohort

This is the cohort that has to absorb the cliff, and its trajectory as a share of net sales is the single cleanest measure of whether the transition is working: 19.0% in Q3 2024, 24.4% in Q2 2025, 28.3% in Q3 2025. It added $1,495M of year-on-year revenue in the quarter while generic erosion removed roughly 7 points of growth, and it did so without a single brand carrying the load alone. Kisqali contributed $542M of the increase, Kesimpta $384M, and the other four $569M between them.

The company's own slide frames a seven-brand cohort including Cosentyx at +35% cc. That is a fair presentation, but it dilutes the signal: Cosentyx contributed $5M of year-on-year growth in the quarter. We track the six because the six are the ones actually replacing lost revenue.

"Our U.S. growth was up 91% in quarter 3. We are the metastatic breast cancer leader in NBRx and TRx. And in early breast cancer, our share is 63%, and we're leading both in the overlapping populations with our competitor and the exclusive population. In particular, I see -- I'd say we see significant growth potential in that exclusive population where we estimate more than 60% of patients are currently not on a CDK4/6 inhibitor." — Vas Narasimhan, CEO, on Kisqali

Assessment: The Kisqali claim is the most consequential disclosure in the brand section. If 60% of the exclusive early-breast-cancer population is untreated with any CDK4/6 inhibitor, the growth here is market creation rather than share capture, and market creation decays more slowly. The NATALEE five-year data presented at ESMO, a 28.4% reduction in recurrence risk with a 0.8 overall-survival hazard ratio still maturing, is what underwrites it.

The eroding cohort

Entresto is the headline and it behaved better than the headline implies. US generic entry landed inside the quarter and the brand still printed $1,877M, down 1% cc, with nine-month sales up 15%. Sequentially it fell $480M from Q2's $2,357M, which is the real measure of the entry and the run-rate the fourth quarter starts from. Tasigna and Promacta, which entered generic competition earlier, show what the mature state looks like: down 48% and 38%.

Assessment: The erosion cohort is behaving predictably and the market has had years to model it. The item that deserves attention is not on this list: Xolair at $440M grew only 3% cc and Tafinlar+Mekinist at $550M only 1% cc, both materially below their nine-month rates of 8% and 9%. Two more brands rolling over inside the same quarter the cliff arrives is the kind of drift that becomes a narrative two quarters late.

Regional mix

Assessment: The US grew 12% in the quarter that absorbed a 7-point generic drag and a negative revenue-deduction adjustment, which is a striking result and says the launch portfolio is US-weighted and working. Europe at +2% cc is the soft spot, and China at $1.0B fell 3%. The US now sits at 43% of net sales against 42%, and rising US concentration is precisely what makes the most-favoured-nation pricing question below more than theoretical.

Key Topics & Management Commentary

Overall Management Tone: Management was confident on the pipeline and conspicuously careful on the numbers, and the gap between the two halves was wider than at any point this year. Prepared remarks spent twelve slides on brands and readouts and arrived at the financial section with a quarter that needed explaining, which management explained fully and without spin. The revealing posture was forward-looking: the company volunteered quarterly guidance it says it normally avoids, pre-announced a 2026 margin headwind before guiding 2026, and declined every invitation to characterise next year's earnings. That is a management team managing expectations down in an orderly way rather than one defending a quarter.

1. The gross-to-net true-up and what Q3 actually was

The single largest wedge between the reported quarter and the underlying quarter was a set of negative US gross-to-net true-ups relating to prior periods, sized by the CFO at roughly $180M and attributed mainly to Cosentyx and Entresto under the Medicare Part D redesign. Most of it relates to the second quarter of this year.

"In the U.S., we had some negative gross to net true-ups first time since the year. Prior, we had mostly positive. But they were mainly related to Medicare Part D redesign, which was new for the industry this year based on invoices for prior periods, mainly quarter 2. And excluding these true-ups, the underlying growth would have been 9% on the top line and 11% on the bottom line as the priority brands and launches continue to offset the increasing generic erosions, mainly for Entresto, Tasigna, and Promacta in the U.S." — Harry Kirsch, CFO

The company's own arithmetic on the effect is explicit: Q1 carried two points of positive true-ups, Q3 carried almost two points of negative ones, and Q4 2024 carried roughly three points of positive ones that now form an inflated base. The CEO closed the loop himself when pressed on whether the year nets out.

