Novartis Raises Its Profit Guide in the Quarter Two of Its Three Big Bets Failed

Results vs. Consensus

Novartis is a Swiss foreign private issuer that reports under IFRS and has no 8-K. Its primary operating measure is "core" (non-IFRS) results, and it guides in constant currencies. Every income-statement figure below traces to the Form 6-K furnished on October 27: Exhibit 99.1 reproduces the media release, and Exhibit 99.2 is the Condensed Interim Financial Report. Growth rates are as reported by the company, with the currency basis labelled where it matters.

Q3 2026 Scorecard

Core EPS consensus spans $2.25 to $2.28 across the sources we track, and net sales $14.53B to $14.60B. On the high core EPS mark the beat is +7.9%; on the high sales mark the miss is -0.5%, and on the low one it rounds to -0.1%. We found no consensus for core operating income that two sources confirm, so that column reads "n/a" rather than a manufactured number.

Q3 2026 vs. Q3 2025

Q3 2026 vs. Q2 2026

Nine Months 2026 vs. Nine Months 2025

All four tables are the income-statement and cash-flow figures as the company presents them, not sums of brand-level or verbally quoted amounts. The quarter-over-quarter column is our arithmetic on two reported quarters. Each Q3 figure also equals the nine-month figure less the first half reported in July, so the quarter reconciles to both filings.

Quality of Beat/Miss

• Revenue: Better than our framework and slightly below the Street. The company's bridge is volume +14 points, generic competition -9 points, pricing -1 point and currency zero. In August we wrote that roughly $800M of US Entresto in the Q3 2025 base was a 5.8-point headwind, and that a low-single-digit print would be consistent with guidance. US Entresto fell to $41M, so most of that headwind arrived, and total US sales still grew 7% to $6,429M. The priority brands carried it. The small miss against consensus is a Street that had moved closer to the guided H2 rate than we had.
• Margins: High quality, with one caveat. The 2.3-point expansion in USD breaks down into +0.9 points from core gross profit, +2.6 points from SG&A, a neutral R&D ratio, and -1.2 points from core other expense, mainly pre-launch inventory provisions. SG&A productivity is the lever management named in July as the only one it had, and it worked. The caveat is that management guided Q4's margin below Q3's and named three reasons: year-end spend, rising royalties on growth brands, and lost Xolair profit-share income.
• EPS: Core EPS grew 9% in USD against core net income up 8%, so share count added about a point. That gap is narrower than last year, when the buyback added four to five points. Core net income grew slower than core operating income because interest expense rose to $468M from $281M on the Avidity-related debt. The IFRS line is almost entirely the September readouts: $2,231M of intangible impairments (del-desiran $1.94B, pelacarsen $0.3B) and a $0.6B onerous-contract provision tied to the del-desiran supply agreement. Core adjustments totalled $4.2B against $1.0B a year ago.
• Cash: The weak line. Operating cash flow fell 16% to $5,503M on unfavourable working capital and higher interest paid, and free cash flow fell 18% to $5,111M. The nine-month figure is $14,002M, down 12%. This is the fourth consecutive quarter of lower free cash flow year over year, and no analyst asked about it.

Brand and Franchise Performance

Novartis reports as a single Innovative Medicines business, so we decompose it by brand and region. The top 20 brands produced $12,052M in Q3, 83.0% of net sales, up 7% cc. The quarter's story is the growth portfolio replacing the eroding one, now with almost no US Entresto left to lose.

Regional Mix — Q3 2026

Assessment: The US is growing again. Over nine months US sales are down 4%, but Q3 grew 7% with US Entresto at $41M, against roughly $0.8B a year earlier. That means the US growth portfolio added well over $1B of year-over-year sales in one quarter. China at +18% cc is now a meaningful second engine, driven by the Leqvio and Kisqali NRDL listings. Europe at +3% cc is the region to watch from here, because Entresto's European regulatory protection expires next month.

Margin Bridge — Core Results, Q3 2026 vs. Q3 2025

Core gross profit includes other revenues ($507M against $449M), so its ratio to net sales overstates the product gross margin. Margin effects are rounded to one decimal and may not sum exactly. The company states the cc expansion as 2.0 points.

The Growth Portfolio

The six brands management designates as priority growth drivers (Kisqali, Kesimpta, Pluvicto, Scemblix, Leqvio and Fabhalta) produced $5,315M in Q3, or 36.6% of net sales. That compares with $5,037M (35.0%) in Q2, 30.3% in Q4 2025 and $3,930M (28.3%) a year ago. The cohort grew 35.2% in USD year over year and 5.5% sequentially. Management's own "priority brands" figure, a wider set that includes Cosentyx and Rhapsido, grew 29% cc.

Kisqali

Sales of $1,821M grew 37% cc, with the US at $1,168M (+38%) and $3B year to date. Management put US new-patient share up three points on the prior quarter, with 53% of patients drawn from the node-negative and N1 populations where Kisqali has no CDK4/6 competitor. Early breast cancer share reached 80% in Germany, 77% in the UK and 55% in China. The quarter also produced the six-year NATALEE analysis presented at ESMO, which showed a clinically meaningful overall survival benefit in the broadest at-risk population, a reduction in the risk of death of nearly 20% (hazard ratio 0.809).

"So taken together, we feel very comfortable with the peak sales guidance on Kisqali." — Vas Narasimhan, CEO

Assessment: Growth slowed from 43% cc in Q2 to 37% as the base gets larger, which is arithmetic rather than a warning. The new question is competitive. An oral SERD from a competitor is coming into the adjuvant setting, and management's defence rests on physician trust in endocrine therapy, the overall-survival data and combination studies. That is a reasonable defence and an untested one. The $10B peak was reaffirmed, and the hard exclusivity date in Q3 2031 still caps the period over which it can be earned.

