Novartis Clears the 40% Margin Bar Two Years Early, Then Guides to Its First Core Profit Decline in Five Years

Results vs. Consensus

Novartis is a Swiss foreign private issuer reporting under IFRS. Its primary operating measure is "core" (non-IFRS) results, and it guides in constant currencies. Every figure below traces to the February 4 media release; growth rates are as reported by the company, with the currency basis labelled where it matters.

Q4 2025 Scorecard

Consensus for core EPS spans $1.97 to $1.99 across the sources we track, and for net sales $13.7B to $13.9B. We use the more widely cited figures above; on the wider revenue range the shortfall is 2.7% to 3.9%. Where a consensus figure is not reliably available for a line, the column reads "n/a" rather than a manufactured number.

Q4 2025 vs. Q4 2024

FY2025 vs. FY2024

The sum of the four reported quarters ties to the full-year net sales figure of $54,532M as presented by the company; the FY numbers above are the income-statement figures from the release, not aggregations of brand-level or verbally-quoted amounts.

Quality of Beat/Miss

• Revenue: Low quality on the headline, better underneath. The company's own Q4 bridge decomposes to volume +18pp, generic competition -15pp (including -3pp from out-of-period US revenue deduction adjustments, mainly Entresto and Promacta), pricing -4pp and currency +2pp. Eighteen points of volume growth is the operating business; the -1% cc print is the LOE cohort and a one-time accounting true-up sitting on top of it. Management put underlying Q4 growth at +3% excluding the adjustment.
• Margins: Mixed quality. Core operating margin of 37.0% expanded only 10bps in USD (70bps cc), and the release attributes the gain to "higher government grant income and lower SG&A expenses, partly offset by higher R&D expenses." Grant income is not a durable margin lever. The SG&A restraint is more interesting, and management was explicit that Q4 SG&A was deliberately flattened rather than allowed to spike as in prior years, which makes the Q4 margin comparison for 2026 harder rather than easier.
• EPS: Core EPS of $2.03 (+2% cc) grew faster than core net income, which fell 1%. The entire delta is the share count: the company retired 66.9 million net shares in 2025 against $8.9B of repurchases. IFRS EPS tells the harsher story, down 11% on a 15% decline in net income driven by higher income taxes. Investors who anchor to core EPS should note that the buyback, not the P&L, carried the quarter's per-share growth.
• Cash: Q4 free cash flow of $1,655M fell 54%, on operating cash flow down 46%. The full year still set a record at $17,596M (+8%), so this is a quarterly phasing issue rather than a break in cash conversion, but it is the line to watch into H1 2026 given the debt-funded acquisition ahead.

Brand and Franchise Performance

Novartis has reported as a single pure-play Innovative Medicines business since the Sandoz separation, so the meaningful decomposition is by brand rather than by segment. The top 20 brands produced $10,765M in Q4 (81% of net sales) and $44,513M for the year (82%). The right way to read the quarter is as two portfolios moving in opposite directions inside one revenue line.

The growth portfolio

The six brands management designates as priority growth drivers (Kisqali, Kesimpta, Pluvicto, Scemblix, Leqvio, Fabhalta) contributed $4,035M in Q4, or 30.3% of net sales, and $14,191M for the year, or 26.0%. Those two percentages are the single most important pair of numbers in this release. A year ago this cohort was a story about the future. It is now roughly a third of the revenue base in any given quarter, which is what makes management's claim that it can grow through the patent cliff arithmetically plausible rather than aspirational.

Kisqali

The headline was the quarter's most misread number. Reported Q4 growth of +44% cc against a Street modelling roughly $1.5B looked like a material deceleration; the release discloses the same line at +54% cc excluding the out-of-period US revenue deduction adjustment, and management put US growth ex-adjustment at 62%. The underlying volume story is intact: US early breast cancer new-patient share above 60% and holding, German early breast cancer share above 80%, with Italy and Spain launching in 2026. The five-year NATALEE analysis presented at ESMO showed a 28.4% reduction in recurrence risk versus endocrine therapy alone and a 29.1% distant disease-free survival risk reduction.

"So in our view versus the consensus, the entire miss really came from these onetime RD adjustments. We remain fully confident on the $10 billion peak sales outlook for the brand." — Vas Narasimhan, CEO

Assessment: We accept the explanation, because the release corroborates it with a disclosed ex-adjustment growth rate rather than leaving it as a verbal assertion. The real Kisqali question is not this quarter's gross-to-net; it is the mid-2031 US loss of exclusivity that sits just beyond the guidance horizon, and whether the CDK2 and CDK4 programs now in the clinic can extend the franchise.

Cosentyx

The one large brand that beat Street expectations outright, at $1,807M versus roughly $1.6B modelled, growing 11% cc in the quarter and 8% for the year to $6,668M. US growth of 9% came from hidradenitis suppurativa and the IV formulation rather than the mature psoriasis base. Novartis holds 51% new-patient share among biologic-naive HS patients and describes that naive market as 2.5 times the size of the switch market, where a competitor is performing better. The Phase III REPLENISH study in polymyalgia rheumatica met its primary endpoint, with submission planned in H1 2026.

