Johnson & Johnson

IThe business

Johnson & Johnson makes prescription medicines and medical devices. It has two segments. Innovative Medicine made $60.4B of the $94.2B of 2025 sales, 64%, and MedTech made $33.8B, 36%.³ About 57% of 2025 sales were in the United States.³ It employs about 138,200 people, and its release datelines say New Brunswick, New Jersey.³² Until 2023 it also sold consumer health products. That business became Kenvue, which went public in May 2023, was split off to holders by an exchange offer in August 2023, and is no longer owned.³

The model is the same for drugs and for devices: spend heavily to find and prove a product, protect it, and keep selling it. The nine-month 2026 gross margin was 67.6%, after $3.7B of amortization of purchased intangibles that sits in cost of goods.⁴ Selling and administration took 25.1% of sales and research 15.8%.⁴ Products reach patients through wholesalers, distributors, hospitals and retailers, and the people who really decide are insurers, government programs and hospital buying groups.³ I read that as: the customer who pays is rarely the one who chooses, which is why evidence and contracts matter.

Stelara, and what replaced it

Stelara, an injected drug for psoriasis and bowel disease, made $10.9B in 2023 and $10.4B in 2024. In 2025 it made $6.1B, down 41%, and in the first nine months of 2026 $2.1B, down 56%.³⁴ The third quarter alone fell 54%, and Stelara took about 6.9 points off the growth of Innovative Medicine.²⁴ The 10-K says biosimilar versions are launching around the world and will keep cutting sales.³ On my estimate it ends 2026 near $2.7B, a fall of about $8.2B since 2023 E (estimate).

The loss did not show up in total sales, because three things grew faster. Darzalex, a blood cancer drug, went from $9.7B in 2023 to $14.4B in 2025, which was about 15% of all sales.³ Tremfya, a psoriasis and bowel disease drug that competes with Stelara in the same clinics, went from $3.1B to $5.2B, and in nine months of 2026 it made $6.1B, up 70%.³⁴ Cancer drugs outside Darzalex grew too. Total sales were $85.2B in 2023 and are guided to $101.2B for 2026.³²

What it sells

This is the nine-month 2026 list for Innovative Medicine, in order of size within each area. Oncology is 45% of the segment, and Darzalex alone is more than half of oncology.

Cancer, immunology and the mind

Oncology made $22.0B in nine months, up 18.7%, or 16.8% before currency.⁴ The 10-Q credits Darzalex with share gains and market growth, and names Carvykti, Tecvayli, Talvey, Rybrevant with Lazcluze, Erleada and the Inlexzo intravesical system as contributors; only Imbruvica fell.⁵ Two pieces of data in the third-quarter release point forward: a single dose of Carvykti gave five-year treatment-free remissions in 50% of patients in an early-line myeloma study, and a Tecvayli and Talvey combination cut the risk of progression or death by 89% in earlier-line myeloma.² Both are company-reported trial results and I can’t check them.

Immunology made $11.4B in nine months, down 3.6%, because Tremfya’s gain of $2.5B was smaller than the $2.7B lost on Stelara and the further falls in Simponi and Remicade.⁴ The next drugs are small so far: Imaavy, which won US approval in the quarter for a blood disorder called warm autoimmune hemolytic anemia, and Icotyde (icotrokinra), an oral psoriasis drug, together lifted the “other immunology” line to $297M from $21M.⁴²⁵ J&J released two-year Icotyde data in the quarter.² Whether it becomes a second Tremfya is one of the things I can’t judge.

Neuroscience made $6.9B, up 20.3%.⁴ Spravato grew 40.5% to $1.7B and Caplyta, which came with the $14.5B purchase of Intra-Cellular on 2 April 2025, made $1.0B.⁴⁵ Both are in patent suits. A court found Spravato’s patents infringed and not invalid in August 2026, and Sandoz appealed; Caplyta’s suit with Zydus was settled in August 2026 on terms not disclosed.⁵

The second wave of losses

Stelara was the first. The 10-Q names others. Opsumit, a pulmonary hypertension pill, met US generics late in the second quarter of 2026, and its third-quarter sales fell 36%.⁵⁴ Uptravi, the other main drug in that area, could see US and European generics in 2027.⁵ Simponi biosimilars reached Europe in the second quarter of 2026, with a possible US entrant later this year, and Imbruvica has been under a Medicare price since 2026.⁵³ None of these is as large as Stelara. Together they made about $9.7B in 2025 C (calculated).³

MedTech

MedTech grew 6.1% in 2025 to $33.8B and 4.9% in nine months of 2026, or 3.5% before currency.³⁴ It has four parts, and they are different businesses.

Cardiovascular is the strong part. Electrophysiology, the heart-rhythm business, made $4.5B in nine months and Shockwave, which treats calcified arteries, grew 16.6%.⁴ The 10-Q says electrophysiology faces competitive pressure in pulsed field ablation catheters, and that Abiomed’s third-quarter sales fell on lower patient volumes.⁵ Surgery is the weak part. Its operational sales fell 1.0% in the third quarter because of China volume-based procurement, a restructuring of the franchise, disruption in the Middle East and competition in staplers and energy devices.⁵ In the quarter the company also received US marketing authorization for Ottava, which it calls the first table-integrated soft-tissue surgical robot.² It has no sales yet that I could find. I compare it with Intuitive’s da Vinci in Section III.