Kesimpta

Sales of $1,484M grew 22% cc. Management put US share at 16% of the total MS market, up three points, and 27% of the B-cell class, up 4.5 points, with 77% of new prescriptions in first line or first switch. Ex-US share runs above 80% in many markets. The every-two-months dosing readout is on track for next year.

Assessment: The steadiest compounder in the portfolio. It now has an awkward neighbour, because remibrutinib in relapsing MS will compete for some of the same early-line oral or self-administered patients. Management frames the two as covering different segments. We expect some cannibalisation and do not think it changes either peak.

Scemblix

Sales of $593M grew 67% cc, with the US up 78% to $414M. US first-line share reached 33%, which management described as "really at market leadership now". Japan holds 56% first-line new-patient share and Germany 33%. Third-line share in key ex-US markets is 77%.

Assessment: In August management guided to US first-line leadership in the second half, and it delivered within one quarter. Growth decelerated from 89% cc to 67% on a larger base, as expected. This remains the cleanest launch in the portfolio.

Pluvicto and the radioligand platform

Sales of $674M grew 20% cc. The US grew only 9% to $502M, while ex-US grew 68% cc on Japan, Europe and early China uptake. The hormone-sensitive prostate cancer label (mHSPC) was approved in the US and China during the quarter, so Pluvicto is now approved across all stages of PSMA-positive metastatic prostate cancer. Management guided to slow uptake through Q4: about five to ten weeks of lead time per new patient, holiday scheduling, and acceleration "next year".

"Overall, I would say we are encouraged by the early uptake indicators." — Vas Narasimhan, CEO

Assessment: In August we called the mHSPC approval the largest revenue catalyst in the next twelve months that did not depend on a coin-flip readout. It arrived on time. But the US business it lands on has been flat for about four quarters, as one analyst pointed out, and management conceded the pattern is "a ramp initially and then you have a slower climb". The approval adds to 2027, not to this year, and the US base needs it more than we had thought.

Leqvio

Sales of $506M grew 63% cc, with the US up 60% and ex-US up 65% cc on China after its NRDL listing. The FDA added language to the US label recognising that lowering LDL-C reduces major cardiovascular events. The ORION-4 outcomes study is now expected to read out in Q4, and VICTORION-2P follows in 2027.

Assessment: After pelacarsen, Leqvio is the cardiovascular franchise. It has two outcomes readouts in twelve months and a mechanism where LDL lowering is already validated. Management cautioned that ORION-4 ran partly during the pandemic. Outcomes data in hand would support the $4B-plus peak, and we would not pay for an outcome beyond that before it reads.

Cosentyx

Sales of $1,857M grew 9% cc. The company stated that US growth "benefited from prior-year revenue deduction adjustments", and the Q3 2025 base carried roughly $180M of negative prior-period gross-to-net true-ups, mainly on Cosentyx and Entresto. Management put the underlying rate at mid-single-digit, as it did in Q2. Hidradenitis suppurativa share holds in the high 40s. Polymyalgia rheumatica received a positive CHMP opinion, and management expects the FDA decision in Q4.

Assessment: Management again volunteered the gap between reported and underlying growth, and we credit that. Cosentyx is a mid-single-digit compounder moving through new indications towards the end of its exclusivity, which is exactly what the $8B peak case requires and no more.

Rhapsido

Sales of $101M were up 58% on Q2's $64M. Management cited over 6,000 US prescribers, over 15,000 patients treated and 26% new-patient share, up about three points, at roughly 50% access to label. The FDA approved symptomatic dermographism in October, which covers about two thirds of chronic inducible urticaria patients. On the back of that approval, management raised combined CSU and CIndU peak sales to above $4B.

"So we are upgrading our peak sales guidance for Rhapsido in CSU and CIndU combined to over $4 billion." — Vas Narasimhan, CEO

Assessment: In August management said it would trade launch speed for net price, and the curve shows that trade. Revenue is ramping at half of label access, with a free-drug programme carrying patients until coverage arrives. The peak raise came from an approved indication, which matches the company's stated rule of giving peak sales only on positive Phase III data.

The renal portfolio

Fabhalta grew 60% cc to $237M and received traditional FDA approval in IgA nephropathy plus a first-in-class approval in Japan. Vanrafia sold $31M and completed submissions in Japan and China. Zigakibart's Phase III readout remains a 2027 event. Under questioning, management conceded that the competitive picture in IgA nephropathy has changed.

"So I think, clearly, the anti-APRILs have changed the landscape. I mean, I think the data has been clear and compelling and really demonstrated that they are the standard of care." — Vas Narasimhan, CEO

Assessment: In August we called the three-asset renal position the pillar that had advanced most quietly and most concretely. This quarter management called competing anti-APRIL antibodies the standard of care, said the positioning of Vanrafia and Fabhalta is "in a state of flux", and described a geographically tailored approach that leans on Japan and China. The $3B Fabhalta peak was reaffirmed across all indications. The US IgA nephropathy opportunity for the two launched assets is smaller than we thought, and zigakibart now carries the US case.

The Eroding Portfolio

Assessment: Entresto, Promacta and Tasigna together produced $1,448M in Q3, 10.0% of net sales, against 10.4% in Q2 and 19.2% across FY25. US Entresto is effectively gone, which is why management can promise faster growth in Q4 and into 2027. The two names we flagged in August behaved differently. Tafinlar plus Mekinist returned to 4% growth, and Xolair held at -25% cc. The Xolair problem has moved off the sales line: Novartis records no US Xolair sales and takes its US economics as profit-share income in other revenues. That income was still growing in Q3, at $391M against $355M. The CFO's guided Q4 margin step-down is the first sign of it reversing. Entresto's ex-US business, $1,124M a quarter and still growing, is the next erosion cohort, starting with Europe from next month.