Assessment: Cosentyx is the portfolio's dependable mid-single-digit compounder on the road to $8B peak, and the PMR indication is a genuine incremental leg. It is also on the list of brands facing exclusivity loss later in the guidance window, which is why management frames it as mid-single-digit growth "on average" rather than accelerating.

Pluvicto and the radioligand platform

Pluvicto crossed $2B for the year and grew 70% cc in the quarter, with US sales up 75% on the pre-taxane launch and PSMA share in that setting quadrupling to 16% since approval. The infrastructure build is the underappreciated part: over 790 US community treatment sites, more than 440 ex-US sites, and new manufacturing coming online in California, Florida, Japan and China. The PSMAddition submission in hormone-sensitive prostate cancer is filed with FDA, NMPA and PMDA, and adds roughly 75% more patients on top of the existing VISION and PSMAfore populations, with two-thirds of eligible patients already sitting with existing treaters.

"So we've really taken this to scale, which positions us well for the Pluvicto launches, ongoing Lutathera business as well as our future RLT portfolio." — Vas Narasimhan, CEO

Assessment: This is the most defensible asset in the portfolio, because the moat is manufacturing and logistics rather than molecule. Radioligand therapy has a half-life measured in days; a competitor with a better molecule still needs the site network and the isotope supply chain. The mHSPC label expansion is the largest single revenue catalyst in the 2026 to 2027 window that does not depend on a coin-flip readout.

Scemblix and Leqvio

Both crossed blockbuster status in the quarter. Scemblix grew 87% cc to $391M with 41% US new-patient share across all lines, and EC approval for all lines of treatment now in hand; first-line US share is in the mid-20s against an ambition of 40% to 50%. Leqvio grew 46% cc to $335M, and the China NRDL listing that began in January is the swing factor for 2026 given how strong the private-pay uptake had already been.

Assessment: Scemblix's first-line share has looked like it was plateauing for several quarters, and management's answer (noisy rare-disease data, access perception rather than access reality, Gleevec loyalists) is plausible but unfalsifiable in any single quarter. Leqvio's China listing is the cleaner near-term catalyst. Both remain on track for their peak-sales guidance.

The renal franchise: Fabhalta, Vanrafia, zigakibart

Fabhalta grew 167% cc to $155M in the quarter and 287% for the year to $505M. The IgAN portfolio contributed half of new-patient market growth, split evenly between Vanrafia and Fabhalta, and the Phase III APPLAUSE-IgAN final analysis showed statistically significant superiority in slowing IgAN progression on two-year eGFR slope. Against that, management amended the zigakibart Phase III protocol to align the proteinuria readout with the interim eGFR readout, pushing it to H1 2027.

Assessment: The strategic logic of the zigakibart delay is sound (arriving with both proteinuria and eGFR data into a crowded anti-APRIL field beats arriving early with half a package), but it is still a nine-month slip disclosed alongside a soft quarter, and it removes a 2026 catalyst. The three-asset renal position is the more durable point: if Fabhalta, Vanrafia and zigakibart all land eGFR outcomes benefits, Novartis owns the nephrologist's office in a way no competitor currently does.

The eroding portfolio

Assessment: Entresto, Promacta and Tasigna together produced $1,658M in Q4, or 12.4% of net sales, versus $10,488M or 19.2% for the full year. That gap between the quarterly and annual share is the erosion happening in real time: the cohort was a fifth of the business across 2025 and is already down to an eighth by the fourth quarter. Note also that Xolair and Tafinlar+Mekinist both grew for the year but declined in Q4, which means the eroding cohort is widening at the margin rather than staying confined to the three named brands.

Key Topics & Management Commentary

Overall Management Tone: Management was confident and unusually well-rehearsed on the parts of the story it controls, treating the 40% margin achievement and the growth-brand trajectory as settled facts rather than claims requiring defense. Where the tone changed was on the pipeline's binary events: on the cardiovascular outcomes readout the CEO volunteered that the company is blinded and cannot explain its own trial's event rate, which is a level of candour that reads as honest rather than evasive but also removes any pretence of forecastability. The handover of the CFO seat was managed with the guidance itself delivered by the incoming CFO, a deliberate signal that the 2026 numbers belong to the new tenure.

1. The gross-to-net adjustment and what Q4 actually was

Three points of the fifteen-point generic drag were out-of-period US revenue deduction adjustments, concentrated in Entresto and Promacta but also hitting Kisqali. The release makes the accounting visible by disclosing both a headline and an ex-adjustment growth rate for the three affected brands, and footnotes that there was "no significant impact on full year." Management was consistent on this across prepared remarks and Q&A.

"...so excluding this adjustment, underlying quarter 4 sales growth would have been positive 3%." — Harry Kirsch, CFO

The mirror-image mechanics matter for 2026 modelling: Q1 2025 carried a positive gross-to-net adjustment of roughly 2 points, which becomes a hard comparison in Q1 2026, while Q4 2026 laps an easier base because of this quarter's negative.

