Medicine for animals — and two customers who could not be less alike
Zoetis makes medicines, vaccines and diagnostic tests for animals. It was spun out of Pfizer in 2013 — remember that, because it matters later — and it is the largest animal health company in the world.
It sells to two entirely different customers, and almost everything that has happened to the share price in the last sixteen months comes from the gap between them.
And then split the same quarter geographically, because that is where the story becomes unmistakable.
American pet owners have become price-sensitive and are taking their animals to the vet less often. Everything else in this analysis — the guidance cuts, the halved share price, the lawsuit — flows from that sentence and from a second one about competition. It is worth holding both in your head from the start.
Chief executive Kristin Peck put it plainly on the Q2 call: "Second quarter results reflected a more pressured Companion Animal market, as lower clinic visits and pet owner price sensitivity reduced demand across parts of our portfolio and heightened competition in key categories."
Four structural advantages that human pharmaceutical companies would pay almost anything for
Before we look at what went wrong, it is worth being precise about what Zoetis is, because the quality here is not a matter of opinion and it has not gone away.
We wrote about Pfizer five days ago and about Eli Lilly last month. Set animal health against human pharmaceuticals and the differences are structural rather than incidental.
| Fiscal year | Revenue | Operating margin | Diluted EPS | Shares out (m) |
|---|---|---|---|---|
| 2016 | $4,888m | 28.7% | $1.65 | 498.2 |
| 2019 | $6,260m | 32.2% | $3.11 | 481.8 |
| 2022 | $8,080m | 36.2% | $4.49 | 470.4 |
| ★ 2025 | $9,467m | 38.0% | $6.02 | 443.8 |
| ★ Nine-year change | +93.7% | +9.3 pts | +264.8% | −10.9% |
Earnings per share compounded at 15.5% a year for nine years while the operating margin rose by nine percentage points. That is not a good business having a good run. That is a structurally advantaged business behaving exactly as its structure predicts.
And the current profitability is still there. This is what the company earns today, in the middle of what everyone agrees is a bad year:
Screens will show you a return on equity of 69.3% for Zoetis, and it is a number that should be ignored rather than admired.
It is an arithmetic artefact of buybacks. Repurchased shares are carried against equity, so a company that buys back aggressively shrinks its own denominator. Zoetis spent $3,235 million on its own stock in 2025 alone, and shareholders' equity fell from $4,770m to $3,331m. Divide roughly $2.3bn of profit by a shrinking $3.3bn of book value and you get 69%. It tells you about the treasury account, not about the business.
Use return on invested capital instead: 21.6%. That figure includes the debt, and it is the one that says whether a dollar put into this company comes back as more than a dollar. It does. We made the same distinction for IBM and for exactly the same reason.
Two guidance cuts in fourteen weeks, and a company now forecasting decline
Here is the sequence, dated, because the dates matter.
| Full-year 2026 | Guidance in early 2026 | Guidance now | Change |
|---|---|---|---|
| Revenue | $9,680–9,960m | $9,120–9,320m | −$600m |
| ★ Organic operational growth | +2% to +5% | −3% to −1% | from growth to decline |
| ★ Adjusted diluted EPS | $6.85–7.00 | $6.15–6.25 | −10% |
| ★ Adjusted EPS organic growth | +9% to +11% | −9% to −5% | a swing of ~16 points |
| Adjusted net income | $2,870–2,950m | $2,570–2,620m | −$315m |
Read the fourth row twice. Zoetis began this year expecting to grow earnings by nine to eleven per cent. It now expects them to fall by five to nine. That is a sixteen-point swing inside eight months, and it is the entire reason a company that traded at thirty-five times earnings now trades at twelve.
We should be fair about one thing. Reported earnings per share actually rose in the second quarter — $1.65 against $1.63, up 1% — and adjusted earnings per share rose 5% to $1.87. Revenue was flat. This is not a collapse; it is a stall, and the stall is concentrated in one country and one customer group.
But the market does not pay thirty-five times earnings for a business that stalls. It pays thirty-five times for one that compounds. The de-rating is not irrational — it is the market repricing a growth company as something else. The only question worth asking is whether it has now overshot.
The most striking fact in this analysis — and still not the reason the stock fell
Librela is an injection that controls the pain of osteoarthritis in dogs. It launched in the United States in 2023, grew quickly, and has since become the most argued-about product in veterinary medicine. Owners' groups have collected thousands of adverse-event reports. Regulators have been drawn in. It is where most of the noise around this company comes from.
We want to make an argument here that we have not seen made elsewhere, and then immediately argue against its importance.