"I think, Harry, if I'm correct, if you net out the prior period upside versus this that really the year-to-date is relatively clean." — Vas Narasimhan, CEO

Assessment: The 9% and 11% underlying figures are the right way to read the quarter in isolation and the wrong way to read the year. If Q1 was flattered by the same mechanism that penalised Q3, then the nine-month reported numbers already are the underlying numbers, and an investor who adjusts Q3 upward without adjusting Q1 downward is double-counting. Credit management for saying so unprompted; the CEO's intervention was the most useful thirty seconds on the call.

2. The deceleration from 21% to 7%

The quarterly cadence slide is the most important page in the deck. Constant-currency core operating income growth ran 27% in Q1, 21% in Q2 and 7% in Q3, and constant-currency net sales growth ran 15%, 11% and 7%. Even on the underlying basis management prefers, Q3 core operating income growth was 11%, roughly half the second quarter's rate.

The mechanics are not mysterious. Generic competition moved from a 3-point drag across the nine months to a 7-point drag in the quarter as Entresto joined Promacta and Tasigna. Core R&D rose to 19.2% of net sales from 18.1%. The core margin stopped expanding for the first time this year: 42.2% in Q2, 39.3% in Q3.

Assessment: The deceleration is the quarter. Nothing in it is a surprise in kind, because the LOE calendar has been public for years, but the magnitude of the step-down arrived in one quarter rather than three, and it lands immediately before a guided-flat fourth quarter and an unguided 2026. A company growing core operating income at 7% with the margin flat is a different asset from one growing it at 21% with the margin expanding 340 basis points, and the market repriced the difference.

3. Q4 guidance: the first flat-to-down quarter

Management broke its own convention and guided the fourth quarter specifically, on the grounds that mid-year US generic entries make the quarterly shape unusually distorted.

"Adjusting for these one-timers, we expect quarter 4 underlying growth to be low single digit on the top line and mid-single digit on the bottom line, reflecting the increasing generic erosion from a full year impact of Entresto U.S. generics but better, obviously, than what we expect to report, including the prior year gross to net adjustments." — Harry Kirsch, CFO

On a reported basis, the simulation slide shows net sales flat in constant currencies and core operating income flat to down low-single-digit. Full-year guidance of high-single-digit sales growth and low-teens core operating income growth was reaffirmed, which is arithmetically consistent only because the first three quarters did the work.

Assessment: Reaffirming the full year while guiding the fourth quarter to flat is honest and it is also the least flattering possible framing of the exit rate. The company will finish 2025 with a strong headline and a fourth quarter that looks nothing like it. Investors price the exit rate.

4. Avidity: a $12B deal and an instruction to cut 2026

Two days before the print, Novartis agreed to acquire Avidity Biosciences for roughly $12B in cash at $72 per share, conditional on the spin-off or sale of Avidity's early-stage precision cardiology programs. The assets are del-desiran in myotonic dystrophy, del-zota in Duchenne exon 44, del-brax in FSHD, a preclinical neuromuscular pipeline, and an extra-hepatic xRNA delivery platform. Closing is expected in the first half of 2026. The company says the deal lifts the 2024-2029 sales CAGR from 5% to 6%.

What was new on this call was the cost, quantified and directed at the models.

"Now we also mentioned yesterday that we do expect some short-term core margin dilution given Phase III trials are basically now starting to run or up and running shortly in the range of 1% to 2% points for the next 3 years. But we are confident that we return to the 40% margin, which we already achieved this year also will return them back to that in 2029. And please make sure that you also model this 1 to 2 points core margin dilution as you finalize your 2026 models for us." — Harry Kirsch, CFO

One to two points of core margin on a roughly $56B revenue base is $560M to $1.1B of core operating income per year, for three years, before any offsetting contribution. The sales contribution is guided to arrive "by 2029" and "post 2030."

Assessment: The strategic logic is defensible and the disclosure is admirably direct. The investment consequence is separable from both. A company that tells analysts, in October, to cut their unpublished 2026 models, and then does not formally guide 2026 until February 4, has created a fourteen-week window in which the consensus number is known to be wrong and the replacement number is unknown. There is no informational advantage in owning that window.

5. Entresto's first generic quarter

Entresto printed $1,877M, up 1% in dollars and down 1% cc, against $2,357M in the second quarter. US generics entered during the quarter. Nine-month sales are still up 15%.