Assessment: We treat +3% cc as the honest read of the operating quarter and -1% as the reported one, and we do not think the distinction changes the investment case in either direction. What it does change is the shape of the 2026 quarterly line, and anyone modelling a clean Q1 recovery is going to be surprised by a comparison that management has now flagged twice.

2. Hitting 40% core margin two years early

The 40.1% FY core operating margin was targeted for 2027 at the Capital Markets Day and arrived in 2025, on 210bps of constant-currency expansion and more than 1,000bps cumulatively over five years. Over that period sales compounded at 8% and core operating income at 15%.

"I hope you agree, but I believe we have really elevated the company to a new level of sales performance, margin profile and as I'll discuss later, free cash flow generation." — Harry Kirsch, CFO

Assessment: This is the single strongest fact in the release and it cuts both ways. It validates the operating discipline of the pure-play transition, and it also means the easy margin expansion is behind the company. Guidance now calls for a return to 40%+ only in 2029, which is an admission that 2026 through 2028 is a margin plateau at best.

3. The 2026 guide and its two halves

Full-year 2026 guidance is net sales growth of low-single-digit cc and core operating income decline of low-single-digit cc, delivered by the incoming CFO rather than the outgoing one. The intra-year shape is what makes it navigable.

"With that, we expect the first half of the year sales to decline low single digit and core operating income to decline low double digit... That said, in the second half of the year, we expect a clear improvement with sales growing mid-single digit and core operating income growing mid- to high single digit." — Mukul Mehta, CFO Designate

Supporting assumptions given on the call: core net financial income and expense of around $1.7B (materially above 2025, on Avidity funding costs), a core tax rate around 16.5%, and an FX tailwind of 2 to 3 points on sales and 1 point on core operating income at late-January rates.

Assessment: A guided low-double-digit core operating income decline in H1 is a real earnings hole, and it is not obvious why an investor needs to own the equity through it when the second-half inflection is both visible and, on management's own framing, dependent on generics annualising rather than on anything new being executed. The guide is credible. It is also a reason to wait.

4. Avidity: the dilution management says the Street missed

The proposed acquisition of Avidity Biosciences brings late-stage RNA-delivery neuroscience programs (DM1 and FSHD) and is expected to close in H1 2026, primarily debt funded. Management quantified the margin effect as 1 to 2 percentage points and was pointed about having disclosed it in October.

"So not everybody has figured this into the consensus. It's okay when people don't follow everything we say, but we have mentioned it to you. And 1 to 2 points, if you take 1.5, it's pretty much what you get when you have a low single-digit increase on sales and a low single-digit decrease in core operating income. So we feel we have implemented exactly that without Avidity would have been unchanged margin basically." — Harry Kirsch, CFO

Assessment: This is the most useful disclosure on the call, because it decomposes the guidance into an operating component and a deal component. Read literally, ex-Avidity the company would be guiding to a roughly flat core margin on low-single-digit sales growth, which is a materially better underlying picture than the headline. It also means the guide is not a statement about the base business deteriorating, and anyone downgrading the operating thesis on this print is reading the wrong line.

5. The US most-favoured-nation pricing agreement

On December 19, 2025, Novartis reached an agreement with the US government aimed at lowering US prices for innovative medicines in exchange for support for continued US manufacturing and R&D investment. The release states the implications are already reflected in both the 2026 guidance and the 5% to 6% 2025 to 2030 sales CAGR. The operational consequence surfaced in Q&A: ex-US launch pricing is now constrained by US net-price reference.

"Ianalumab is more complex as we get to launches in 2027 in the G7 countries. There, of course, it's on the entire market of U.S. net price, not just Medicaid. And so we're working through strategies." — Vas Narasimhan, CEO

Assessment: The quantified impact is inside guidance, so the near-term risk is contained. The structural point is not. A newly-launched asset whose ex-US price now feeds back into its US net price faces a genuinely different launch calculus, and management's own answer distinguishes between Rhapsido (manageable, already launched, Medicaid-rebate exposure only) and ianalumab (harder, whole-market exposure, launching 2027). That is a two-tier problem the company has not yet solved and did not pretend to have solved.

6. The CFO transition

Harry Kirsch confirmed this was his final earnings call after 13 years as CFO and 23 years at the company, handing over to Mukul Mehta in mid-March. Mehta delivered the 2026 guidance section himself.

"Before handing over, I'd like to briefly acknowledge that this will be my final earnings call as CFO of Novartis. It has been a privilege to serve in this role in the last 13 years." — Harry Kirsch, CFO

Assessment: A well-choreographed handover with continuity: Mehta has been at Novartis since 2003 and worked closely with Kirsch for a decade, and the incoming CFO owning the guidance is the right structural signal. The residual risk is that a departing CFO's final guide is by convention a conservative one, and a new CFO's first guide is by convention more conservative still. We would not treat the 2026 numbers as a stretch target.