Orthopaedics, and the split

In October 2025 J&J said it intends to separate Orthopaedics, and the 10-Q expects that to finish in mid-2027; the proxy calls it the planned separation of DePuy Synthes.⁵¹⁰ It made $9.3B in 2025, about 10% of sales, with growth of 1.1% that year and 4.8% in nine months of 2026.³⁴ The 10-Q does not give the method. Separation costs were $0.7B in nine months of 2026, and they sit outside adjusted earnings.⁴ For an owner, a spin-off hands out shares in a second company, and the value stays with holders. What changes is the mix and the stranded costs, and I can’t judge those.

Where the margin goes

GAAP pre-tax earnings were 17.1% of sales in nine months of 2026, against 39.7% a year earlier, because the 2026 figure carries a $5.5B talc charge and the 2025 figure a $7.0B talc reversal.⁴⁵ The adjusted figure strips those out. Adjusted pre-tax earnings were 32.3% of sales, against 35.4%.⁴ About 1.3 points of the fall is a single $1.0B charge for the Firefly Bio purchase, and most of the rest is mix as Stelara shrinks, plus higher research and selling costs.⁵⁴ For the full year 2025 the adjusted margin was 33.6%, against 32.6% in 2024.⁸ Section V takes the gap between adjusted and cash apart.

Debt is modest for the size. On 27 September 2026 J&J had $44.6B of debt and $17.2B of cash and securities, a net debt of $27.4B, about the same as the $27.2B a year earlier.⁵ Loans and notes payable due within a year were $8.9B, which includes $3.2B of bonds.⁵ The company also holds a $12.5B 364-day credit line, and says it can issue up to $20B of commercial paper.⁵

IIThe moat

What protects Johnson & Johnson’s profit? It is two businesses with different answers, and in the drug business the answer is drug by drug. Each protection has a date, and the dates are what matter to an owner.

What protects it, source by source

Patents. The 10-K gives dates for only two drugs. Darzalex’s two Genmab patent families end in the United States in 2029, with compound or use protection in Europe in select countries to 2031 and 2032, and J&J owns separate patent portfolios for the under-skin form and the intravenous form.³ The company pays Genmab 12% to 20% of Darzalex sales, which was about $2.4B in 2025.³ Tremfya’s composition patent family ends in the United States in 2031.³ For every other drug I have only the litigation notes. That is the first thing I can’t date: when a biosimilar of Darzalex can really be sold. The patent date is not the market date, and Stelara is the proof. I give the Darzalex range in Section VI.

Evidence. The second protection is evidence that doctors act on. The release reports data in Tremfya for spinal pain in psoriatic arthritis, in Carvykti for long remissions and in the Tecvayli and Talvey combination.² Evidence like that is why a plan or a hospital keeps a drug on its list even when a cheaper copy exists. I treat it as real but thin beside a patent, because I can’t measure it from the filings.

Habit and training. In devices the protection is the surgeon. A surgeon trained on a stapler, a catheter or a knee system is slow to change, and the hospital buys the service and the instruments for years. This is the weaker part of the group today: the 10-Q names competition in pulsed field ablation, staplers and energy devices.⁵

Evidence it is widening

Darzalex grew 17.0% before currency on a base of $10.4B, Tremfya 67.9%, Carvykti 45.5% and Spravato 39.7%.⁴ Gross margin is 67.6% in a year when a high-margin drug lost $2.7B of sales, and the adjusted pre-tax margin was still 32.3% with a $1.0B research purchase inside it.⁴ I read that as the portfolio replacing profit as well as sales.

Evidence it is narrowing

Three things on the page show the edges. First, Stelara. The same company, the same sales force and the same doctors are on course to see about 75% of a $10.9B drug go by the end of 2026, on my fourth-quarter estimate. Second, government prices. Imbruvica has carried a Medicare price since 2026. Erleada, which made $3.6B in 2025, is on the list with prices from January 2028, and CMS has until 30 November 2026 to publish any agreed prices.¹³³ The 10-K says CMS will remove Xarelto and Stelara from the list from 2027.³ Third, the group has chosen to settle with the government on prices. In January 2026 J&J signed a voluntary agreement that includes patient affordability initiatives, participation in pricing programs and continued US investment, and a press report says it carries an exemption from certain tariffs.⁵¹⁴ I don’t have the terms, so I can’t say what it costs.

Verdict. The protection is wide and has dated holes. Wide, because no drug is more than 15% of sales and the newest ones are growing 40% to 70%. Dated, because Darzalex, 17% of my 2026 sales estimate, reaches 2029 on its patents, and Medicare and biosimilars can reach it earlier or later than that.