Assessment: Better than a bear would have modelled for the entry quarter, and irrelevant to the run-rate question. The sequential decline of $480M is the number to carry forward, and the fourth quarter is the first with a full quarter of exposure. The CFO's own framing of Q4 as "a full year impact of Entresto U.S. generics" is the honest read: the quarter just reported is the transition, not the destination.

6. Cosentyx, the $74M adjustment, and an HS market that is not growing

Cosentyx was flat in dollars and down 1% cc at $1,698M. Excluding a one-time US revenue-deduction adjustment of $74M, global growth was roughly 4% cc and US growth moved from +1% to +9%. Management reaffirmed the $8B+ peak-sales ambition and mid-single-digit full-year growth. The more interesting disclosure came under questioning on hidradenitis suppurativa, where a subtle change in the slide language was noticed.

"So what I can say is that we feel confident that our share has stabilized after the competitor entry. I think we have not seen the market growth that we had originally hoped for that we -- there's clearly a lot of patients who can benefit from biologic therapy with HS. We continue to see this as a $3 billion to $5 billion-plus market, but it's clearly going to take longer for that market to develop." — Vas Narasimhan, CEO

Separately, roughly 25% of Cosentyx HS patients have moved to every-other-week step-up dosing, which management frames as a retention mechanism rather than a share gain.

Assessment: Share has stabilised, which removes the acute risk, and the market is not expanding at the pace underwriting the peak-sales number, which replaces it with a slower one. The $8B ambition was reaffirmed in the same breath as an admission that one of its growth legs will take longer than planned. That is not a contradiction management has to resolve this quarter, but it is one that accrues.

7. Rhapsido: approval in hand, access the gating item

Remibrutinib was approved by the FDA as Rhapsido, the first BTK inhibitor for chronic spontaneous urticaria, with no boxed warning, no contraindications and no routine laboratory or liver monitoring. Management sizes the uncontrolled population at roughly 400,000 of 1.5 million treated patients. EU and China submissions are complete; Japan is slated for later in the year.

"Right now, our focus is on sampling through patient start form, getting through patient start forms and negotiating with payers to ensure broad access in the early part of next year. I think once we get to the early part of next year, we get that base up through sampling in this initial phase, we would then start to expect a more rapid uptake through Q2 forward next year, where I think there will be the opportunity then to really drive uptake." — Vas Narasimhan, CEO

The indication stack behind CSU is the real asset: chronic inducible urticaria reading out next year, food allergy, hidradenitis suppurativa, relapsing MS and myasthenia gravis all in Phase III.

Assessment: A clean oral label in a symptomatic, high-burden condition is a genuinely good launch setup, and management is being realistic rather than promotional about the timeline: meaningful revenue is a second-quarter-2026 story, not a 2025 one. Nothing in the 2026 numbers should assume otherwise. The MS readouts are where this asset either becomes a franchise or stays a dermatology product.

8. Ianalumab: two positive Phase IIIs the market did not reward

Both NEPTUNUS trials in Sjögren's disease met their primary endpoint on ESSDAI reduction with a favourable safety profile, and the FDA granted Fast Track designation. The market's reaction to the ACR abstract had been poor, and management was pressed on it directly. Under questioning, it emerged that secondary endpoints did not reach statistical significance because of the testing hierarchy.

"I mean, I think for us, the most important thing is that we make a compelling proposition to patients and physicians. And then if we deliver a strong launch, then I think, obviously, the markets will do what the markets will do, but presumably will follow. I think -- we will present detailed data on Thursday, and I think that will help at least understand where our conviction comes from." — Vas Narasimhan, CEO

Management's argument is that ESSDAI is a physician-facing aggregate and that patients experience fatigue, salivary flow, sleep and steroid burden, which the full dataset presented later in the week would address.

Assessment: A hierarchically-failed secondary is a regulatory technicality and a commercial signal at the same time. In a disease with no approved therapy the primary endpoint is enough to file, and it is not obviously enough to launch into. Management's answer was an argument about what will matter at launch rather than evidence that it will, and it deferred the substance to an event two days later. This is a real asset with an unresolved commercial question, and the quarter did not resolve it.

9. Pluvicto and the radioligand infrastructure

Pluvicto grew 45% cc to $564M with US growth of 53%, driven by the pre-taxane mCRPC approval. Over 730 prescribing sites are live against a target of roughly 900 to support the hormone-sensitive indication; nine of ten patients are within 30 miles of a treating site; roughly 70% of sites use the pre-filled syringe. The PSMAddition trial in mHSPC reduced risk of progression or death by 28% with a 0.84 overall-survival hazard ratio, with global submissions planned for the fourth quarter.

