Merck’s Keytruda Cliff Is Real. The Panic Might Be Overdone
Roughly half of Merck‘s revenue sits in one injection. Keytruda and its newer under-the-skin version, Keytruda Qlex, were reported at about $31.7 billion of sales in 2025 (that is the figure in press coverage of the results, for example BioPharma Dive coverage), against total company revenue of $65.0 billion. That works out to 49 percent, and the key US patents behind it start expiring in 2028. Merck trades around $146.87 as I write this, only 5.9% below its 52-week high of $156, and nearly double the 52-week low of $75. So the obvious question is why anyone pays up for a company with a cliff on the calendar.
My answer is that the cliff is real but the market is pricing a smaller fall than the headline suggests, and the price already leaves little room for an unpleasant surprise. I would call Merck a fair-value holding with a specific job to do before 2028, not a bargain and not a disaster in waiting.
Half the revenue, one molecule
Start with the arithmetic, because most cliff commentary skips it. Merck’s 2025 revenue was $65.0 billion, up from $64.2 billion in 2024, $60.1 billion in 2023 and $59.3 billion in 2022. Growth has been a slow 1 percent last year and about 3 percent a year across those three years. Take out Keytruda at around $31.7 billion and the rest of the company sold about $33.3 billion. That remainder is a large business on its own, but it is the part that has to carry the weight.
Now run a rough stress case. Suppose Keytruda sales fall by half after biosimilars arrive. That removes about $15.9 billion, or 24 percent of 2025 revenue. To hold total revenue flat, the other products would need to grow from $33.3 billion to about $49.2 billion, a 48 percent increase. Spread over four years, that is roughly 10 percent a year. Over five years it is closer to 8 percent. This is my own arithmetic, not a company forecast, and the half-decline is an assumption I picked to make the problem concrete.
Ten percent a year sounds hard for a mature portfolio. It is less hard when a company is launching drugs and buying assets, which is what Merck has been doing. But it is a real hurdle, and it is the number I would keep in front of me instead of the scarier “$30 billion at risk” framing.
Why oncology patents behave differently
Small-molecule pills lose most of their sales within a year or two of generic entry. Antibodies like Keytruda tend to erode more slowly, because biosimilars are harder to make, hospitals and oncologists are cautious about switching a cancer patient’s therapy, and the drug is embedded in many treatment regimens. I read that as a reason to model a slope rather than a drop. It is not a reason to model no decline.
The timing also matters. The 2028 date refers to key US patents beginning to expire, according to the coverage I found, and biosimilar makers still need to win approval and build supply. Pricing pressure can arrive before volume moves. Medicare price negotiation in the US is a separate lever that can compress revenue independent of any patent, and I have not tried to size it here because I could not verify a figure I trusted.
The subcutaneous shield
This is the part I care about most.
Merck’s main defense is the subcutaneous version. The FDA approved Keytruda Qlex in September 2025, according to BioPharma Dive coverage. It combines pembrolizumab with an enzyme that lets the drug be given as an injection under the skin in a few minutes, instead of a longer intravenous infusion. Patients and clinics get a convenience benefit. Merck gets something more valuable, which is a new patent estate. Trade coverage reports protection running into the 2040s, though I would treat the exact year as something to confirm in Merck’s own filings before relying on it.
The logic is simple to state and hard to execute. A biosimilar copies the intravenous drug. If most patients have already moved to Qlex by the time copies arrive, the biosimilar competes for a smaller pool. The catch is that doctors and payers decide, not Merck. If insurers reimburse the two formats at similar rates and clinics like the shorter chair time, conversion could go fast. If reimbursement is awkward, or if competing biosimilar makers seek their own subcutaneous versions, conversion could stall.
I dropped the specific conversion target that older articles about this stock quote, because I could not verify it. Merck’s investor updates are where to look for the current number, and in the 2025 annual report on SEC EDGAR.
The rest of the pipeline is a bet on many small wins
No single product replaces $31.7 billion. The realistic outcome is a stack of products in vaccines, cardiovascular, and other areas each adding a few billion, plus whatever acquisitions bring in. Merck has the cash generation to buy: operating income was $22.1 billion in the latest fiscal year, a 34 percent margin, and net income was $18.3 billion, a 28 percent net margin. Companies with that much cash have options, and they do not have to wait for their own labs.