Key Topics & Management Commentary

Overall Management Tone: Management was assured on the operating business and brief on everything that went wrong. The margin, the guidance raise and the priority brands came across as settled facts. Each September failure got a few sentences, and the details went to future congress presentations. In Q&A the posture moved from July's pre-emptive candour to deferral: at least six questions were answered with some version of "wait for the investor event" or "wait for February". On the commercial business this was the most confident call of the four quarters we have covered. On the pipeline it was the least forthcoming.

1. Two of the three headline readouts failed, in five days

On September 4, Novartis announced that the pelacarsen Lp(a)HORIZON outcomes trial did not reduce cardiovascular events against placebo in more than 8,000 patients, although it lowered Lp(a). On September 8 it announced that del-desiran's Phase III HARBOR trial in myotonic dystrophy type 1 missed its primary endpoint of video hand opening time, with "evidence of clinical activity" in secondary and exploratory analyses. Three days before the first failure, remibrutinib had met its primary endpoint in both relapsing MS trials. Of the three readouts we said in August had no pre-committed definition of success, one won and two lost outright, so the "clinically meaningful" debate never applied.

"While the del-desiran readout in DM1 was not what we had hoped, we saw evidence of clinical activity in the data and are working through the development path for this medicine." — Vas Narasimhan, CEO

The wider pipeline ledger for the quarter also leaned negative. Remibrutinib in myasthenia gravis was stopped for futility. The Alzheimer's and ALS programmes of VHB937 were discontinued. The autoimmune trials of the YTB323 CAR-T are on a global temporary pause while the company assembles a safety package, which management said would take "a few more months". The follow-on Lp(a) siRNA, DII235, is "under assessment following HORIZON". Against that: Rhapsido, Pluvicto and Fabhalta label expansions; pelabresib filed in the EU; priority review for del-zota and for the malaria combination KLU156; and an abelacimab interim analysis that let the study continue unchanged.

Assessment: This is the scenario our Bear 3 pillar was written for, and it is why we assigned no expected value to any of the three readouts. Our estimates do not move on the failures, and neither does management's 5% to 6% 2025 to 2030 sales CAGR, which it restated on September 8 and again on this call. What moves is the 2030s. Pelacarsen was the company's entry into a new cardiovascular class. Del-desiran was the lead asset of a $12B acquisition, and its first Phase III failed. The replacement story for the decade after Kisqali and Cosentyx lose exclusivity now rests more heavily on remibrutinib and on assets the company has not yet named.

2. The impairment arithmetic: $1.94B written off, $5.54B left, same probability of success

The interim report discloses the test in unusual detail. Before the test, the del-desiran in-process R&D asset carried $7.48B; the recoverable amount came in at $5.54B, so the charge was $1.94B, about 26%. The valuation used a 12% post-tax discount rate, unchanged from the purchase price allocation. A further $0.6B provision covers a long-term manufacturing supply agreement that became onerous once HARBOR changed expected capacity use. The company also disclosed the sensitivity: a 10-point change in probability of success moves the recoverable amount by about $1.0B, and a 10% change in forecast sales by about $0.7B.

"We did take a partial impairment charge of $1.9 billion. Primarily driven by the timeline shifting out by the additional clinical work that has to be completed." — Vas Narasimhan, CEO

Under questioning, management confirmed that the model assumes another Phase III study and a probability of success similar to the original one.

Assessment: The write-down reflects delay, not a lower chance of approval, and that is the assumption to watch. A failed Phase III followed by an unchanged probability of success for the next one is a generous position. The company's own sensitivity says that cutting it by 10 points would remove another $1.0B, and the notes warn that further impairment could reach the asset's full remaining carrying amount. None of this touches core earnings. It does mean that the asset behind most of this year's $15.4B of net debt growth is already carried about a quarter below its acquisition-date value, and the remaining $5.5B depends on a regulator conversation that has not happened yet.

3. Remibrutinib: the readout that worked, and an $8B-plus combined peak

REMODEL-1 and REMODEL-2 reduced the annualised relapse rate against teriflunomide by 35.2% and 31.2%. The pooled analysis showed a 29.2% reduction in six-month confirmed disability progression, nominally significant, and a 20.5% reduction in the three-month measure that did not reach significance. Adverse-event discontinuations were 3.1% against 5.8%, with no liver signal. The FDA submission is planned for Q4 with a priority review voucher. Management put the MS peak at above $4B, excluding non-relapsing secondary progressive MS, on top of the newly raised above-$4B urticaria peak.

"And as we all know in multiple sclerosis, with so many options, it starts with safety. And if you cannot demonstrate a clear, clean safety profile, understandably, neurologists will look for other options." — Vas Narasimhan, CEO

The ADR rose 6.0% on the September 1 topline and fell 3.3% on October 23, when the full data were presented and published. The full data showed relapse reductions against an oral comparator in the low-to-mid 30s and a disability signal that was only nominal.

Assessment: In August we said the commercial question was whether a relapse-rate win without a clear disability win would put the asset behind the B-cell antibodies rather than alongside them. The data answered that it will be positioned mainly on safety and oral convenience, and management's own segmentation says so: about 30% of naive patients, oral switchers, and patients who want to come off an anti-CD20 after years of infusions. That is a credible $4B franchise and a narrower one than a disability win would have opened. The asset's value is now spread across MS, urticaria and the HS readouts due before the November investor event.

4. Guidance raised for the first time this year

Full-year core operating income guidance moved up to "in line to grow low single-digit" from "decline low single-digit". Net sales guidance stayed at low-single-digit growth. The second-half framing moved up on both lines, and the CFO flagged the Q4 margin in advance.