"So for Pluvicto overall, I think we're on the steep part of the curve right now. We see -- as you saw, very strong growth in quarter 3. We would expect very solid growth in quarter 4. It's important to note in quarter 4, we always have a slowdown in the Thanksgiving and Christmas holidays." — Vas Narasimhan, CEO

Assessment: The distribution build-out is the durable part. Management's claim that it has "solved the challenge of rolling out radioligand therapy in the United States" is a platform claim, not a product claim, and if it holds it lowers the launch cost of roughly ten further radioligand assets in the clinic. The $5B peak-sales path now depends on the mHSPC label rather than on execution, which is a better problem to have.

10. US policy: weekly meetings, no agreement

By the date of this call three peers had reached agreements with the US administration on Medicaid pricing and tariffs. Novartis had not.

"I'd say we're meeting with the administration weekly to look at what are the best solutions we can come up with. It is important to note that the President was very clear on the four parameters, and I think those are the four parameters that are in discussion. And we'll have to see in the coming weeks and towards the end of the year if we can come to a proposed approach that makes sense for all involved." — Vas Narasimhan, CEO

Management's stated view is that the negotiations do not address the underlying issues, which it identifies as pharmacy benefit managers, 340B, and ex-US countries underpaying for innovation. Separately, the previously announced $23B five-year US commitment was characterised as mostly R&D operating expense rather than capital expenditure, with capex staying in a 2.5% to 3% of sales range.

Assessment: An unsigned agreement is an open-ended liability of unknown size, and the US is 43% of sales and rising. The capex clarification is genuinely reassuring and removes a modelling risk. The pricing question does not get smaller by waiting, and a company negotiating weekly without an outcome is carrying a drag that cannot yet be put in a spreadsheet.

11. Capital allocation and the buyback that carried per-share growth

The $15B buyback completed in July; a new up-to-$10B programme targeted for completion by end-2027 has commenced. Nine-month repurchases totalled 66.4 million shares for $7.5B, and shares outstanding fell 56.3 million from December 31. Net debt was $20,365M at September 30, down $3,419M from $23,784M at July 1 on strong cash generation, but up $4,224M from $16,141M at the end of 2024. Ratings are Aa3 and AA-.

Assessment: Core net income grew 6% cc and core EPS grew 10% cc: four points of the ten came from the share count. That is legitimate and it is also finite, and the $12B Avidity consideration lands on this balance sheet in the first half of next year on top of a new $10B repurchase authorisation and a dividend that consumed $7.8B. Something in that stack gives. The sequential net-debt reduction of $3.4B is the quarter's most under-discussed positive, and nobody asked about it.

Guidance & Outlook

Implied Q4 ramp. Take the quarterly simulation at face value rather than working back from the full-year guide, which is stated in constant currencies against a year in which currency costs roughly two points of core operating income. A flat fourth quarter on the prior year's $13,153M puts net sales near $13.2B. Flat to a low-single-digit decline on the prior year's $4,859M puts core operating income between roughly $4.7B and $4.9B. Against that, a share count down about 4% and a higher core financial expense leave core EPS near $1.95 to $2.05 for the quarter, and $8.90 to $9.00 for the year once added to the $6.94 already banked.

Street at. One to two margin points on a roughly $56B revenue base is $560M to $1.1B of core operating income, or about $0.25 to $0.50 per share after a 16.2% core tax rate on roughly 1,900M shares. That is the per-share size of the dilution the CFO asked the Street to put into its 2026 models. Our own estimate, built bottom-up from a 38% to 39% core margin on low-single-digit sales growth, lands at $8.80 to $9.10. That is flat to down on 2025, and no formal 2026 guidance exists until February 4.

Guidance style. Novartis raised its full-year core operating income guidance to low-teens from low-double-digit growth in July, and reaffirmed here rather than raising again, despite nine months tracking at 11% sales and 18% core operating income growth against a high-single-digit and low-teens guide. Reaffirming a guide the company is running ahead of, while simultaneously guiding the fourth quarter to flat, is a deliberate reset of the exit rate rather than conservatism.