7. Rhapsido: the launch that could matter most

Rhapsido (remibrutinib) is in early US launch in chronic spontaneous urticaria with over 2,000 healthcare-provider starts through the sampling and bridge program, a clean label (no boxed warning, no contraindication, no required routine lab monitoring, no liver safety issues), and onset described as fast as two weeks. Management framed access as the gating factor, with a steady pickup weighted to H2.

"And I think that when we benchmark that versus other highly successful dermatology launches, it's right in line with some of the most successful dermatology launches." — Vas Narasimhan, CEO

The indication stack behind it is the real asset: chronic inducible urticaria (one subtype submitted, two more reading out in H1 2026), food allergy (positive Phase II, broad Phase III program initiating, data presenting in Q1), relapsing MS (two Phase III readouts mid-2026), hidradenitis suppurativa (2028), plus SPMS and myasthenia gravis.

Assessment: If the MS readouts land, Rhapsido becomes the most important asset in the company, because an oral with a clean label competing against injectable B-cell therapies addresses a population management sizes at 25% of US and 65% of ex-US disease-modifying-therapy patients not on B-cell agents. Management's own base case is notably modest (an oral will not match monoclonal efficacy in MS), and we prefer that framing to the alternative. This is genuine optionality that we are not paying much for today.

8. Pelacarsen and the HORIZON event-rate problem

The Lp(a) cardiovascular outcomes trial reads out mid-2026, later than originally expected because the event rate has come in below the modelled rate. The trial is powered for a 20% relative risk reduction at the 70 mg/dL cutoff and 25% at the 90 mg/dL cutoff, and management confirmed the study is running to its originally specified event count with no protocol changes.

"I wish we knew, Steve. Honestly, obviously, I can only give you an opinion. I can't actually give you a fact because we're completely blinded, and we have no database insights." — Vas Narasimhan, CEO

Management's hypothesis is that optimal management of other risk factors, particularly LDL lowering, has suppressed background event rates, and that Lp(a) becomes the dominant risk factor only at higher thresholds. They also conceded the study can succeed at a mid-teens relative risk reduction, and that such a result would raise a commercial question rather than settle one.

Assessment: This is the largest single binary in the story and the honest answer is that nobody, including management, can currently handicap it. A clean win creates a new drug class and validates an entire Lp(a) portfolio including the annually-dosed partnered asset ready to move straight to Phase III. A statistically significant but clinically marginal win is arguably the worst outcome, because it forces a launch decision on an asset whose value depends on payers accepting a mid-teens risk reduction. We assign this option value, not expected value.

9. Capital allocation and a balance sheet that is no longer pristine

Net debt rose to $21.9B from $16.1B, roughly 1.0x FY core operating income. The bridge is disclosed: $17.6B of free cash flow more than offset by $9.2B of treasury share transactions, $7.8B of dividends and $5.2B of net M&A and intangible outflows. In 2025 the company repurchased 77.6 million shares for $8.9B, completed the $15B program in July and started a new up-to-$10B program with $7.7B remaining, targeted for completion by end-2027. The proposed dividend is CHF 3.70, up 5.7%, the 29th consecutive increase in Swiss franc terms. Long-term ratings are Aa3 and AA-.

Assessment: Leverage of roughly 1.0x core operating income is unremarkable for a AA- credit, and none of this constrains the business. It does constrain the pace of the buyback, which is the mechanism that carried per-share growth this quarter, and Avidity is to be debt funded on top. Investors relying on the buyback to bridge a flat core-earnings year should size that expectation against $7.7B remaining across two years and a balance sheet absorbing an acquisition.

10. The 2025 to 2030 framework

Management reiterated a 5% to 6% sales CAGR for 2025 to 2030 and a return to a 40%+ core margin in 2029, explicitly including the Entresto impact and the US MFN agreement. The bridge shown was front-loaded generic erosion, a middle column of dynamic growth brands, and a probability-weighted pipeline contribution spanning ianalumab, Pluvicto and actinium PSMA, pelacarsen and the Avidity assets.

Assessment: A 5% to 6% CAGR from a $54.5B base through the largest patent expiry in company history is a credible but not conservative target, and it is doing the work of justifying the current multiple. Note what the margin guidance concedes: three years before the company gets back to the level it just reached. The 2029 date, not the CAGR, is the number that should shape position sizing.

11. Kisqali's franchise horizon and the oral SERD question

Kisqali's US loss of exclusivity is guided to mid-2031 with pediatric exclusivity, later in some ex-US markets, which sits just outside the 2030 guidance window. Asked whether oral SERDs threaten CDK4/6 combinations in the adjuvant setting, management argued that physicians want a CDK4/6 inhibitor for eligible patients and will pair it with whichever endocrine therapy becomes standard, including oral SERDs.

"That's what gives us confidence in the $10 billion-plus guidance that we have and are sticking to." — Vas Narasimhan, CEO

The in-house response is a CDK2, CDK2/4 and CDK4 program set, all now in the clinic, plus HER2 and neobombesin radioligand assets in breast cancer.