Librela — the molecule is bedinvetmab — is a monoclonal antibody that blocks nerve growth factor. Nerve growth factor, or NGF, is a protein involved in transmitting pain signals. Block it and the pain stops. It is elegant, and for a while it was one of the most exciting ideas in pain medicine anywhere.
Anti-NGF drugs were developed for humans too, and they failed — for one specific reason.
| The human drug | Tanezumab, an anti-NGF monoclonal antibody for osteoarthritis pain, developed by Pfizer in partnership with Eli Lilly. |
| What happened | ★ In March 2021 an FDA advisory committee voted 19 to 1 that a risk-mitigation plan could not make the benefits outweigh the risks. The specific risk was rapidly progressive osteoarthritis — accelerated destruction and collapse of the joint the drug was meant to relieve. |
| The outcome | The European regulator's committee also rejected it. The FDA issued a complete response letter. Pfizer and Lilly discontinued the global development programme entirely. |
| ★ In dogs, June 2026 | A paper in the Journal of the American Veterinary Medical Association (von Pfeil, Adams, Clark et al., published 26 June 2026) documents cases of rapidly progressive osteoarthritis in dogs treated with bedinvetmab — severe joint laxity, loss of a femoral condyle, bilateral fibular fractures — and notes that the manufacturer categorised the associated ligament ruptures, fractures and joint luxations as non-serious. |
★★ The drug class at the centre of Zoetis's most controversial product was rejected 19–1 by an FDA advisory committee for human use, on precisely the safety signal now being documented in dogs — and the company that abandoned it in humans is the company Zoetis was spun out of.
We are not veterinarians and we are not making a clinical claim. NGF biology is not identical across species, dogs are not people, and the dose, duration and endpoints all differ. Zoetis states that more than 25 million doses have been distributed globally and that no individual adverse event has been reported at a rate above "rare" as the European regulator defines it — fewer than ten per ten thousand animals treated. The company also updated the US label in 2025, on its own submission, to add neurologic signs observed after approval.
What we are saying is narrower and, we think, useful: the honest question about Librela is not "was there a bad news cycle?" but "is the mechanism itself the problem?" — and there is a documented, expensive, nineteen-to-one precedent suggesting that in at least one species it was. That is a materially harder question than the one most commentary is asking, and an investor should know it exists.
Having made that case, we have to be honest about scale, because scale is what decides whether a controversy is an investment issue.
Librela is the headline. It is not the reason the shares halved. Four and a half per cent of revenue declining 8% does not take a company from thirty-five times earnings to twelve. The reason is duller, larger and more worrying, and it is in the next two parts: the base business is under attack.
One note in Zoetis's favour that deserves stating. The company has launched two successor long-acting pain antibodies — Lenivia and Portela — in Canada and the European Union, with approvals in Great Britain and Switzerland. It has not retreated from the category. Whether that is conviction or sunk cost is a question the next two years will answer.
Two franchises, two better labels, and generics arriving underneath
Zoetis's two largest companion-animal franchises are parasiticides — flea, tick and worm prevention, led by Simparica Trio — and dermatology, the itch and allergy business led by Apoquel and Cytopoint. Between them they are the engine of the company. Both are now being attacked with specific, dated, credible products.
| Zoetis franchise | The attacker | The evidence |
|---|---|---|
| ★ Simparica Trio (parasiticides) | Credelio Quattro — Elanco | Reached $100m of net sales in under eight months — the fastest pet-health blockbuster in Elanco's history. In July 2026 Elanco published head-to-head data claiming it kills black-legged and lone star ticks faster and more consistently through day 28 than Simparica Trio, and it carries an FDA-approved Lyme disease prevention claim. |
| ★ Apoquel (dermatology) | Zenrelia — Elanco | More than half a million dogs treated. Elanco ran a head-to-head non-inferiority study against Apoquel itself — which is to say, it set out to prove it was as good, and priced accordingly. |
| Cerenia, Convenia | Generic manufacturers | Zoetis named generic competition on both as a driver of the US companion-animal decline in the second quarter. These are older, established products — exactly the long-lived brands that were supposed to be safe. |
This is the honest bear case, and it has nothing to do with Librela. A competitor with a demonstrably better tick label and an approved Lyme claim, aimed at your largest franchise, at a lower price, into a customer base that has just become price-sensitive for the first time in a decade. That combination is how pricing power ends.
Two things soften it, and they are real.