Analyst Q&A Highlights

Whether a White House pricing agreement is close

The call opened on policy. With three peers having reached agreements with the administration, the first question asked whether the industry would converge on a template and when Novartis would follow. The answer separated the company's public position on the underlying issues from its private negotiating posture, and committed to no date beyond the end of the year.

Q: "It's a question about policy, Vas. And we've seen now two companies do deals with the White House around Medicaid and tariffs. And I wondered from your perspective, how much you felt we could see the industry do a cookie cutter of those deals or whether there are meaningfully greater challenges for some companies and when we should expect something from Novartis?"
— Matthew Weston, UBS

A: "That said, I think, as you point out, there are I think now three companies that have reached agreements with the administration. I'd say Novartis has -- I can't speak to what other companies are doing. We've been in conversations with the administration since the beginning of the year as we've had the various turns in these discussions. And I'd say we're meeting with the administration weekly to look at what are the best solutions we can come up with."
— Vas Narasimhan, CEO

Assessment: Weekly meetings and no agreement, eleven months into the process and after three peers have signed. The honest reading is that Novartis's exposure differs enough from the signatories' that a template does not fit, which is information, but not the kind an investor can size. This stays an unquantified overhang into 2026.

The scope and persistence of the gross-to-net true-up

Three separate questions probed the $180M adjustment. The most useful established which brands carried it and, critically, that the Entresto portion self-extinguishes as the brand erodes, while the Cosentyx portion sits on a brand that stays.

Q: "So one quick question on the rebate adjustment. Is there anything you can call out other than the Cosentyx? And also, did any drug benefit from the rebate adjustment in the Q3?"
— Qize Ding (for Simon Baker), Rothschild & Co Redburn

A: "So overall, of course, when you see the amount that is prior period is roughly $180 million. You see that this has about this 1.5 almost rounding the 7% to 9%, if you will, effect on the quarter. And Cosentyx is a big piece of it. Another big piece of it is Entresto actually where patients got quicker into the catastrophic as part of the Medicare Part D redesign. And of course, that part really should go away as Entresto kind of goes away."
— Harry Kirsch, CFO

Assessment: The split matters more than the total. The Entresto component disappears with the brand, so it is a 2025 problem. The Cosentyx component sits on a brand management is guiding to $8B, so it is a recurring modelling input. Management's later clarification that the adjustment relates mainly to the second quarter of this year, and that Q3's stated underlying rate already assumes the new channel mix persists, is the sentence a modeller should anchor on.

Whether margins survive a full year of generic exposure

The most direct question of the call asked whether margins can hold flat through 2026 before any acquisition dilution. The answer was a two-sided bridge: gross margin deteriorates on mix, and selling and administrative productivity is the offset.

Q: "So firstly, just a clarification on margins for Harry. So 3Q margins were a little bit below Street. I guess, partly on gross margin, which is sort of first impact from generics. I wonder if you could just talk about gross margin, EBIT margin as we think about a full year of Entresto impact in '26. My simple question is, can you maintain margins stable next year through the full year of generics before we model the underlying Avidity dilution?"
— Sachin Jain, Bank of America

A: "So on the margins, of course, when you have a product like a small molecule, high-priced products like the 3 going off patent, especially Entresto being so big, there's a slight negative mix effect. Now Kisqali is also a super high-margin product, right, and growing significantly. So that's partly offsetting. But we have also a significant productivity efforts, especially in our manufacturing and supply chain. ... And overall, to close that long answer on a short question, basically, the gross margin headwinds, I do expect to be offset by SG&A productivity."
— Harry Kirsch, CFO

Assessment: This is the most load-bearing answer on the call and it is a commitment, not a forecast. Core gross margin already contracted 44 basis points in the quarter while core SG&A fell 68 basis points, so the mechanism is visible and it worked. The concern is that an SG&A-funded margin defence has a floor, and it is being asked to absorb both the mix shift and an acquisition's Phase III spend at the same time.

The shape of 2026 and the Part D base effects

A detailed question asked what the quarterly cadence of 2026 looks like given Part D seasonality, the generic calendar, and whatever cost actions are running underneath. The answer declined the self-help framing and reduced the whole thing to two base effects and two variables.