"On the core margin, we do expect a decline in Q4 versus Q3. This reflects the typical year end phasing of our spend. Also, higher royalties as growth drivers scale up would play a part And finally, we will also see the impact of lower other income as a result of the Xolair biosimilars in the US." — Mukul Mehta, CFO

Assessment: This is the "guide to be beaten, and do not spend the beat" pattern we described in August, now with the beat spent. After a first half that beat its guide by about five points and a Q3 at +10%, the raise is earned and conservative. The bar it sets for Q4 is modest: holding the full year at "in line" needs roughly 7% cc core operating income growth in Q4, against Q3's 10%. That is why we put our FY26 estimate above the Street rather than at it.

5. Where 41.6% came from: SG&A did all of it

Core SG&A fell 6.8% to $3,078M, from 23.8% of net sales to 21.2%. On the margin bridge that is +2.6 points, more than the total 2.3-point expansion. Core gross profit added 0.9 points, the R&D ratio held, and core other expense, mainly pre-launch inventory provisions, cost 1.2 points. Management now describes third-party spend as "upwards of $11 billion", against the "upwards of $10 billion" it cited in July, and named agency, travel, consulting and professional services as the categories where technology and AI are cutting cost.

"If you look at SG&A for the quarter, we have actually seen SG&A 7% lower in Q3 versus previous year, while the company from a top line perspective has shown a growth of 4%." — Mukul Mehta, CFO

Assessment: In July the CFO said SG&A productivity was the only margin lever, and that we would read a miss on it as a genuine signal. This is the opposite signal: a single quarter that carried the whole margin. Two cautions keep us from annualising it. The CFO himself said that extrapolating one quarter "is not the right thing". And the 2029 target for a 40%-plus margin was not pulled forward, even after two quarters above 40%. Management is telling you the lever works and that the headwinds coming at it are large.

6. Q3 cleared the Entresto base we flagged in August

In August we wrote that roughly $800M of US Entresto in the Q3 2025 base was 5.8 points of headwind, and that a low-single-digit Q3 would be consistent with guidance. The quarter printed +4% cc with US Entresto at $41M, and the US region grew 7%. The guided acceleration in Q4 follows, because the Q4 2025 base holds almost no US Entresto.

"In the quarter, Novartis delivered sales growth and core margin expansion. And that supported an upgrade to our full year 2026 bottom line guidance." — Vas Narasimhan, CEO

Assessment: This is the replacement-power argument proven in a single quarter. The company lost about three quarters of a billion dollars of US sales from one brand and still grew the US 7%. A management team that called the year "a year of two halves" in January has delivered the turn on schedule. The question has moved from whether Novartis can grow through the US cliff to what the next cliff costs, and that is topic 7.

7. The other-revenue problem: Xolair biosimilars and Entresto in Europe

The CFO set out the main margin headwinds between now and 2029. Entresto loses European regulatory protection next month, and the company will defend secondary patents country by country. US Xolair biosimilars reduce the profit-share income that flows almost dollar for dollar into core operating income.

"...we would also have to deal with Entresto ex-US genericization as well as the US Xolair biosimilar competition coming and hence impacting other revenue, which is essentially pretty much a dollar on other revenue all comes down to the bottom line." — Mukul Mehta, CFO

He sized the other-revenue erosion at two to 2.5 percentage points of margin to absorb on the way to 2029. On Entresto in Europe, management expects no impact in Q4 and gave no country timing.

Assessment: This explains why a company running at 41.6% guides only to "40% by 2029 latest". Up to 2.5 points of margin from lost profit-share income is a large amount to absorb, and SG&A productivity is the stated offset. Profit-share income was still growing in Q3, so the erosion has not started in the reported numbers. When it does, a quarter like this one becomes hard to repeat. For 2027 the mix improves on the top line, with less loss of exclusivity, and worsens on the margin.

8. Ianalumab misses its approval window

In July management said the Sjögren's approval was on track for Q3, with no advisory committee and mid-cycle meetings complete. The quarter ended without an action. Management now describes ongoing label discussions with the FDA and expects an action by year-end. A follow-up question established that the discussion is about which patients benefit most, not about the endpoint.

"I think it is not about the endpoint. The endpoint results are robust. This is about how best to describe the benefits, which patients derive the greatest benefit." — Vas Narasimhan, CEO

Assessment: A label discussion about responder subgroups fits the heterogeneous response that management described in July, when it said there are "super responders and patients who respond less well". A narrower label is now the likely outcome. A competitor has also reported positive data in the same disease with a different mechanism. The asset's multi-blockbuster claim was always about the 2027 indications, and the Sjögren's launch looks smaller and later than it did in July.

9. Pluvicto: approval in hand, revenue in 2027

The mHSPC approval came on time, and management positioned it as a 2027 driver. US growth was 9%, first-line new-patient share in the pre-taxane setting is 15%, and management described the work of moving use earlier in the castration-resistant setting as slow behaviour change in community practices. The company also said it sees no impact on site behaviour from draft guidelines on patient release after radioligand therapy.

"...it is important to note we expect about 5 to 10 weeks of lead time for new patients to be treated, a combination of doing the testing required as well as scheduling." — Vas Narasimhan, CEO

Assessment: The peak-sales outlook was unchanged, and the shape changed. A ramp followed by a slow climb is the pattern management now describes for each new setting. Ex-US at +68% cc, led by Japan and China, is growing fast enough to carry the brand while the US works through behaviour change. We moved our expected mHSPC contribution from this year to 2027.

10. Kidney: "the anti-APRILs have changed the landscape"

Asked how the company sees its kidney franchise given the intensifying competition, management conceded more than we expected. It called the anti-APRIL class the standard of care in IgA nephropathy. It said positioning of its two launched renal assets is "in a state of flux" and will be tailored by geography. It pointed to next year's zigakibart readout, its own anti-APRIL, as the decision point.

"...but we are in a state of flux at the moment as we kind of think through how best to position Vanrafia and Fabhalta in the context of the anti-APRILs." — Vas Narasimhan, CEO

Assessment: This is a material change to a pillar we had treated as the quietest strength of the thesis. Fabhalta's $3B peak was reaffirmed, but it spans several indications, and the US IgA nephropathy setting is the one management now says belongs to someone else's mechanism. Zigakibart's 2027 readout becomes the asset that decides whether Novartis competes in the class or works around it.