IIIThe competitive advantage

Is J&J better than its peers at turning science into cash? Strong protection is one thing, and a company that uses it well is another. I compare it with the four other healthcare companies I have written up: Eli Lilly, AbbVie, Novo Nordisk and Intuitive Surgical, using the figures from those pages.

Where J&J wins

It wins on spread. Lilly earns 64% of sales from one molecule, Novo 74% and AbbVie 32%; J&J’s largest drug is 15% and the second, Stelara, was 6.5% in 2025.⁹¹⁶¹⁵³ It has also been through a cliff on a large product already and grew through it, as AbbVie did with Humira. It wins on balance sheet: net debt of $27.4B is about 0.8 times my estimate of 2026 earnings before interest, tax and depreciation, $36B, and about 1.0 times if the talc reserve and opioid payments are added C (calculated).⁵ S&P affirmed its AAA rating in April 2025 after the Intra-Cellular deal; I found no later action and could not confirm Moody’s grade, which the 10-K gives only as investment grade.⁷³ AbbVie, by comparison, carries $70B of net debt on its page.¹⁵

And it wins on the dividend record: 64 straight yearly increases, the latest on 14 April 2026 from $1.30 to $1.34 a quarter, $5.36 a year.⁶

Where it loses

It loses on growth. Lilly is growing six times as fast on sales, and Intuitive’s procedures are growing at about 14% against MedTech’s 3.5% before currency.⁹¹⁷⁴ It loses on margin: AbbVie’s adjusted operating margin was 48.7% in its latest quarter and J&J’s adjusted pre-tax margin is about 32%, because J&J carries a device business, which usually earns less than a protected drug.¹⁵⁴ And it loses on the number I can’t change: the price. At $276.32 the stock sits 204% above my base, more than any of the other four pages, even though J&J is the least concentrated and has the strongest balance sheet.¹

J&J’s robot against Intuitive’s

Ottava is the one head-to-head. J&J says it is the first table-integrated soft-tissue robot and it received US marketing authorization in the third quarter.² The Intuitive page covers the same event from news coverage and treats it as a pressure on da Vinci in simple benign surgery, not in cancer work.¹⁷ I would not size it. Ethicon, J&J’s surgery arm, sells the staplers and energy devices used in robotic cases, and Surgery is growing 0.8% before currency, so a robot is a reasonable attempt to defend a franchise that is under pressure and not a new engine.⁴

Who can compete with J&J? The 10-K says the group faces substantial competition in all its markets and does not name rivals.³ I have no filing-based view of the other myeloma, psoriasis or ablation competitors, and I can’t put a number on their effect. That is part of why the bear case is as low as it is.

IVManagement

How he is really paid

Joaquin Duato is chairman and chief executive.¹⁰ His total pay for 2025 in the summary table was $32.8M, up from $24.3M in 2024 and $28.4M in 2023. The parts were a $1.6M salary, $15.0M of stock awards, $5.9M of options, a $4.3M cash bonus, a $5.3M rise in the value of his pension and $0.6M of other pay.¹⁰ The company says the median employee made $91,000 and the pay ratio was 360 to 1.¹⁰ The committee’s own measure, which leaves out pension changes, was $27.1M.¹⁰

What does that pay him to do? The annual bonus has two parts. Seventy percent is financial, on operational sales, adjusted operational EPS and free cash flow, in equal parts, and the 2025 result paid 122.7% of target. Thirty percent is strategic goals set by the board committee. The executives’ bonuses overall paid 118.3% of target.¹⁰ The long-term award is 60% performance shares, half on three-year cumulative adjusted EPS and half on shareholder return against a peer group, 30% options and 10% share units. The 2023 to 2025 performance shares paid 113.6% of target.¹⁰

Two points matter here. First, the measures are adjusted. Adjusted EPS leaves out the amortization of intangibles and items such as litigation, so buying a company costs the pay measure its interest and little else, while buying early research is charged, as the $1.0B Firefly deal was in 2026.⁴ The free cash flow measure is operating cash less capital spending, as the company defines it, so it is struck before any acquisition.²¹⁰ That is neutral by design. It does not discourage deals, which are the largest use of cash after the dividend. Second, the committee and the audit committee reviewed every item left out of the measures for 2025, which is the right check.¹⁰

Who runs it

Duato is also chairman, so one person leads both the company and the board. Marillyn Hewson is lead independent director, and the proxy says all nominees are independent except the chairman and chief executive.¹⁰ At the April 2026 meeting a proposal to require an independent chair got 23.4% of the votes cast for or against, and the pay vote got 94.7% C (calculated).¹⁸

Their own money

Duato owned 356,297 shares in the proxy table, about $98M at $276.32, plus 1,051,370 options and share units he can exercise or vest within 60 days.¹⁰¹ The rule is twelve times salary for the chief executive, which is $19.2M, so he is far above it.¹⁰ The 21 directors and executive officers together held 890,696 shares, about 0.04% of the company C (calculated).¹⁰⁵ The largest holders are Vanguard at 10.1%, BlackRock at 7.7% and State Street at 5.5%, on filings from February 2026 for Vanguard and January 2024 for the other two.¹⁰ I did not check the Form 4 record for purchases or sales.