Assessment: The oral SERD argument is sound as far as it goes: SERDs substitute for endocrine therapy, not for CDK4/6. The genuine issue is that the company's largest future brand goes off patent roughly twelve months after the guidance window closes, which means the 2025 to 2030 CAGR is defined over precisely the period that excludes its biggest cliff. That is not a criticism of the guidance, but it is a reason not to extrapolate it.

Guidance & Outlook

The guidance is unusually legible because management gave the half-year split alongside the annual numbers, which is not their habit. The reason is mechanical: Entresto, Promacta and Tasigna generics entered the US market mid-2025, so H1 2026 laps a pre-erosion base and H2 laps a post-erosion one. Layered on top, Q1 2026 must also lap the roughly 2-point positive gross-to-net adjustment that flattered Q1 2025. The result is a first quarter that will look considerably worse than the business is performing.

Implied core EPS math: taking the midpoint of a low-single-digit core operating income decline against the FY25 base of $21,889M, adding the guided 1-point FX tailwind, deducting the guided ~$1.7B of core net financial expense, taxing at the guided 16.5% and assuming continued buyback at a pace consistent with $7.7B remaining through 2027, we get 2026 core EPS in the region of $8.60 to $8.90 against $8.98 delivered. Note that the step-up in net financial expense alone, roughly $660M versus the ~$1.0B implied in 2025, is about $0.28 of after-tax EPS. Guidance does not include Avidity revenue, and management did not provide a core EPS guide.

Street at: the pre-print consensus had 2026 core operating income above the guided range on our read of the sell-side reaction, with the shortfall concentrated in the Avidity dilution rather than the base business. Estimated consensus reductions circulating on the day were characterised as modest, in the region of half a point at the midpoint, weighted toward H2.

Guidance style: Novartis raised guidance twice during 2025 and then delivered at or above the raised range on both lines, guiding high-single-digit sales and delivering 8%, guiding low-teens core operating income and delivering 14%. That is a company that guides to be beaten. Applied to 2026, a "low-single-digit decline" from a management team with this track record, handing over the CFO seat, is more likely to be a floor than a ceiling.

Analyst Q&A Highlights

Liver monitoring risk in the MS program

The call opened on whether FDA-mandated liver monitoring inside the Phase III trials will end up in the label, a live concern because competitors in the class have carried Hy's law cases into REMS programs and warnings. Management's answer separated the structural argument (the molecule already carries an approved label with no liver safety discussion) from the procedural one (monitoring was added out of caution, and is more limited than competitors').

Q: "You talked about avoiding liver monitoring in MS given no high law. Competition recently has been vocal avoiding monitoring in the label when monitoring has been involved in the studies could be difficult. So I wonder if you could just give us any color on FDA conversations around this topic and whether monitoring in the studies picked up events that required dose changes?"
— Sachin Jain, Bank of America

A: "I think first on liver, I think we should first take a step back and note that we already have an approval and an approved label without any liver safety discussion in the label, which just points to the fact that remibrutinib structurally does not have the off-target toxicities we believe that the structures of some of the other molecules do."
— Vas Narasimhan, CEO

Assessment: Management answered the structural question convincingly and did not answer the specific one, which was whether in-study monitoring picked up events requiring dose changes. The absence of a direct denial is not evidence of a problem, but it is the kind of gap worth revisiting when the readouts land mid-year. The existing clean label in urticaria is a genuine asset in this argument.

Whether the Q4 gross-to-net drag persists into 2026

The most consequential modelling question of the call. The exchange established that the adjustments are out-of-period true-ups in both directions, which sets up a hard Q1 comparison and an easier Q4 one, and that the underlying brand performance is what the ex-adjustment growth rates show.

Q: "So how much of the gross to net impact that was impacting fourth quarter carries through into the next year because of a different channel mix versus how much is one-off? And so to what extent does that give you an easy base for comparison in 2025 into 2026?"
— Graham Parry, Citi

A: "So on Kisqali, I mean, one thing to note is that actually in quarter 1 of '25, with a positive gross to net as Mukul pointed out. So in quarter 1, that's a higher base due to one-timers. As Vas mentioned, the quarter 4, what we have noted here out-of-period adjustments. So if you take that out, it's really the true quarter 4 performance. And then quarter 4 of '26, there should be a bit of a lower base because of this negative this year. But overall, the -- basically, these gross to net adjustments are all out of period, so one-timers and the underlying is what you see."
— Harry Kirsch, CFO

Assessment: A clean, specific answer that a departing CFO had every incentive to leave vague. The disclosure of both directions, positive in Q1 2025 and negative in Q4 2025, is what makes it credible; a company managing perception would have flagged only the one that helps. This exchange is the reason we accept the +3% cc underlying read on the quarter.

Whether Q4 SG&A discipline is repeatable and what it implies for the margin guide

A direct challenge to the arithmetic of the 2026 guide: if the fourth quarter's cost control persists, the guided margin dilution looks too large, particularly if the acquisition closes later than assumed. The response separated a deliberate phasing change from ongoing productivity, and then restated the deal dilution as the whole of the guidance delta.