11. Cash fell, net debt fell, and the buyback rose into the sell-off

Free cash flow of $5,111M fell 18% on lower operating cash flow, which the company attributes to working capital and higher interest paid. Net debt fell $2.0B in the quarter to $37,374M, the first decline this year, and trailing leverage eased to about 1.7x core operating income from about 1.9x at June 30. The buyback accelerated: the company repurchased 12.7M shares for about $1.9B in Q3, against 18.2M for $2.8B in the whole first half. Of that, $0.7B ran under the up-to-$10B programme, which has $4.9B left to execute by the end of 2027. Shares outstanding fell 12.1M in the quarter, against 7.5M in the first half.

"Absolutely right that as our share price took an inordinately significant reduction, we actually increased the pace." — Mukul Mehta, CFO

The interim report also discloses agreements with the US government that exclude qualifying Novartis imports from Section 232 pharmaceutical tariffs until January 2029, tied to planned US investment. The tariffs took effect for Novartis on July 31, and none have been assessed.

Assessment: Buying more stock after a 14% one-day drop is the right reflex, and it is the first sign this year that the buyback can still flex. The cash line is the problem. This is the fourth straight quarter of lower free cash flow, it has no analyst attention, and the company now carries $46.3B of gross debt with liquidity down to $9.0B. The tariff exclusion removes a tail risk we had not priced, and it was disclosed only in the notes.

12. Meet Novartis Management resets the medium term on November 18 to 19

Management will present updated medium-term guidance for 2026 to 2031, its long-term growth outlook and updated peak sales for key growth drivers in London. It also aims to name one to three mid-stage assets per therapeutic area for investors to start modelling. Most forward questions on this call were deferred to that event, or to the Q4 results on February 3, 2027.

"All of that taken together, we reaffirmed our 5% to 6% constant currency sales guidance for 25 to 30 and our mid single digit constant currency long term growth outlook. We have done all the modeling. We feel very comfortable with those numbers." — Vas Narasimhan, CEO

Assessment: The new guidance window will include 2031, Kisqali's loss-of-exclusivity year, and it is being set three months after two failed headline readouts. For a Hold, this is the most important event of the next twelve months. A 2026 to 2031 CAGR that holds near the current 5% to 6% with a credible bridge into the mid-2030s would change our view. A lower number, or one that leans on unnamed assets, would confirm it.

Guidance & Outlook

The guide combines the CFO's raised second-half framing with the CEO's 2027 posture: fewer generic headwinds, "dynamic growth" from the growth brands, and an Entresto Europe headwind to be handled country by country. Formal 2027 guidance comes with the Q4 results.

Implied Q4 math: the second-half sales guide of mid-to-high single digits, with Q3 at +4% cc on a $13,909M base, implies Q4 sales growth of roughly 6% to 12% cc on the $13,336M Q4 2025 base. The second-half core operating income guide of high single to low double digits, with Q3 at +10% cc, implies Q4 core operating income growth of roughly 4% to 14% cc on a $4,929M base. With nine months at -2% cc, a flat full year needs about 7% Q4 growth, and +2% needs about 16%. Applying that range, about $436M of Q4 core net financial expense (the $1.7B full-year guide less $1,264M for nine months), a 16.5% core tax rate and about 1,887M shares gives Q4 core EPS of roughly $2.08 to $2.32 and FY26 core EPS of $8.95 to $9.15, against $8.98 in FY25. The implied Q4 core margin of about 36.5% to 38.5% compares with 37.0% a year ago, consistent with the guided step-down from Q3.

Street at: the Street carried FY26 core EPS of roughly $8.80 to $8.83 entering the print, and FY27 at about $9.80. Adding the quarter's $0.21 beat to the pre-print Street gives about $9.01 to $9.04, close to our midpoint, so the raised guide adds little beyond the beat itself. FY27 at $9.80 implies about 8% growth from our FY26 midpoint, which needs the margin to hold against the other-revenue headwind.

Guidance style: unchanged and consistent. In 2025 the company guided low, raised twice and delivered at the top. In 2026 it guided a low-double-digit first-half profit decline and delivered -7%, reaffirmed rather than raised in July, and raised in October once the beat was banked. The raised guide should be read as a floor again. The real guidance event is the 2026 to 2031 update in three weeks.

Analyst Q&A Highlights

Whether Q4 is the template for 2027

The opening question asked management to weigh the pushes and pulls for 2027 and whether the implied Q4 growth rate is the best proxy. Management listed the drivers, fewer generic headwinds and the Entresto Europe exposure, and declined to endorse a proxy.

Q: "So just to wrap the question, is the implied Q4 sales growth rate the best proxy to think about 2027? Thank you."
— Sachin Jain, Bank of America

A: "I think overall, when you look at next year, as you say, we have less generic headwinds. We will be navigating the Entresto Europe headwinds, which will just navigate country by country. ... I mean, I think it is important to note in Q4, we are lapping a number of gross to net dynamics in the previous year. I think it would be best for us to provide, you know, more detailed guidance as always in January."
— Vas Narasimhan, CEO

Assessment: The answer does not endorse Q4 as the template and gives a reason: Q4 laps prior-year gross-to-net dynamics, which suggests its growth rate will flatter the run-rate. That is a useful caution from a management team that does not usually volunteer one. Anyone extrapolating a high-single-digit Q4 into 2027 should haircut it for Entresto Europe and for the base effect management just named.

What the del-desiran impairment assumes

The questioner noted that the impairment removed only about a quarter of the carrying value and asked whether the delay drove it and whether the probability and peak-sales assumptions were barely changed. Management confirmed that it expects another Phase III and models a similar probability of success.