Three and a quarter billion dollars, spent just before the fall
This is the weakest part of the case, and we are not going to soften it.
| Fiscal year | Buybacks | Dividends | Total debt | Shareholders' equity |
|---|---|---|---|---|
| 2021 | $743m | $474m | $6,743m | $4,543m |
| 2023 | $1,092m | $692m | $6,755m | $4,997m |
| 2024 | $1,858m | $786m | $6,744m | $4,770m |
| ★ 2025 | $3,235m | $889m | $9,493m | $3,331m |
In 2025 Zoetis spent $3,235 million buying its own shares — more than in the previous two years combined — and increased total debt by $2,749 million in the same year. Free cash flow that year was $2,283 million. The buyback alone exceeded it by nearly a billion dollars, before the dividend.
Within months of that spending, the shares fell from a 52-week high of $157.21 to $77.17.
We do not know the precise average price Zoetis paid, and we will not invent one. What we can say is arithmetically unavoidable: the company borrowed money to buy $3.2 billion of its own stock in the twelve months before that stock halved. Whatever the average, a very large part of that money is gone.
The fair defence. Nobody times a de-rating. Zoetis was buying a business it believed was compounding at nine to eleven per cent, at a multiple that had been normal for that business for a decade. Directors who refuse to buy their own shares at any price are not exercising judgement either. And Zoetis had, at the time, one of the best long-run records in the industry to point to.
The reply. A management team that has just guided earnings down by sixteen points inside eight months, having spent a record sum on its own equity at the top, is a management team whose forecasting we should discount. That is not an accusation of bad faith. It is what the evidence supports. And a securities class action now alleges rather more than that — see Part VIII.
★ A new combined CFO and COO arriving straight after a guidance reset is usually one of two things: a company installing someone to run the cost base harder, or a company preparing to change how it reports and guides. Both are plausible. Both would be visible within three quarters. We would treat the next two results as the real test of this appointment, and of the credibility of the guidance that comes with them.
Comfortably. But look at what happened to the growth rate.
Zoetis yields 2.75% at $77.17 — the highest it has ever been, purely because the price fell. The dividend itself is in no danger whatsoever, and the arithmetic is quick.
| Test | Value | Reading |
|---|---|---|
| ★ Cover on free cash flow | 2.7× | Free cash flow per share of $5.67 against $2.12 of dividend. Our house test is free cash flow, not earnings — and it passes easily. |
| Cover in dollars | $2,283m vs $889m | 2025 free cash flow against 2025 dividends paid. |
| Payout on earnings | 34.0% | A third of profit. Room to fall a long way before it binds. |
| Funded by | operations | Operating cash flow of $2,904m in 2025 against $621m of capital spending. The dividend is not funded by debt. The buyback partly was — a different question. |
| Balance-sheet room | 2.0× | Net debt to EBITDA. Interest covered 14.4×. Current ratio 3.08. Altman-Z 4.73 — comfortably in the safe zone, and here the score is actually applicable. |
| Declared | New quarterly dividend | Increase |
|---|---|---|
| Dec 2021 | $0.325 | +30.0% |
| Dec 2022 | $0.375 | +15.4% |
| Dec 2023 | $0.432 | +15.2% |
| Dec 2024 | $0.500 | +15.7% |
| ★ Dec 2025 | $0.530 | +6.0% |
Four consecutive years of 15% or better, then 6%. That decision was taken in December 2025 — five months before the first guidance cut and eight months before the second. The board knew something about 2026 before the market did, and the dividend told you.
We would not call this a warning that was hidden. It was published. It simply was not read, because a 6% increase from a company everyone assumed was compounding at 10% looks like a rounding error rather than a signal. It was a signal.
What would force a cut: essentially nothing on the current shape of the business. Free cash flow would have to fall by more than 60% and stay there. The realistic risk is not a cut but that the dividend grows at 5–6% rather than 15% for several years — which, for an investor buying at a 2.75% yield, is a materially different proposition from the one the last decade advertised.
★ One genuine caution. Zoetis returns roughly three and a half times as much through buybacks as through dividends. The buyback is the real capital-return vehicle here, and Part VI is about how badly it has been executed. An investor treating the dividend as the whole of the shareholder return is looking at the smaller and better-managed half.
Verified afresh, 25 August 2026
The securities class action is live and is the material legal exposure. It is In re Zoetis Inc. Securities Litigation, case number 26-cv-04401 in the Southern District of New York, brought under the Securities Exchange Act of 1934 against the company and certain officers. The class period runs from 14 January 2025 to 6 May 2026. The lead-plaintiff deadline passed on 27 July 2026 and the case remains at an early procedural stage.
The allegation, stated fairly: that the defendants made materially false and misleading statements about the strength, competitive positioning and growth trajectory of the companion-animal portfolio — Librela, Apoquel, Cytopoint and Simparica Trio — while concealing that FDA safety warnings were eroding veterinarian confidence in Librela and that lower-priced competitors were capturing share in dermatology and parasiticides. The claimed corrective disclosures run from August 2025 to the 21.5% single-day fall on 7 May 2026.