Capital over a full cycle

The test of an allocator is what the cash did. From 2016 to 2025 J&J’s operating cash flow was $225B. It paid $106B in dividends, 47% of that cash, spent $107B on deals, 47%, bought back $54B of stock and spent $38.5B on plants and equipment C (calculated).¹¹³ The sum is $305B, about $80B more than the cash the business made. The difference came from new debt and from the Kenvue separation, which brought $4.2B from its public offering and $8.0B of borrowing that was passed to Kenvue in 2023.³ Part of the cash I count as deals before 2023 may include milestone payments that were booked elsewhere in investing, so the ten-year total is a floor.

The dividend was the steady part. Cash dividends paid per share rose from $3.15 in 2016 to $5.14 in 2025, 5.6% a year, and the board raised the quarterly dividend to $1.34 on 14 April 2026, the 64th straight year of increases.¹¹⁶ I checked the count against the company’s release. The share count fell, from 2,789M diluted in 2016 to 2,429M in 2025. Most of the fall came between 2022 and 2024, when it dropped by 235M shares. Buybacks of $7.5B in those two years, at an average close of about $158, would retire only about 47M, so I read the rest as the Kenvue exchange offer of August 2023, which I did not trace share by share E (estimate).¹¹³¹ The buybacks are another story: $5.96B in the first nine months of 2026 bought 6.5M shares in the third quarter at an average of $261.16, which the 10-Q calls a systematic plan to meet the needs of compensation programs.⁵ Over the same nine months option exercises brought in a net $2.3B, and the diluted share count in the third quarter was 2,448.6M, up 0.8% on the year before.⁵⁴ So the buybacks are best read as part of stock pay, and in my model I charge stock pay as an expense and do not deduct buybacks as well E (estimate).

The deal record

Deals are where an owner needs to look hardest, because they decide the next ten years of earnings. Here is the record, with the outcome I can verify.

The pattern is mixed. Shockwave and Cardiovascular look good so far, and Intra-Cellular is on a good path with a short record. Actelion’s drugs are now losing exclusivity, and Abiomed’s growth has slowed since the price was paid. The deals total $107B in ten years, an average of $10.7B a year and $11.8B a year in the last three, so a company that makes $19B of owner cash a year has been spending most of it, plus the dividend, on keeping the pipeline full. The record says buying pipeline is not safe. It is why I charge a deal budget to every year of the valuation.

VFinancials

Reported earnings and cash differ here more than at most large companies. I read three versions: the GAAP figure, the company’s adjusted figure, and what is left for an owner. Everything below is fiscal years of 52 or 53 weeks; 2025 ended on 28 December. Sales before 2023 include the consumer business, which became Kenvue, so I do not quote a ten-year growth rate as like for like: the raw figure is 3.0% a year, but for 2021 and 2022 the company later restated sales to $78.7B and $80.0B without Kenvue.¹¹³

Margins

GAAP pre-tax earnings are noisy because talc swings them: a $5.5B charge in the third quarter of 2026, a $7.0B reversal in the first quarter of 2025 and about $5.1B of charges in 2024.⁵³ The adjusted pre-tax margin was 32.6% in 2024, 33.6% in 2025 and 32.3% in nine months of 2026, and the company guides adjusted EPS of $11.10 for 2026, up 2.9%.⁸⁴² The gap between adjusted and GAAP is mostly the amortization of purchased intangibles, which was $3.7B in nine months, plus litigation, restructuring and separation costs.⁴ The amortization is real money spent on past deals even though it is not cash now.

What an owner can really take out

Start from operating cash flow, take off capital spending and stock pay, and you have what the business made for its owners before deals. Then take off cash paid for acquisitions and purchased research. The table puts three periods side by side.

In the twelve months to September an owner could take out $19.2B after stock pay, which is 74% of adjusted earnings of $26.0B and 2.8% of the market value C (calculated). In 2025 the same ratio was 70% and in 2024 about 77% C (calculated).³⁸ After the $4.4B of deals in those twelve months, $14.7B was left, more than the $12.7B of dividends but well below dividends plus $7.9B of buybacks, which came to $20.6B. The cash balance covered the difference: it fell from $19.7B at the end of 2025 to $16.9B.⁵ The Auris judgment is another $0.8B that went out in January and is not in the table, so the real owner cash was a little lower still C (calculated).⁵

That is the central fact of this page. Earnings of $26B look like $19B of cash, and $14.7B once deals are counted. The gap from adjusted earnings to cash is about $7B a year. I can’t split it line by line from the filings. Part is working capital, which grew by $1.5B over nine months as receivables and inventories rose faster than payables.⁵ Part is cash taxes and the timing of deferred taxes, part is payments on restructuring and litigation, and part is the timing of other accruals. I charge the recurring part of it in the model as “other cash costs” at 3.0% of sales in 2027, falling to 2.0%, and I charge the talc reserve separately as a claim E (estimate).