Q: "Also on the impairment, noting it was only around a quarter. The total carrying value. ... So did mention the delay to launch. Was that the key reason for the write down?"
— Kerry Holford, Berenberg

A: "I mean, we still assume we will need to do another phase 3 study to ultimately bring the medicine to patients. In terms of the math for the write-off, I mean, when you have to do an additional study, a of course, the phase 3 program, you know, the timelines are the timelines. We assume a similar probability of success. ... So, I mean, I think this is. I would not read too much into these numbers."
— Vas Narasimhan, CEO

Assessment: The most revealing exchange of the call. "We assume a similar probability of success" after a Phase III miss is the assumption that holds $5.54B on the balance sheet, and the interim report's own sensitivity prices a 10-point cut at about $1.0B. "I would not read too much into these numbers" is the request; the notes, which warn of impairment up to the full carrying amount, argue the opposite.

Whether 40% comes before 2029

With two consecutive quarters above 40%, the question was whether the 2029 target could move forward, possibly to next year. The CFO held the date and explained why, naming the two other-revenue headwinds and sizing them.

Q: "So you now have 2 quarters in a row with margins above 40%. I think the guide has been a suggestion to return to above 40% by 2029. So can you maybe just talk to how we could see that progressing into next year? Could you see 40% earlier than 2029? Maybe even into next year?"
— Graham Parry, Citigroup

A: "So we remain confident on our commitment to get to 40% margin by 2029. And the main drivers of getting. For us to get back to that would be continued momentum on the top line, as well as continued productivity movements. ... I think during this time, we would also have to deal with Entresto ex-US genericization as well as the US Xolair biosimilar competition coming and hence impacting other revenue, which is essentially pretty much a dollar on other revenue all comes down to the bottom line."
— Mukul Mehta, CFO

Assessment: A clear no, with the mechanism stated. The CFO then sized the other-revenue erosion at two to 2.5 points, which sets the expectation for 2027. This answer is more important than the quarter's margin: a business printing 41.6% that will not commit to 40% for three years is telling you the current level is not the run-rate.

Whether the SG&A cut is structural

A two-part question first asked for 2027 mechanics, which the CFO deferred. The follow-up asked whether the SG&A reduction has structural parts, such as AI-driven productivity, or is cost containment during the loss-of-exclusivity phase.

Q: "The implicit question was that the SG&A we have seen, the reduction we have seen in third quarter are there any structural elements there that we need to factor in when we are thinking about the future, like AI driven productivity or, you know, a certain cost reductions which means for future, you will not have those costs."
— Rajesh Kumar, HSBC

A: "And if I look at Q3, taking 1 quarter and saying that is the estimate now for the next 3 years, I think is extrapolating it for the full year is not the right thing. But from a-- if we needed a data point to say how we get to that 40% margin by 2029 through SG&A, then I think here is a good data point that we have that we are absolutely on the right path."
— Mukul Mehta, CFO

Assessment: The CFO refused to let the quarter become the baseline and still claimed it as proof of the plan. Both halves are fair. Our read is that part of the cut is structural, because the named categories are ones where technology lowers demand, and part is timing, because Q4 spend is guided up. We model SG&A leverage continuing at a slower rate than Q3's.

Why ianalumab missed its approval window

The question recalled the July statement that the Sjögren's approval was on track for Q3, noted the FDA's real-time data requests, and asked for the reason and the new outlook. Management described label discussions and moved the expected action to year-end.

Q: "So we are now a third of the way through Q4 and no news. On the second quarter call, the company also referred to data FDA was requesting. So real time requests, at least as of that point. So curious on the reason for the delay, and what is the outlook? Thank you."
— Steve Scala, TD Cowen

A: "So we are in ongoing label discussions with the FDA and providing them the analyses that they are requesting to finalize their review as well as to support them on their approach to labeling. We do expect an action by the end of the year."
— Vas Narasimhan, CEO

Assessment: A missed date with no explanation volunteered. The later clarification, that the discussion concerns which patients benefit most, points to a narrower label. After pelacarsen's slip in July, this is the second dated commitment in two quarters to move right, and we would hold management to the year-end date.

Why the remibrutinib peak came in two tiers

The question asked why the urticaria peak was raised now while other peak updates wait for the investor event, and whether the split reflected disappointment with the MS data. Management gave the arithmetic and its rule for peak-sales guidance.

Q: "Vasant, you raised the Remi peak for CSU and CIndU to $4 billion-plus, but you said wait for the CMD for other peak increases. I am just curious as to why the 2 tier timing."
— Matthew Weston, UBS

A: "$4 billion for CSU and CIndU, $4 billion for the current MS readout, so not including non-relapsing SPMS, that indeed equals $8 billion-plus. ... As you know, our standard is usually not to include peak sales for drugs that have not actually read out positive in the phase 3 studies."
— Vas Narasimhan, CEO

Assessment: A clean answer that sets a floor: $8B-plus across indications that have already read out positive, with HS, food allergy, chronic pruritus and rosacea outside it. The rule is conservative and useful, because the next peak raise needs a positive Phase III first. The question behind the question, whether MS disappointed, got no direct answer, and the 3.3% fall on the full-data day suggests the market thinks it did.

Whether US Pluvicto has plateaued

The question noted roughly four quarters of flat US sales and asked whether the pre-taxane setting is fully penetrated, and how the hormone-sensitive launch would ramp. Management described slow behaviour change in the castration-resistant setting and a gradual mHSPC uptake.