Q: "The SG&A expenses in the fourth quarter were extremely tight. Very good performance there, helped you to gear the margin in underlying terms. As I think about the margin for 2026, obviously, you do have the Avidity dilution. But if that SG&A control continues, I struggle to see how you're going to get as much dilution, especially if avidity doesn't quite close as quickly as maybe it could. So can you just talk about the repeatability of that SG&A performance in the fourth quarter into 2026?"
— Michael Leuchten, Jefferies

A: "Historically, we always had quite an increase in quarter 4. So we took this year to say, look, this is inefficient to have such a peak in a quarter where you have 1 to 2 weeks of Christmas and you have also the U.S. Thanksgiving and so on. It shouldn't be actually a big peak here. So we took that in order to even a bit out."
— Harry Kirsch, CFO

Assessment: The most valuable exchange on the call, and the answer was better than the question expected. Q4 SG&A was not a productivity windfall, it was a phasing decision to stop back-loading spend into a short quarter, which means it does not repeat as incremental savings but does not reverse either. Combined with the follow-on that ex-Avidity the margin would be "basically unchanged," this decomposes the entire guidance cut into one identifiable, disclosed, temporary line item.

What a win looks like on the Lp(a) outcomes trial

The questioning returned repeatedly to the delayed cardiovascular outcomes readout, pressing on whether a statistically significant but smaller-than-powered benefit would still be commercially viable. Management conceded the trial can win below its powering assumption and declined to say what it would do in that case.

Q: "Where is the latest magnitude of efficacy? I think the original design was a 20% benefit in the broader population, a 25% benefit in a narrow population with a longer study and maybe some other tweaks. Is that sort of the minimum? Is there a possibility that you could actually have a benefit for either of those groups that were statistically significant, but didn't quite hit the her? And if so, would that still be a product with strong commercial prospects?"
— James Gordon, Barclays

A: "We can win on the study with a relative reduction that's lower than that. And so certainly, there is the opportunity to win with CVRR in the mid-teens. I think we have to evaluate, I think, for patients who have no other option. And if we were to win at that lower level, what would be the right approach to bring it to market. And that's something we'll have to see based on the data."
— Vas Narasimhan, CEO

Assessment: An unusually candid answer that widens rather than narrows the distribution of outcomes. "What would be the right approach to bring it to market" is not the language of a company confident it will have a commercially compelling result at the low end. The readout is now a three-way outcome (clear win, marginal win, failure) rather than a binary, and the middle branch is the one the market is least likely to have priced.

Ex-US launch strategy under most-favoured-nation pricing

A line of questioning on how the new US pricing framework changes rest-of-world launch plans for assets that have not yet launched globally. The answer drew a sharp distinction between an already-launched asset with limited exposure and a 2027 launch with whole-market exposure.

Q: "Given we're now in an MFM world, how should we be thinking about ex U.S. launch plans for what are clearly mostly significant assets. I'm basically just pushing my life to see how specifically you're comfortable being about changing in rest of world launch strategies for important assets like ianalumab."
— Peter Verdult, BNP Paribas

A: "Absolutely, it's our aspiration to get these medicines launched in all of these markets given the patients that need them. But we certainly can't adversely affect the U.S. market. And so we're just going to have to be thoughtful about looking at where are there opportunities to price appropriately for the value that ianalumab brings."
— Vas Narasimhan, CEO

Assessment: "We certainly can't adversely affect the U.S. market" is the sentence that matters. It is an explicit statement that ex-US launch scope is now subordinate to US price protection, which is a structural change to how a global launch gets planned. Management said the strategy is still "in the works" and promised more clarity during the year. That is an honest non-answer and a genuine open risk to ex-US revenue on the 2027 launch cohort.

Frontline share in CML and whether the ramp has stalled

A challenge on whether first-line share for the CML franchise has plateaued in the US over recent quarters. The response attributed the appearance to data noise in a rare disease, an access-perception gap rather than an access gap, and a residual cohort of prescribers loyal to the legacy standard of care.

Q: "Vas, just on Scemblix, you helpfully provide the share data across lines of therapy for the product. It looks like it's plateauing in first line in the U.S. a little bit over the last few quarters. What's stopping the momentum to continue?"
— Michael Leuchten, Jefferies

A: "One thing to note is that data is very noisy because with CML, it's a rare disease, most physicians only see 1 or 2 patients. And so the data here always is getting restated. Overall, our view based on our internal assessments is we continue to see steady share growth on the frontline setting. Actually, I would say our frontline share growth is ahead of our plan and our original planning assumptions."
— Vas Narasimhan, CEO

Assessment: The claim that frontline share growth is ahead of internal plan is not verifiable from external data, and the acknowledgement that third-party scripts get restated is a convenient defence against any quarter that looks soft. That said, the diagnosis is specific and testable: if the constraint is access perception rather than access, share should step up as the message lands, and mid-20s to 40% is a large enough gap that progress will be visible within a few quarters. We will hold management to it.

Rationale for delaying the fourth renal asset

On why the anti-APRIL program's readout was pushed to align proteinuria with interim kidney-function data rather than reading out earlier on the surrogate alone. The answer was a competitive-positioning argument in a field filling up quickly.