What it takes to stand still

Standing still means keeping sales where they are while patents end. The business pays for part of that inside the margin: research was $14.7B in 2025, 15.6% of sales, and 15.8% in nine months of 2026.³⁴ The rest it buys, as acquisitions and purchased research. That cost is not in the margin or in operating cash, and so it is easy to forget.

How much is enough? I can’t know, and the record is wide. Deals averaged $10.7B a year over ten years, $7.2B without Actelion, and $11.8B a year from 2023 to 2025 with Shockwave and Intra-Cellular C (calculated).¹¹³ I charge 5.5% of sales every year in the base, which is $5.9B in 2027, and 7.0% in the bear and 4.5% in the bull E (estimate). That is lower than the last three years, because I assume the large purchases are rarer than the small ones. Each extra point of sales, about $1.1B a year, moves the base value by roughly $6 a share.

The talc reserve, on the other hand, is not a stand-still cost. It is a debt for the past, and the 10-Q puts it at about $9.0B in present value, of which about half is a current liability.⁵ I treat it as a claim ahead of the shares and not as a yearly cost.

The balance sheet

On 27 September J&J had $44.6B of debt and $17.2B of cash and securities, so net debt of $27.4B C (calculated).⁵ Beside it are two other claims: the talc reserve of about $9.0B and about $0.9B still to pay on opioid settlements.⁵ Together that is $37.3B ahead of the shares, $15 a share C (calculated). A further $2.6B for the option to buy Sail is not a liability until J&J uses it.⁵

The AAA label deserves a plain check. S&P affirmed its AAA rating on 25 April 2025 with a stable outlook, and said it could lower it if adjusted leverage were projected above 1.0 times on a sustained basis.⁷ The 10-K says only that J&J keeps investment-grade ratings from Moody’s and S&P, and I could not confirm Moody’s grade.³ My own arithmetic is simpler: net debt is about 0.8 times earnings before interest, tax and depreciation, and 1.0 times with talc and opioids included, on my estimate of $36B for 2026 E (estimate). That is low, and it is the reason a $14.5B purchase did not cost the rating. It is also the reason I do not put a premium on the balance sheet beyond the debt figure: a company this low on debt can still lose value if the cash flows fall.

Goodwill and purchased intangibles are $95.1B of $197.4B of assets, so another large write-down of an acquired product is possible.⁵ The 10-Q shows a $0.4B write-down in 2026 on a research asset from Abiomed.⁵ Equity is $81.5B.⁵ Net interest cost was only $136M in nine months.⁴

Tax

The adjusted tax rate was 17.3% in 2025, 16.3% in 2024 and 17.1% in nine months of 2026.⁸⁴ The GAAP rate was 14.0% in nine months of 2026 and 21.5% a year before, because talc and other special items are taxed unevenly.⁴ I use 17.5% on adjusted operating income and I flag the range of 14% to 21% in Section VI E (estimate).

The shares

The price has risen 46% since 31 October 2025, when it closed at $188.87, and 15% since 14 April 2026, when it closed at $240.10 C (calculated).¹ The close of $276.32 is within 1% of the highest close in the ten-year record, $278.43 on 3 September 2026.¹ At that level the dividend yield is 1.94% on the $5.36 annual rate C (calculated).⁶ The price is 24.8 times the 2026 consensus EPS of $11.12 and 23.9 times the 2027 consensus of $11.55, which is the figure the Eli Lilly page uses for its Johnson & Johnson row at $273.20.¹²⁹

Johnson & Johnson, ten years

VIValuation

First, how this goes badly wrong

Before the numbers, I asked what would most likely ruin the business or the investment, and worked back. Three things could, and none needs a surprise.

• The big drugs end sooner, or cheaper, than their patents say. Darzalex is about 17% of my 2026 sales estimate and its US patents end in 2029.³ Stelara is on course to lose about 75% of its 2023 sales by the end of 2026, two years after copies began to arrive.³⁴ If Darzalex did the same after 2029, and Tremfya after 2031, the newer launches would need to be very large to fill the hole. This is my bear case, worth $39.
• Money goes out for the past and into deals that do not pay. J&J has $9.0B of talc reserve and $0.9B of opioid payments to make, $27.4B of net debt, and a record of spending $11B a year on deals.⁵¹¹ The ovarian talc settlement of $5.5B only becomes final if at least 95% of claimants join, and the mesothelioma cases are outside it.⁵ A run of poor deals plus a bigger talc bill is how a strong balance sheet becomes an ordinary one.
• The price already pays for the good outcome. At $276.32 the stock is 23.9 times the 2027 consensus, and a stock priced like that can fall a long way with no change in the business if its growth only meets expectations.¹² The first two risks sit inside my bear case. This third one is what the rest of this section measures.