★ Note what makes this uncomfortable: the allegation is not about accounting or fraud in the usual sense. It is that management described a business as growing while it was being taken apart. That is the same question Part VI raises about the buyback, arriving from a different direction. We express no view on the merits — securities class actions follow essentially every large share-price fall in America, and most settle for a fraction of the claim. But it is live, and it is not trivial.
The separate consumer litigation is resolved in Zoetis's favour. A federal court in New Jersey dismissed the putative class action brought by eight pet owners over Librela on 22 October 2025, ruling on the third amended complaint that the design-defect claims failed because the plaintiffs identified no feasible alternative design. That materially reduces the product-liability overhang, and it deserves stating as clearly as the bad news.
★ The risk we would rank first is none of the above. It is that competitive share loss in parasiticides and dermatology proves permanent. Everything else here — price-sensitive owners, fewer clinic visits, even Librela — is recoverable. A veterinarian who has spent two years comfortably prescribing a rival product is not.
The cheapest this business has ever been, and the reason has a name
| Measure | Value | Reading |
|---|---|---|
| Share price, 24 Aug close | $77.17 | Market capitalisation $32.35bn; enterprise value $40.11bn |
| ★ 52-week range | $72.38 – $157.21 | 50.9% below the high, 6.6% above the low. The stock has more than halved in twelve months. |
| P/E on trailing earnings | 12.7× | Against a decade spent in the thirties. |
| ★ P/E on company guidance | 12.4× | On the midpoint of the company's own reduced adjusted EPS guidance of $6.15–6.25. Not on a hopeful consensus — on the number management has already cut twice. |
| ★ Free cash flow yield | 7.3% | $2.28bn of free cash flow. Price to free cash flow of 13.7×. Even if free cash flow fell 20%, the yield would still be 5.8%. |
| EV / EBITDA | 10.4× | On a business earning a 37% operating margin. |
| Net debt / EBITDA | 2.0× | Interest covered 14.4×. Altman-Z 4.73 — safe zone, and here the score genuinely applies. |
| Dividend yield | 2.75% | $2.12 annualised, covered 2.7× by free cash flow. |
| ⚠️ Our DCF feed | $188.78 | Implying +145%. Reported and not used — it extrapolates a growth rate the company has just told us has stopped. The mirror image of the fault we flagged on Deere. |
| All-time average target | $153.48 (33 targets) |
| Average over the last year | $116.15 (20 targets) |
| Average over the last quarter | $87.71 (7 targets) |
| ★ Average over the last month | $83.33 (3 targets) — the live number. Median across all live targets: $85. Range $80 to $160. |
| ★ Recommendations | 1 strong buy · 14 buy · 17 hold · zero sells, from 32 analysts. Consensus label: Hold. ★ Not one sell rating on a stock that has halved — the sell side has cut its price targets by 46% while refusing to call the business broken. We read that as an honest reflection of the situation rather than as cowardice. |
| Fiscal year | Consensus EPS | Analysts | P/E at $77.17 |
|---|---|---|---|
| 2026 | $6.29 | 10 | 12.3× |
| 2027 | $6.52 | 12 | 11.8× |
| 2028 | $7.09 | 6 | 10.9× |
| 2029 | $8.34 | 3 | 9.3× |
| 2030 | $9.34 | 2 | 8.3× |
Note the analyst counts falling away — ten, twelve, six, three, two. The 2029 and 2030 figures are the opinion of two or three people and should be read as illustrations, not forecasts. The honest window is 2026 and 2027, and it says this: consensus expects earnings to trough this year and grow about 4% next year. Nobody is forecasting a collapse. Nobody is forecasting a return to 10% either.
So: what does 12.4× assume? It assumes that a business earning a 70.9% gross margin and a 21.6% return on invested capital, in end markets that grow because people keep acquiring pets and keep spending more on them, will not grow again. That is a strong assumption to make about a company that has grown earnings at 15.5% a year for nine years and whose international business grew 6% last quarter.
It is not an absurd assumption. Competitive share loss can be permanent, and we said in Part V that it is the risk we rank first. But at thirty-five times earnings you needed the growth story to be right. At twelve times, you need it merely not to be catastrophically wrong. That is a very different bet, and it is a much better one.
I have spent a good deal of my life looking for businesses where nobody stands between the company and its customer with the power to set the price. They are rarer than you would think. In most of medicine there is an insurer, or a government, or a pharmacy benefit manager, and whatever the doctor and the patient believe the medicine is worth, someone else decides what will actually be paid.