Q: "The data has been pushed out a little bit in order to have the eGFR data on the label at launch. But as you think about sort of the strategy here with your other assets, whether you had the UPCR data first and then adding the eGFR data. Is this a case that the data were quite close together so it is worth having a delay, just trying to understand the rationale here versus the mechanism?"
— James Quigley, Goldman Sachs

A: "I think when we looked at the number of competitors entering in the [ anti-APRIL ] space. We asked ourselves, given we already have a strong portfolio in the nephrologist office, what would give us the most compelling data package to kind of cut through all of the various launches that are ongoing... So we thought that was prudent given that the time that passes here is three quarters, it's not the end of the world, and then we would have a much more compelling data set to provide to FDA."
— Vas Narasimhan, CEO

Assessment: A defensible strategic trade that is nonetheless a delay, and the framing "it's not the end of the world" is doing some work. The stronger part of the answer is the portfolio logic: three renal assets each carrying kidney-function outcomes data would be a category-defining position with the prescribing specialist, and arriving with a half-package into a crowded field would waste that. We accept the reasoning and note the lost 2026 catalyst.

What They're NOT Saying

1. No core EPS guidance. Novartis guides sales and core operating income only. With core net financial expense rising to ~$1.7B and the buyback pace constrained by the Avidity funding, the gap between core operating income and core EPS is wider and less predictable in 2026 than it has been. Guiding the top of the P&L and not the bottom is convenient in exactly the year the bottom gets harder.
2. No quantification of the MFN agreement. The release says the December 19 agreement's implications are "reflected in" 2026 guidance and the five-year CAGR. It does not say how many points of either. An investor cannot separate the policy drag from the generic drag, which means they cannot tell whether the 5% to 6% CAGR absorbed a small concession or a large one.
3. No Avidity revenue or accretion timeline. Management quantified the margin dilution precisely (1 to 2 points) and said nothing about when the DM1 and FSHD programs contribute revenue or when the deal turns accretive. The cost is guided; the benefit is not.
4. No answer on whether in-study liver monitoring caught events. The MS liver-safety question had two halves. The structural half was answered at length. The specific half, whether monitoring in the trials picked up events requiring dose changes, was not addressed.
5. No 2026 quarterly shape beyond the H1/H2 split. Management flagged that Q1 laps a positive gross-to-net comparison and guided halves rather than quarters. Given how much this quarter's print turned on out-of-period adjustments, the absence of quarterly guidance in a year with known base distortions leaves the first print unusually wide open.
6. No update on the residual buyback pace. $7.7B remains under a program targeted for completion by end-2027, alongside a debt-funded acquisition and net debt already up $5.8B year over year. Management restated the capital allocation framework as unchanged without committing to a 2026 repurchase run-rate.
7. Little said about Xolair and Tafinlar+Mekinist rolling over. Both grew for the full year and declined in Q4. Neither featured in the priority-brand commentary or drew a question. The erosion cohort widening beyond the three named LOE brands is the kind of drift that only becomes a narrative two quarters late.

Market Reaction

• Pre-print setup: NVS closed at $149.86 on February 3, up 8.7% year to date against the S&P 500's +1.1%, up 8.7% over the trailing 30 days and up 42.1% over the trailing twelve months. The 52-week closing range entering the print was $101.96 to $152.22, so the stock was sitting within 1.6% of its 52-week closing high.
• Reaction session (before-open reporter, so the print day is the reaction day): opened at $153.82, a 2.6% gap up, traded a $151.42 to $154.71 range and closed at $152.91, up 2.0% or $3.05, on a session when the S&P 500 fell 0.5%.
• Volume: 4.4 million shares against a 30-day average of 1.6 million, 2.8 times normal.
• Listing divergence: the US ADR closed up 2.0% while the Swiss ordinary line was quoted modestly lower during the European session. The two listings read the same release differently, and coverage on the day carried both "earnings beat, stock jumps" and "shares slip as record patent expiry overshadows profit beat" headlines.

The ADR's reaction is best understood as a beat-and-shrug rather than a re-rating. A 2.0% move on 2.8 times volume, into a fresh closing high, on a day the index fell, is a market that liked the core EPS beat and the record cash flow and declined to punish a guide it had been pre-warned about in October. That the stock finished 1.2% below its intraday high, having gapped up 2.6%, is the more informative detail: buyers faded through the session as the market worked through the 2026 core operating income line.

The divergence between the two listings is worth sitting with. The ADR buyer was pricing a $2.03 core EPS beat and a record $17.6B of free cash flow. The Swiss holder, who has ridden the stock up 42% over a year and holds it as a domestic large-cap staple, was pricing a company that has just told them core profit falls next year and does not get back to a 40% margin until 2029. Both are correct. The gap between them is the gap between a quarter and a cycle, and it is the reason this initiation lands at Hold rather than at either extreme.

Street Perspective

Debate: Is 2026 a trough to look through or a signal to step aside?