Q: "Just trying to understand the margin cadence over 2026. I know you're not giving a '26 guidance at the moment. But very helpfully, you said it will be a year of 2 halves. ... If you were not cutting the costs, would the cadence be a lot more steeper? And what have your actions done to offset that impact? So what is the mix impact versus self-help?"
— Rajesh Kumar, HSBC

A: "So for next year, you will have a very high base Q1 right, with the 2 points of growth that we got from the -- and you will have a relatively low base in Q3 from the 2 negative points this year. And other than that, it's all about launch uptake and generic erosion of the three main products."
— Harry Kirsch, CFO

Assessment: "It's all about launch uptake and generic erosion" is the correct model and also a refusal to quantify either. The usable output is the base map: Q1 2026 laps a flattered comparison and Q3 2026 laps a penalised one, so the first quarter of next year will look worse than the business is and the third will look better. Management separately described 2026 as depressed in the first half and much stronger in the second.

Whether the Kesimpta quarter was clean

Kesimpta's 44% cc growth ran well ahead of its trend, which drew a follow-up on whether inventory or ordering patterns flattered it. The answer was short and unhedged.

Q: "Harry, Kesimpta looks like a very strong quarter in Q3 that looks somewhat off trend. And I'm just making sure that as we go into Q4, we aren't going to learn that it was lumpy one way versus the other. Can you just confirm that was underlying operational growth?"
— Matthew Weston, UBS

A: "Yes, it was mainly underlying operational growth, a little bit of inventory, but not much."
— Harry Kirsch, CFO

Assessment: Taken at face value, and there is no reason not to. Management also disclosed that it sees no impact to date from the competing subcutaneous launch and that roughly 25% of US multiple-sclerosis patients remain untreated with any B-cell therapy. A clean $1.2B quarter growing 44% is the second-largest contributor to the cohort and the least controversial line in the print.

What the raised five-year CAGR is actually built on

The upgraded 2024-2029 sales CAGR from 5% to 6% was attributed half to Avidity and half to the existing business, and a question asked what the existing-business half consists of. The answer pushed the marketed portfolio forward and the pipeline back.

Q: "Novartis raised the long-term revenue guidance yesterday, half of which was attributed to the existing business. Of the half attributed to the existing business, how much is due to currently marketed products? And how much is due to higher sites for the pipeline agents?"
— Stephen Scala, TD Cowen

A: "Yes, Steve, I think we can provide better midterm guidance on that and meet the management. But most of that is in-line brands. ... And there is probably some in there of what we expect will be a strong launch for remibrutinib, so Rhapsido and the label expansion for Pluvicto. ... I think any other pipeline assets we would expect to have limited ramp in this period, just given how long it takes to ramp up these launches when you think out to '29."
— Vas Narasimhan, CEO

Assessment: This is a better answer than it first appears. A five-year framework carried mostly by brands already selling, with the pipeline explicitly excluded from the period, is a low-assumption framework. It also means the 6% CAGR is not where pipeline optionality is priced, and an investor paying for the pipeline is paying for something outside the guidance horizon entirely.

Why acquire an IL-6 asset eight years after CANTOS

The final question asked why the company was returning to inflammation-driven cardiovascular risk reduction via the Tourmaline acquisition, given its own earlier trial in the field produced a commercially inadequate effect size. The answer named the design change that is supposed to make the difference.

Q: "Given the proof of concept established by the CANTOS trial 8 years ago, what new evidence compelled Novartis to go down the same pathway and acquire Tourmaline at this time?"
— Stephen Scala, TD Cowen

A: "So I think we clearly understand that IL-1 beta and hitting the inflammasome has a powerful effect on cardiovascular risk reduction. But in that trial, where we did an all-comers study of patients who had a prior event without, I think, focusing down, you saw the challenge of having a significant CVRR. ... And we think by targeting now prospectively patients right after an event who are at elevated CRP levels as a marker of elevated inflammation, we can then have a much more compelling cardiovascular risk reduction than the kind of 14%, 15% that we saw in the CANTOS study."
— Vas Narasimhan, CEO

Assessment: The thesis is patient selection rather than mechanism, which is a reasonable conclusion to draw from a trial that worked biologically and failed commercially. It is also the same bet the company is running in Lp(a), where the readout has already moved. Management is accumulating cardiovascular outcomes exposure faster than it is resolving any of it, and a competitor is ahead on this specific asset.