What the price implies

At $276.32 Johnson & Johnson is worth $677B on 2,449M diluted shares, the 2,407M outstanding on 27 September plus 41M of dilution C (calculated).¹⁵⁴ Add $27.4B of net debt, the $9.0B talc reserve and $0.9B of opioid payments and the whole business is priced at $714B, or 7.1 times 2026 sales C (calculated).⁵ The price is 24.9 times the 2026 adjusted EPS guide of $11.10, 23.9 times the Nasdaq consensus of $11.55 for 2027 and 20.0 times $13.84 for 2028 C (calculated).²¹² The consensus for 2029 is $14.59, so the Street itself sees growth speeding up from 4% in 2027 to 20% in 2028 and then slowing.¹² I can’t explain the 2028 jump from the filings. For comparison, the Eli Lilly page shows J&J at 23.7 times 2027, AbbVie at 18.4 and Merck at 16.1 on its own closing prices.⁹

The owner yield is lower than the headline. On the twelve months to September, owner cash after stock pay was $19.2B, or 2.8% of the market value. Take off a stand-still deal budget of about $5.5B and it is $13.7B, or 2.0% E (estimate). The ten-year Treasury yields 4.90%.¹⁹ In my 2027 base, owner cash before interest is $16.8B, a yield of 2.5% E (estimate). The choice that moves my base most is charging a stand-still deal budget of 5.5% of sales every year; with no deals the base would be about $125 E (estimate), but J&J spent $107B on deals in ten years, so I think leaving them out would flatter the cash.

What does the price pay for? Holding the margin flat at 33.0%, close to what the company earned before the Firefly charge, the price needs sales to grow 15.6% a year for ten years C (calculated). That takes sales from $101B to $430B by 2036, more than four times. Sales grew 7.4% in nine months of 2026 and 5.6% a year from 2022 to 2025 on the restated figures C (calculated).⁴³ The company has said it has “clearer line of sight to double-digit growth by the end of the decade.”² Nothing in the filings points to a decade of it, with Darzalex’s patents ending in the middle of the period.

What I can’t judge

Six inputs drive this value, and none of them can be known today. I show them as ranges for that reason, and I do not trust one number.

When Darzalex copies arrive, and at what price. The two Genmab families end in the United States in 2029 and J&J has separate patents on the under-skin and intravenous forms.³ I found no settlement or ruling that dates a biosimilar. Medicare could also reach it. A biologic can be selected eleven years after licensure, and Darzalex was licensed on 16 November 2015, so the list due in early 2027 is the first it could appear on E (estimate).¹³²⁰ But the FDA record flags that approval as an orphan one, a 2025 law widened the exemption for rare-disease drugs, and I can’t tell if Darzalex qualifies.¹³²⁰ Moving the copy date two years either way moves the base by only about $2 a share, because the cash arrives late in the model and is discounted. The larger risk is a deeper price cut than I assume. Whether Tremfya and Icotyde last. Tremfya’s patent runs to 2031 and Icotyde and Imaavy are brand new.³² The final cost of talc. The reserve is $9.0B and I add nothing to it in the base; $5B more costs about $2 a share, which tells you how small it is beside the market value.⁵ What the Orthopaedics split does. I keep Orthopaedics in the model through 2036 because holders get its shares, and I charge separation and stranded costs inside “other cash costs” as an estimate, without knowing the size.⁵ Tariffs and the government agreement. The January 2026 agreement, per a press report, trades price commitments for an exemption from certain tariffs, and sits beside a $55B US investment plan through 2029; the terms are not public.¹⁴ The 10-Q shows tariff recoveries of about $0.2B in MedTech cost of goods in the quarter, and I do not know what produced them.⁵ What the deals earn. Intra-Cellular, Halda, Firefly and Sail are young, and I can’t value them.

The discount rate

I use 9.0%, the same hurdle as on my Eli Lilly, AbbVie and Novo Nordisk pages. The ten-year Treasury yields 4.90% and Damodaran’s implied equity risk premium was 4.20% on 1 October.¹⁹²¹ J&J’s month-end beta against the S&P 500 ETF is 0.44 over ten years, 0.19 over five and −0.11 over three, which gives 6.7%, 5.7% and 4.4% C (calculated).¹²² Those numbers say the stock has moved little with the market, and they argue for a lower rate than 9.0%, more so than for the other pages. They say nothing about whether the cash flows hold, so I keep 9.0% for comparison with the others and show 7% and 8% in the table. At 7% my base is $140, still about half the price.

Three cases

The model is built line by line. I take sales as ten lines: Darzalex, Tremfya with the newer immunology drugs, the other cancer drugs, Stelara, neuroscience and the rest of Innovative Medicine, then cardiovascular, surgery, vision and orthopaedics in MedTech. I set the operating margin before purchased intangibles and acquired research, tax it at 17.5%, and take off capital spending above depreciation, working capital at 10% of added sales, other cash costs, and deals at 5.5% of sales. Stock pay is already in the margin and I do not add it back. What is left is owner cash before interest. I then take off net debt, the talc reserve and the opioid payments, as claims ahead of the shares.