Bull view: The guided decline is almost entirely an acquisition the company disclosed three months ago, and management stated that ex-Avidity the core margin would be roughly unchanged. The base business grew volume 18 points in the quarter. Buying a high-quality compounder in the one year its optics are bad, with a visible H2 inflection and a reiterated 5% to 6% CAGR, is the whole trade.

Bear view: The stock is up 42% in twelve months and sits at a 52-week high going into a year with core earnings down, free cash flow already showing a 54% quarterly decline, net debt up $5.8B, and a margin that does not recover to the level just achieved until 2029. There is no urgency to own the H1.

Our take: The bull case is right about the business and wrong about the timing. The Avidity decomposition is genuinely reassuring and we credit it fully. But an investor who agrees the H2 inflects has roughly two quarters to establish a position at a lower multiple, against a guided low-double-digit core operating income decline in H1 and a Q1 that laps a flattering base. Quality does not require paying up on the first day of the trough.

Debate: Can the growth portfolio actually outrun the largest patent cliff in company history?

Bull view: It already has. FY25 delivered 8% cc growth through the Entresto, Promacta and Tasigna US entries, with the six priority brands now 30% of quarterly sales. Kisqali is heading for $10B, Kesimpta for $6B+, Cosentyx for $8B, Scemblix for $4B+, and Pluvicto has crossed $2B with a label expansion filed that adds 75% more patients.

Bear view: The cliff is not one event. Entresto is the loud one, but Cosentyx and Xolair have exclusivity issues later in the window, Kisqali itself goes off patent in mid-2031 just past the guidance horizon, and Q4 already shows Xolair and Tafinlar+Mekinist rolling over. Each new growth brand has to replace an ever-larger base.

Our take: The bulls have the better of this one on the evidence available. Growing 8% cc through the first year of a record LOE cohort is not a forecast, it is a result, and the growth cohort's 30% share of Q4 revenue is the arithmetic that makes the next few years work. The bear point that matters is not 2026, it is 2031, and it argues for how long to own the stock rather than whether to.

Debate: How should the pipeline optionality be valued?

Bull view: Seven pivotal readouts in 2026, an oral asset with a clean label reading out in MS mid-year with a market sized at the majority of patients not on B-cell therapy, food allergy Phase III initiating off positive Phase II, ianalumab with breakthrough designation in Sjögren's and NEJM-published ITP data, and a first-in-class Lp(a) outcomes trial. None of this is in the multiple.

Bear view: The pipeline is where the disappointments live. The Lp(a) readout has already slipped once because the event rate came in below model and management is blinded and cannot explain it. The fourth renal asset slipped three quarters. Management's own MS base case is that an oral will not match monoclonal efficacy. Optionality that keeps moving right is not free.

Our take: We side with the bulls on breadth and the bears on any single asset. Seven pivotal readouts is a portfolio, and a portfolio with this many shots does not need the marquee one to land. But we would not pay for the Lp(a) outcome specifically, because the CEO's own answer describes a three-way distribution in which the middle branch, a mid-teens risk reduction, creates a launch problem rather than a franchise. Value the breadth, discount the headline.

Model Update Needed

Valuation: at the $152.91 close, NVS trades at 17.0x FY25 core EPS of $8.98 and roughly 17.2x to 17.8x our FY26 core EPS range. On IFRS EPS of $7.21 the trailing multiple is 21.2x. FY25 free cash flow of $17,596M against a market capitalisation of roughly $296B (1,939M shares at $152.91) is a free cash flow yield near 5.9%, which is the most attractive line in the valuation and the one that best reflects the quality of the underlying business. The dividend proposal is CHF 3.70, up 5.7%, the 29th consecutive Swiss franc increase.

Valuation impact: a 17x multiple on a trough-year core EPS is neither demanding nor cheap for large-cap pharma with a 5% to 6% top-line CAGR, an AA- balance sheet and a near-6% free cash flow yield. What it does not offer is compensation for the specific risks in front of it: an H1 with core operating income down low-double-digit, a mid-year cardiovascular readout management explicitly cannot handicap, and an unquantified policy concession embedded in the five-year guide. We see fair value roughly in line with the current price and would want either a mid-teens multiple or a de-risked pelacarsen before paying up.

Thesis Scorecard Post-Earnings

This is our initiation on NVS, so the pillars below are established here rather than carried forward, and the status tags are the starting point against which subsequent quarters will be scored.

Overall: Thesis established. The operating case is stronger than the headline print suggests, and the guidance cut decomposes almost entirely to a disclosed, temporary, deal-related item rather than to deterioration in the base business. The offsetting problem is timing: every bear point above is concentrated in the next twelve months while every bull point pays out beyond them.

Action: Hold. This is a business we would like to own and a price at which we do not need to hurry. The specific conditions that would move us to Outperform are a de-risked mid-year Lp(a) readout, confirmation that H1 tracked to or better than the guided low-double-digit core operating income decline, or a multiple in the mid-teens on 2026 core EPS. A downgrade would require the growth portfolio itself to decelerate, which nothing in this quarter suggests.