Q: "But for the past 4 quarters or so in The U. S, sales have been flat. So is the PSMAfore indication sort of fully penetrated from a revenue perspective in The U. S?"
— James Quigley, Goldman Sachs

A: "So we do wanna see continued growth in mCRPC, and then on top of that, HSPC as well, we do expect there to be steady uptake. I do think that it is important to note that this is going to take time to develop. ... We think this is just the nature of you see a ramp initially and then you have a slower climb as you try to change behavior in the physician."
— Vas Narasimhan, CEO

Assessment: Management did not dispute the plateau, which is an implicit yes. "A ramp initially and then a slower climb" now describes both settings, so mHSPC should be modelled as a step plus a slow slope rather than a re-acceleration. The international business, Japan and China in particular, is where the brand's growth sits for the next year.

What follows pelacarsen in cardiovascular

The question asked what data would decide the fate of the follow-on Lp(a) asset and of the anti-inflammatory programme in the same space. Management said it does not need competitor data, that the full HORIZON dataset at the American Heart Association meeting will inform the decision, and that any further cardiovascular bet would be high risk.

Q: "Just on the cardiovascular pipeline... What data are you waiting for to make a decision? Are you waiting for competitor data on Lp(a)?"
— Thibault Boutherin, Morgan Stanley

A: "I think on Lp(a), we do not need to wait for necessarily competitor readouts. I think we have a lot of insights based on the study... I think given what we have seen, I mean, anything we do would have to be considered a high risk, high reward approach..."
— Vas Narasimhan, CEO

Assessment: A frank answer. Describing the next cardiovascular bet as high risk, high reward after a failed outcomes trial is honest, and it tells you the company's cardiovascular strategy beyond Leqvio is being redesigned. We value none of it until a Phase III design is announced.

What They're NOT Saying

1. Nothing on the reversal of Q2's one-time items. In July the CFO volunteered that phasing added about one point to Q2 sales and about five points to Q2 core operating income, and that it would reverse in the second half. This quarter's prepared remarks, release and Q&A do not mention it. We cannot tell whether Q3 absorbed it or whether part of it is in the Q4 margin step-down.
2. No discussion of free cash flow. Free cash flow fell 18% in the quarter and 12% for nine months, the fourth consecutive quarterly decline. The CFO described the cause in one sentence, and none of the 23 analyst questions raised it.
3. The onerous-contract provision appears only in the notes. The $0.6B charge on a del-desiran manufacturing supply agreement does not appear in the release or the call. Its cash cost runs over the remaining contract term, and the company says it may be adjusted.
4. The ex-US launch-pricing strategy under most-favoured-nation pricing, a third quarter without it. In February management promised to clarify it during 2026. In April it was described as "a 2027 story". This quarter the only reference was a remark that Kisqali ex-US is "not MFN constrained".
5. No 2027 framework at all. Questions on 2027 growth, gross margin, SG&A run-rate and peak sales were all deferred to the November event or to February, even though the company sized a 2027 FX effect and a 2029 margin headwind.
6. No detail on the YTB323 safety issue in prepared remarks. The CAR-T autoimmune programme across lupus nephritis, lupus, systemic sclerosis, myositis and vasculitis is on a global temporary pause. Management addressed it only when asked, without describing the issue, and gave a restart hope of next year.
7. No update on first-line ITP. In July the ianalumab first-line ITP readout was listed for the second half of 2026. It does not appear in this quarter's release, call or pipeline news, and the pipeline table still shows a 2027 submission.
8. No mention of the competing DM1 therapy. The interim report names "progress of a competing DM1 therapy" as a factor that could materially change del-desiran's recoverable amount. The call did not discuss it.
9. The Section 232 tariff exclusion was not mentioned on the call. An exclusion from US pharmaceutical tariffs until January 2029 is good news and a disclosed policy agreement. It appears only in the interim report's notes, without the size of the related US investment commitment.

Market Reaction

• Pre-print setup: NVS closed at $153.81 on October 26, up 11.6% year to date against the S&P 500's +12.1%, up 5.7% over the trailing 30 days and up 19.1% over the trailing twelve months. The 52-week closing range was $121.80 to $168.62, so the ADR entered the print 8.8% below its closing high.
• The September path: from $154.61 on the day of our August note, the ADR rose 6.0% to $161.25 on 4.3 million shares on the September 1 remibrutinib topline. On September 8, the first US session after the pelacarsen release and the day of the del-desiran release, it fell 13.9% to $137.70 on 10.3 million shares. It made its closing low of $137.16 on September 11, then recovered to $156.27 by October 22. On October 23, the day of the full MS data, it fell 3.3% to $151.12 on 4.4 million shares.
• Reaction session (before-open reporter, so the print day is the reaction day): premarket indications were up about 0.3%. The ADR opened at $154.15, up 0.2%, traded between $154.13 and $156.99, and closed at $156.86, up 2.0% or $3.05, near the session high. The S&P 500 rose 0.6%. The Swiss line traded up about 2%.
• Volume: 3.7 million shares against a 30-day average of 1.9 million, 1.9 times normal.
• Since the print: $156.08 on October 28, $161.10 on October 29 (+3.2%) and $160.79 on October 30. That is 17.2% above the September closing low and 4.6% below the 52-week closing high.

The failures were priced in September; the print priced the recovery. The del-desiran impairment drove a 61% fall in IFRS operating income and moved the stock not at all, because the market had marked both failures down by 13.9% in one session seven weeks earlier. What the print added was evidence that the base business can carry the company without them: a guidance raise, a 41.6% margin and growth through the Entresto cliff. The ADR closed near its high on almost twice normal volume and kept going two sessions later. That is a market adding to positions, not one reacting to a surprise.

The full-data sell-off is the reading that matters for the 2030s. The stock rose 6.0% on the remibrutinib topline and gave back 3.3% when the full MS data showed a nominal disability signal and relapse reductions in the low-to-mid 30s against an oral comparator. The market is valuing remibrutinib as a safety-and-convenience franchise, which matches management's own segmentation and is below the case that disability data could have opened.