Two choices matter. I treat the $9.0B talc reserve and the $0.9B of opioid payments as debt and do not count them in the yearly cash, so there is no double count; the $5.5B ovarian settlement is already inside the $9.0B. I take no credit for the tax deduction on those payments, which would add about $2B, or $0.80 a share. And I charge deals every year, because a company that loses a patent every few years has to buy or build the next one. If I left deals out, the base would be $125, and I think that would be wrong.

My base 2027 operating income is $35.4B. After about $0.2B of net interest and a 17.5% tax rate, that is about $11.84 a share before acquired research, against the consensus $11.55 for 2027. My 2028 is $12.66 against $13.84 E (estimate).¹² So I am close to the Street on the near year, higher only because I keep acquired research out of the margin and charge it in the deals line, and lower on the next, and the difference grows as the dates begin to matter.

Bear. Darzalex copies arrive on the patent date and take about a quarter of its sales a year from 2031 to 2033, Tremfya is overtaken by cheaper rivals before its patent ends, the other cancer drugs stall, Medicare cuts reach several of the larger drugs, MedTech grows 2% to 4%, margins slip to 27.6% and deals take 7% of sales. Sales fall to $91B by 2036. Base. Darzalex grows to $21B in 2029, Tremfya and its successors to $17B in 2031, Stelara fades to $0.2B, MedTech grows about 4% a year to $52B, and copies and Medicare take most of Darzalex by 2036. Sales end at $120B, 1.7% a year. Bull. Darzalex holds to $24B, Tremfya and Icotyde reach $21B, cancer and mind drugs grow 4% to 20% a year, MedTech reaches $58B, the margin rises to 36% and deals fall to 4.5% of sales. Sales end at $154B.

Where is the cliff in the base? It is not one year. Sales peak at $123B in 2031 and are $120B in 2036. MedTech carries growth in the 2030s, from $37B to $52B, while Innovative Medicine falls from $80B to $68B E (estimate). The year-ten cash flow is cut by 10% before the long-run value, for what follows 2036, and then grows at 3%.

The price sits 204% above my base and 91% above my bull. That is the plain result. Why might the market be right? Because it can see what I can: the largest cancer drug growing 17% a year, an immunology drug, Tremfya, growing 68% in the quarter, a business that has already absorbed an $8B loss, a AAA balance sheet, and a company that says it has a clearer line of sight to double-digit growth.⁴² It may be paying for what I leave out, such as a long Darzalex life, an Icotyde bigger than Tremfya, or a Medicare outcome kinder than I assume. If those are worth more than I can price, my values are too low. It may also be paying for safety, because a stock that has moved little with the market in the past can trade at a lower rate of return. The $39 bear is a fair picture of the world in which the dates bite and the deals do not pay.

The same dollar, elsewhere

The ten-year Treasury pays 4.90%.¹⁹ Against that, my base returns 5.1% a year from this price, my bear 3.9% and my bull 6.3% E (estimate). The nearest comparisons on this site are AbbVie, where my base equals the price only at 5.6%, Eli Lilly at 6.3% and Novo Nordisk at 7.3%.¹⁵⁹¹⁶ So J&J’s price sits further above my base than any of those, even though it has the least concentrated sales and the strongest balance sheet. That is not a contradiction. Quality and price are two separate questions, and here the business is excellent and the price is high. Novo is the closest of the group to its base. J&J’s price has risen 46% in twelve months on results that met guidance.¹ At $276.32 an owner gets about 2.8% in owner cash, plus growth, which is close to the Treasury yield plus a few points.

From my value to the price, one change at a time

How the odds shrink

If the case needs many things to go right at once, multiply them. The price needs, roughly, all of these. Darzalex keeps growing and holds its price past 2030: 40%. Tremfya and Icotyde together reach $25B or more: 35%. MedTech grows 6% or more a year with a rising margin: 40%. Talc closes at or near the $9.0B reserve: 70%. Deals earn their cost: 60%. Those are my guesses. Multiplied they give 2% E (estimate). They overlap, so perhaps 5% to 10% in practice. A price that needs something like a one-in-fifteen outcome leaves no room for error, even for an excellent business.

The margin of safety

My margin-of-safety price is $64, the base value less 30%, a cushion for the six things I cannot judge. The price is 4.3 times that. At $64 my base would return 11.2% a year and the dividend yield, at $5.36, would be 8.4% E (estimate). The bear case is $39. I don’t expect the price to go there. The point is what an owner would need to see before the numbers favour them, and it is not close.

VIIThe case for the price

What most people miss

Three facts cut in J&J’s favour and get little attention. First, the replacement shows in the numbers: in nine months of 2026 Tremfya made $6.05B against $2.11B for Stelara, and its US composition patent runs to 2031.⁴³ Second, operating cash flow rose through the Stelara loss: $22.8B in 2023, $24.5B in 2025 and $26.2B in the twelve months to September 2026 C (calculated).¹¹³⁵ Third, the largest overhang has a number. The talc reserve is $9.0B, the $5.5B ovarian settlement is inside it, and that settlement is conditioned on at least 95% of the remaining claims joining.⁵

The bull case

In plain words: J&J lost a $10B drug and grew anyway, and the next drugs and devices are early in their life. The chairman and chief executive, Joaquin Duato, said on 13 October that “the accelerating momentum of our focused portfolio” gives the company “clearer line of sight to double-digit growth by the end of the decade.”² The company raised its 2026 adjusted EPS guide to $11.10, won approval for Imaavy in a new condition and received US marketing authorization for Ottava.² The Nasdaq consensus has EPS of $11.55 in 2027 and $13.84 in 2028, and at 20.0 times 2028 that is below Lilly at 25.4 times.¹²⁹ On that view the shares are a fair price for a compounder with a 64-year dividend record.