What They're NOT Saying

1. No 2026 guidance of any kind, in the quarter the numbers changed. Management pre-announced a margin headwind, described the year as depressed in the first half, and declined to guide. The CEO's response to a direct question on 2026 earnings was that the company is "focused on clearing out 2025." Pre-announcing the bad news without the framework it belongs in leaves consensus formally stale for fourteen weeks.
2. No Avidity funding detail. A $12B all-cash consideration was quantified for its margin effect and not for its balance-sheet effect. Nothing was said about debt issuance, about the interaction with a new $10B buyback authorisation, or about what happens to the core net financial expense line in 2026. No analyst asked.
3. No update on pelacarsen or the Lp(a) outcomes trial. The company's largest binary cardiovascular readout went unmentioned in the release, unmentioned in prepared remarks, and unasked in seventeen questions. Silence on the biggest single swing factor in the pipeline, in a quarter where management bought another cardiovascular inflammation asset, is conspicuous.
4. No quantification of what the raised CAGR assumes for the acquisition. The 5%-to-6% step-up is described as roughly half Avidity, half existing business. Avidity's contribution is guided as "multi-billion-dollar sales contribution by 2030," outside the CAGR window that ends in 2029. The arithmetic of a 2029-ending CAGR being lifted by the acquisition was not quantified beyond the slide's promise of substantial sales growth by 2029.
5. No commitment on the US pricing agreement. Management would say only "in the coming weeks and towards the end of the year." There is no stated magnitude, no stated structure, and no statement of whether the 2026 framework being built assumes an agreement or its absence.
6. Nothing on Xolair or Tafinlar+Mekinist decelerating. Both grew materially below their nine-month rates in the quarter and neither appeared in the priority-brand commentary or drew a question. The interim financial report notes only that a Xolair biosimilar was introduced in some European markets in September. The erosion cohort widening beyond the three named LOE brands is the drift that becomes a narrative late.
7. No peak-sales update on the six growth brands. Cosentyx's $8B was reaffirmed at the brand level, and Pluvicto's $5B path was restated as conditional on the hormone-sensitive label. Kesimpta, Scemblix, Leqvio and Fabhalta, collectively $2,037M in the quarter, carry no updated peak framing at all.

Market Reaction

• Pre-print setup: NVS closed at $129.14 on October 27, up 32.7% year to date against the S&P 500's +16.9%, up 4.4% over the trailing 30 days and up 11.7% over the trailing twelve months. The 52-week closing range entering the print was $96.77 to $132.39, so the stock sat 2.5% below its 52-week closing high.
• Reaction session (before-open reporter, so the print day is the reaction day): opened at $125.52, a 2.8% gap down, traded a $123.06 to $125.75 range and closed at $123.43, down 4.4% or $5.71, on a session when the S&P 500 rose 0.2%.
• Volume: 4.6 million shares against a 30-day average of 1.5 million, 3.2 times normal.
• Shape of the session: the stock gapped down 2.8% to open at $125.52, printed a session high of $125.75 just $0.23 above that open, and closed at $123.43, within $0.37 of the session low of $123.06.

A 4.4% decline on 3.2 times average volume, closing near the low, following an in-line revenue print, a core operating income beat and a reaffirmed full-year guide, is not a reaction to the quarter. It is a reaction to the next five quarters. Three things landed together and all three cut the same way: constant-currency core operating income growth fell from 21% to 7%, the fourth quarter was guided to flat-to-down on a reported basis, and the chief financial officer told analysts to take one to two margin points out of 2026 models that will not be formally reset until February.

The setup amplified it. The stock came in up 32.7% year to date, nearly double the index, and 2.5% off a 52-week closing high, which is a position with no cushion for a guide-down. The stock never traded more than $0.23 above its opening print and finished $0.37 off the low, so the gap down was not faded at any point in the session. The distinction matters: a gap that fills is positioning, a gap that never fills is estimate revision.

The most telling omission is what the price did not respond to. Novartis reported $6.2B of quarterly free cash flow, cut net debt by $3.4B sequentially, and disclosed two positive Phase III trials in a disease with no approved therapy. None of it mattered on the day, because none of it changes the 2026 earnings line, and the 2026 earnings line is the only thing this print changed.

Street Perspective

Debate: Is the deceleration the cliff arriving, or one distorted quarter?

Bull view: The bull case on the Street is that Q3 is mechanically depressed and the underlying rate is 9% on sales and 11% on core operating income. Strip out $180M of prior-period true-ups that relate mostly to the second quarter, and this is a high-single-digit grower absorbing the worst quarter of its LOE calendar without breaking the margin. Nine months at 11% sales growth and a 41.2% core margin is the real company.