Set against the August note, the ADR is 4.0% higher on a business that is clearly better operationally and carries less optionality. Our FY26 estimate rose about 1% at the midpoint, so the multiple did most of the work. The rally from $137 has been faster than any estimate change could justify on its own, and it leaves the stock within 5% of its closing high three weeks before a guidance reset.

Street Perspective

Debate: Is the 41.6% margin repeatable, or the high-water mark?

Bull view: Two consecutive quarters above 40%, SG&A down 7% while sales grew 4%, and a cost programme aimed at an $11B third-party spend pool where technology is lowering demand. The full-year guide went up, and the company has a record of guiding low and beating. A business that can cut SG&A this hard through a patent cliff can defend its margin through the next one.

Bear view: Management guided Q4's margin down, will not move the 2029 target forward, and named two to 2.5 points of other-revenue erosion still to come from Xolair biosimilars and Entresto in Europe. R&D has to keep rising to rebuild a pipeline that just lost two late-stage assets. One quarter of SG&A outperformance does not offset three structural headwinds.

Our take: The bears have the better of the arithmetic. The CFO's own answers, no earlier 40% and a sized other-revenue headwind, say that Q3 is near the high end of the near-term range. SG&A productivity is real, but it is mostly needed to stand still against the headwinds. We model a FY26 core margin of about 39.5%, with 2027 roughly flat to it.

Debate: Does the pipeline still bridge the 2030s after September?

Bull view: The most valuable of the three readouts, remibrutinib in MS, succeeded, and remibrutinib now carries an $8B-plus combined peak with HS, food allergy and other indications still to come. Kisqali has overall survival in the adjuvant setting, Rhapsido's peak was raised, and del-zota, pelabresib and abelacimab are advancing. Management restated the 5% to 6% CAGR the day after the second failure and says it has "done all the modeling".

Bear view: Pelacarsen and del-desiran failed, the Avidity asset has already lost a quarter of its value, the CAR-T autoimmune programme is paused, two neuroscience programmes and one remibrutinib indication were stopped, ianalumab slipped, and the IgA nephropathy position has been overtaken. The medium-term reset in three weeks will include 2031, the Kisqali loss-of-exclusivity year, with less late-stage support than it had in July.

Our take: Through 2030 the bulls are right, because the in-market brands carry the guidance and none of the failures were in it. After 2030 the bears are right, and the burden of proof now sits with the November event. We expect the 2026 to 2031 range to be set conservatively, and the substance to be in the mid-stage assets the company names. The CAGR number matters less than whether those assets are credible.

Debate: After a 17% rebound, what is left for the stock?

Bull view: At about 16.4x the Street's FY27 earnings, with sales growth improving as the loss-of-exclusivity cohort shrinks, a buyback that has re-accelerated, net debt falling and a tariff exclusion in hand, the stock is reasonably priced for a business returning to high-single-digit earnings growth.

Bear view: On 2026 earnings the stock trades near 18x, above its five-year average multiple. It has recovered almost all of its September losses even though two pillars of the long-term story failed, free cash flow is falling, and a guidance reset that could disappoint is three weeks away. The rebound has priced the beat and left nothing for the pipeline.

Our take: Neither side has the edge until November 18. The bull case depends on FY27 at $9.80 and a margin that holds. The bear case depends on the reset underwhelming. At this price we are paid to wait for the reset, not to anticipate it.

Model Update Needed

Valuation: at the $156.86 reaction-day close, NVS trades at 17.1x to 17.5x our FY26 core EPS range and 17.5x FY25's $8.98. At the $160.79 close of October 30 the range is 17.6x to 18.0x, and about 16.4x the Street's FY27 estimate of $9.80. Market capitalisation at that close is about $304B on 1,888.6M shares outstanding. Against our FY26 free cash flow estimate of $15.5B to $16.0B, the free cash flow yield is 5.1% to 5.3%, down from 5.5% to 5.6% in August.

Valuation impact: our estimate rose about 1% at the midpoint and the multiple rose about 0.6 of a turn. In August we said fair value was close to the price at roughly 17x and that we wanted a de-risked cardiovascular readout or a mid-teens multiple before paying up. The cardiovascular readout de-risked the wrong way. The mid-teens multiple appeared from September 8 to mid-October, at 15.0x to 15.3x our current range at the $137 low, and has gone. On a business that is operationally stronger but carries less optionality after 2030, we see fair value slightly below the current price and would revisit on the November guidance or near $140 to $145.

Thesis Scorecard Post-Earnings

The pillars below are the standing thesis set at the October 2025 initiation and carried in our thesis of record. They are graded against this quarter, not re-derived.

Grading last quarter's commitments

Overall: the operating thesis strengthened again and the optionality thesis weakened sharply. Of thirteen commitments from July, eight were delivered or beaten, one was partial and four were missed. Every miss is in the pipeline or policy column, and every delivery is in the commercial and financial column. That split is the thesis in one table. The base business is executing better than we underwrote. The pipeline that was supposed to extend it past 2030 has lost two of its three headline assets and a quarter of its largest acquisition's value. Net of both, the thesis is unchanged in rating and narrower in duration.

Action: hold. Of the two upgrade triggers left from August, the cardiovascular readout resolved negatively, and the mid-teens multiple came and went in September. The downgrade watch items moved in opposite directions. Free cash flow conversion declined again, which we flagged as a downgrade risk. Net debt fell rather than rising, which removes the other. Our new upgrade trigger is the November 18 to 19 investor event: 2026 to 2031 guidance near the current 5% to 6% with credible named assets for the mid-2030s, or a pullback to around $140 to $145, about 15.5x to 16x our FY26 range. Our new downgrade trigger is a fifth consecutive quarterly decline in free cash flow with Q4 results, against a weak $1,655M Q4 2025 base, or 2027 guidance implying margin regression larger than the 2 to 2.5 points the CFO has already sized. For an investor with a position, this quarter is a reason to keep it. For one without, the guidance reset in three weeks is the better entry decision than this price.