That flexibility is also a risk. Large deals often cost a lot up front and take years to pay back. Merck’s trailing earnings per share of $1.25 looks strangely low against forward earnings of $8.00, and that gap is why the trailing P/E of 117.5 is meaningless as a valuation signal. I cannot tell from my data what drove the low trailing figure, but a large one-time charge tied to buying a pipeline asset is the usual cause in pharma, and I would check the latest 10-Q on SEC EDGAR to see it. Anyone looking at the 117.5 times figure and concluding Merck is wildly expensive is reading the wrong number.
What the price asks for
Use the forward multiple instead. At 18.4 times forward earnings, Merck is priced like a mature, slow-growing company. That already includes some of the cliff. The five-year average P/E in my data, 73.8, is inflated by the same kind of one-off charges, so I ignore it as a benchmark. Price to sales tells a cleaner story: 5.4 now against a five-year average of 4.3. So the stock is more expensive on revenue than it has usually been, even as the revenue is about to face a headwind. Price to book is 8.5 against 6.0.
I take two things from that. Buyers are not treating the cliff as an emergency, and they are paying more per dollar of sales than they typically did. The stock has re-rated on confidence that Qlex and the pipeline will fill the gap. If that confidence is wrong, the multiple is what gets hurt first.
| Metric | Value | Context |
|---|---|---|
| Share price | $146.87 | 52-week range $75 to $156 |
| Market value | $362.4 billion | about 5.6 times 2025 revenue |
| Revenue, 2025 | $65.0 billion | $59.3 billion in 2022 |
| Net margin | 28% | operating margin 34 percent |
| Forward P/E | 18.4 | based on $8.00 expected EPS |
| Price to sales | 5.4 | five-year average 4.3 |
| Dividend yield | 2.29% | $3.36 a share over 12 months |
| Analyst target | $163 | 11% above price; range $125 to $180 |
Income while you wait
The dividend is $3.36 a share over the last twelve months, a yield of 2.29%. Against forward earnings of $8.00 that is a payout of about 42 percent, which leaves space to keep raising it even if earnings dip. That matters for anyone holding through the transition, because a yield above 2 percent is a modest but real payment for waiting. I compared it with other income options in my piece on dividend stocks worth holding for real income; Merck does not yield as much as some, but it has more cushion than most.
The analyst view is favorable and not extreme. Of 18 analysts covering the stock, 83% rate it a buy, and the average target of $163 sits 11% above the current price. The lowest target is $125, which is 15 percent below the price, so the worst-case analyst still sees a loss of that size rather than a collapse. I would not lean on these numbers. Analysts anchor to guidance, and guidance is precisely what the cliff puts at risk.
Where I would be wrong
The case against my view is specific. If subcutaneous conversion is slower than Merck hopes, biosimilars would hit a bigger share of intravenous sales than the market assumes, and the 10 percent annual growth hurdle for the rest of the business would look too easy. A drug loss of more than half in the first two years after entry would break my arithmetic. A second way to be wrong is on the pipeline side: if two or three important late-stage programs fail in the same year, there is no stack of small wins to write about.
And there is one point where I remain uncertain. I cannot see the mix of Keytruda sales between the intravenous and subcutaneous versions from my data, so my view on how fast conversion is going is based on the approval and the logic, not on numbers. That is the biggest hole in this analysis.
I also have to admit the opposite risk. If Qlex works and Merck buys well, today’s multiple could look cheap in two years. A company with this much profit does not stay quiet, and a fixed date focuses management. I made a similar point about a different company in my Intel piece, where the deadline was closer and the cushion thinner.
The number I would watch
Quarterly revenue was $16.6 billion in the most recent report, 5 percent higher than a year earlier. That growth rate is what I would track. If it stays at or above 4 percent through the next few reports while Keytruda is still growing, the company is building its replacement revenue at the right time. If it drops below 2 percent before 2028, the rest of the business is not growing fast enough to fill a gap even half the size I used in my stress case, and I would want a lower price. Around $130 to $135, roughly 10 percent below where it trades today and near the analyst low, I would be more interested. Around $146.87, a holder could reasonably wait for the conversion data.
Analysis and opinion only, not investment advice. Figures come from Merck’s filings on SEC EDGAR and its investor materials, plus the press coverage linked above; valuation multiples are approximate and were checked on September 22, 2026.