Bear view: The bear camp points out that the company's own chief executive said the year-to-date nets out, which means the first quarter was flattered by the same mechanism. On that basis the nine-month reported numbers already are the underlying numbers, and the quarterly path from 27% to 21% to 7% in core operating income is exactly what it looks like. The fourth quarter is guided flat, which is the company's own forecast of no growth.

Our take: The bears have this one, and management handed it to them. If Q1's two points of positive true-ups offset Q3's two points of negative ones, the adjusted-Q3 argument cannot be made without also adjusting Q1 down, and almost nobody will do that. The deceleration is real, it is structural rather than accounting, and the guided fourth quarter confirms it rather than contradicting it.

Debate: Does the growth cohort outrun the cliff, and on what timetable?

Bull view: The optimistic argument is that this quarter is the proof. Six brands added $1,495M of year-on-year revenue and reached 28.3% of net sales from 19.0%, in the exact quarter Entresto lost US exclusivity, and the company still grew 7% cc. Kisqali is growing 91% in the US into an exclusive early-breast-cancer population where more than 60% of patients are not on a CDK4/6 inhibitor. Rhapsido, Pluvicto in hormone-sensitive disease and ianalumab are all incremental from here.

Bear view: The sceptical argument is about timing rather than direction. The cohort grew $506M sequentially while Entresto fell $480M, so it barely covered one brand's decline in the entry quarter, and Entresto has three more quarters of erosion ahead of it. Meanwhile the replacement revenue carries a worse gross margin, which is why core cost of goods rose 95 basis points.

Our take: Both are right and they resolve at different horizons. The cohort wins over three years; it does not win over the next three quarters. An investor with a 2028 horizon should not care about this debate, and an investor with a 2026 horizon should care about nothing else. The mix-driven gross-margin erosion is the under-discussed part and it is permanent, not transitional.

Debate: Is Avidity a good deal or an expensive one?

Bull view: The constructive view is that $12B buys three late-stage neuromuscular assets with loss of exclusivity beyond 2042 and no US drug-pricing negotiation exposure, plus an extra-hepatic delivery platform, in a company whose entire problem is that its best assets go off patent in the 2030s. The 2024-2029 CAGR moves from 5% to 6% on the strength of it. Three years of one to two margin points is a small price for a decade of duration.

Bear view: The critical view is that the company is paying $12B in cash for assets whose revenue contribution management itself places at 2029 and beyond, while carrying a $10B buyback authorisation and a dividend that consumed $7.8B this year. The margin dilution is certain and immediate; the contribution is uncertain and distant.

Our take: We think the deal is strategically right and the disclosure was exemplary, and neither of those is an argument for owning the stock today. The specific problem is sequencing: the cost was quantified in October and the offsetting framework does not arrive until February. Buying a stock during a period when the company has said consensus is wrong, and has not yet said what it should be, is a decision to take avoidable estimate risk.

Model Update Needed

Valuation impact: At $123.43 the ADR trades at roughly 13.8x our $8.90 to $9.00 FY2025 core EPS estimate and roughly 13.6x to 14.0x a FY2026 estimate of $8.80 to $9.10. That is a fair price rather than a cheap one once the denominator is adjusted for what management has already told the Street is coming. A mid-teens multiple on a 2026 number that has been formally guided, rather than informally pre-cut, would be the entry point. We are not setting a price target at initiation; the earnings base resets on February 4 and a target struck against a number the company has disowned would be decoration.

Thesis Scorecard Post-Earnings

This is our initiation on NVS. The pillars below are established here rather than carried forward, and their status tags are the baseline against which every subsequent quarter is scored.

Overall: Thesis established. The operating case is stronger than the share-price reaction implies, and the transition from the old portfolio to the new one is visibly working on the only metric that settles it. The offsetting problem is that every bull pillar pays out beyond 2026 and every bear point is concentrated inside it, and the company has pre-announced that the 2026 number will be lower without yet saying how much lower.

Action: Hold. We want to own this business and we do not want to own the next fourteen weeks. The specific conditions that would move us to Outperform are a formal 2026 guide at or better than flat core EPS on February 4, a signed US pricing agreement with a disclosed magnitude, or a fourth quarter that comes in above the flat simulation. A downgrade would require the growth cohort to decelerate below the pace of Entresto's erosion, or for the SG&A productivity that is currently holding the margin to run out before the acquisition dilution does.