I found no named investor or analyst statement on J&J that I could open, so I attribute the case only to the company and the consensus. The Nasdaq mean one-year target is $287.88, with no firm named.¹²

Against reality

Much of the case is already in the price. My bull has sales growing 4.3% a year to $154B with the margin rising to 36%, and it is worth $145 E (estimate). The price asks for 15.6% a year, or $430B of sales in 2036 C (calculated). If J&J reached the company’s aim of double digits, the price is possible: 10% growth for ten years at a 36% margin is worth $210 at my 9.0% rate and $337 at 7% E (estimate). But sales grew 7.4% in nine months of 2026, and currency added 1.4 points of that.⁴

The argument is weakest on time. Darzalex’s US patents run to 2029 and Tremfya’s to 2031, and Medicare can reach both.³ The Street’s EPS for 2029 grows only 5% on 2028.¹² Past deals ran to $35.2B for Actelion and $14.5B for Intra-Cellular.¹¹⁵ Multiplied out in Section VI, the five things the price needs come to about 2%, or 5% to 10% after overlap E (estimate).

What a 10x would take. Ten times $276.32 is $2,763, about $6.8T of market value C (calculated). At 20 times earnings that needs about $338B of net income, 12.4 times this year’s adjusted $27B, or about $1.26T of sales at the same margin. By 2036 that is 29% growth a year, against the 3% J&J managed over nine years with Kenvue in it. The Eli Lilly page values Lilly, the largest drug company on this site, at about $1.27T.⁹ The one case I can show is NVIDIA, from about $0.51T on 2 July 2021 to about $5.8T on 6 October 2026, in an AI chip boom C (calculated).²³²⁴ A 10x for a company this size is close to impossible, and the price does not need it.

Where I come down

The replacement of Stelara moved me: it is done, and cash flow rose through it. The balance sheet moved me too. The price did not: it needs 15.6% growth for ten years, and nothing I can open says that is likely. What would raise my values is a Darzalex biosimilar date after 2032 or a Medicare exclusion, Tremfya and Icotyde together passing $25B, and two years of deals under $6B.

VIIIRisks, and what would change my view

The three ways this goes badly wrong are at the start of Valuation. These are the rest.

• Medicare price-setting. Imbruvica has carried a government price since 2026, and Erleada, a $3.6B drug, is on the list for 2028.³¹³ The next list is due in early 2027. A rare-disease exemption widened in 2025 may protect the myeloma drugs, but I can’t tell, and a list that includes Darzalex or Carvykti would move my numbers.¹³
• The government agreement. J&J has signed voluntary agreements with the US government on patient affordability, pricing programs and US investment, and a press report says they include an exemption from certain tariffs.⁵¹⁴ I don’t have the terms, the length or the price cost, so I can’t say what it gives up or what ends it.
• Talc beyond the reserve. About 76,500 plaintiffs have direct claims, the ovarian resolution depends on 95% participation, the mesothelioma docket is outside it, and the company also faces litigation in Canada.⁵ The company says it cannot estimate a loss beyond the amounts accrued for the matters it lists.⁵ I add nothing to the reserve in the base and show $5B more in the swing table.
• Other litigation. A class action over delaying Stelara biosimilars was certified in December 2025 and is partly on appeal; Biosense Webster is appealing an antitrust verdict and a permanent injunction; two qui tam suits over Zytiga pricing and HIV drug promotion continue; and an amended false-claims complaint naming DePuy Synthes was filed in June 2026.⁵ I can’t size them beyond the reserve.
• The Orthopaedics separation. The method is not set in the 10-Q, costs have been $0.7B so far this year, and the stranded costs after the split are unknown.⁵
• Devices under pressure. Electrophysiology faces pulsed field ablation competition, Surgery faces China volume-based procurement, and Abiomed’s third-quarter sales fell.⁵
• Royalties. The largest drug pays Genmab 12% to 20%, about $2.4B in 2025, and Tremfya pays about 5%.³ Both rise with sales.
• Accounting gap. Adjusted EPS was $8.19 in nine months of 2026, against GAAP EPS of $4.49. The gap is talc, purchased-intangible amortization, restructuring and separation costs, but a reader who uses only adjusted numbers sees more earnings than the business pays out.²⁴
• Currency. Nine-month sales rose 7.4% reported and 6.0% before currency, so 1.4 points came from exchange rates.⁴ The 2026 guide assumes a euro at $1.15.²