In animal health there is no such person. The dog gets sick, the owner takes him to the vet, and the owner pays. There is no formulary, no prior authorisation, no committee. And because the dog is a member of the family, the owner pays without much argument. I have long thought this one of the most attractive structural arrangements in commerce, and Zoetis is the largest company that enjoys it.
There is a second advantage that I think is even less appreciated. Human pharmaceutical companies live and die on the patent clock. We wrote about Pfizer five days ago; it has told its investors to expect seventeen billion dollars of revenue to disappear to expiries by 2030. Animal health does not work like that. The products are smaller, cheaper to develop, and they persist for decades on the habit and trust of the veterinarian who prescribes them. There is no cliff. There is a long, gentle slope, and often not even that.
On those two advantages Zoetis compounded its earnings per share at fifteen and a half per cent a year for nine years, and lifted its operating margin from twenty-nine per cent to thirty-eight. It was never cheap, and I never bought it, and until about sixteen months ago I thought that was simply the price of admission for a business of this quality.
The shares are now half what they were. The multiple has fallen from the mid-thirties to twelve and a half times the company's own guidance. So the only question that matters is whether something has broken, or whether the market has merely stopped believing in growth that will turn out to be there.
I want to tell you the most interesting thing I found, and then explain why it is not the answer.
The product at the centre of all the noise is called Librela — an injection for the pain of arthritis in dogs. Owners' groups have collected thousands of adverse-event reports; the regulator has been involved; it is the subject of a great deal of anger on the internet. What almost nobody has written is what kind of drug it is. It is a monoclonal antibody that blocks nerve growth factor — a protein that carries pain signals. Block it and the pain stops.
That idea was tried in people. The drug was called tanezumab. In March 2021 an advisory committee to the American Food and Drug Administration voted nineteen to one that no risk-mitigation scheme could make its benefits outweigh its risks. The risk in question was rapidly progressive osteoarthritis — the accelerated destruction of the very joint the drug was meant to relieve. The European regulator's committee said no as well. The programme was abandoned. And the company that abandoned it was Pfizer, working with Lilly — Pfizer being the company Zoetis was spun out of in 2013.
This June, a paper in the Journal of the American Veterinary Medical Association documented the same condition in dogs given the animal version, with photographs of collapsed joints and fractured bones, and noted that the manufacturer had classified the associated ligament ruptures and fractures as non-serious.
I am not a veterinarian and I am not going to pretend the biology is settled. Dogs are not people, the doses differ, and Zoetis says twenty-five million doses have gone out with no adverse event above the regulator's definition of "rare". But I know an unfashionable question when I see one, and it is this: is the problem the product, or is the problem the mechanism? There is a nineteen-to-one precedent suggesting that in one species, at least, it was the mechanism. An investor should know that question exists, because most of the commentary is asking a much easier one.
And now let me take it away from you, because Librela is not why this stock halved. The whole class of canine arthritis antibodies is about a hundred and five million dollars a quarter — call it four and a half per cent of the company. If it went to zero tomorrow Zoetis would survive it comfortably. The pet-owners' class action, incidentally, was dismissed by a federal judge in New Jersey last October.
The real problem is duller and larger, and it is in two parts.
The first is that the American pet owner has run out of money. Clinic visits are down. Owners are trading down to cheaper alternatives. United States companion-animal revenue fell eleven per cent in the second quarter. That is a recession in the household budget arriving at the veterinary practice, and I regard it as temporary, because people do not permanently stop caring for their animals.
The second is competition, and I regard it as considerably more serious. Elanco has launched a flea and tick product called Credelio Quattro that reached a hundred million dollars of sales in under eight months — the fastest in its history — and in July published head-to-head data claiming it kills ticks faster and more consistently than Zoetis's Simparica Trio, with an approved claim for preventing Lyme disease that Zoetis does not have. It has launched a second product aimed squarely at Apoquel, Zoetis's dermatology franchise. And generics have arrived on two older brands that were supposed to be safe forever.
A better label, at a lower price, aimed at your two largest franchises, arriving in the same year your customer becomes price-sensitive for the first time in a decade. That is how pricing power ends, and a veterinarian who spends two years getting comfortable prescribing somebody else's product does not come back when the economy improves. Of everything in this file, that is the risk I would watch.
Which brings me to management, where I have less patience than I would like to have. In 2025 Zoetis spent three billion two hundred and thirty-five million dollars buying its own shares — more than the two previous years combined and nearly a billion more than its entire free cash flow — while adding two and three-quarter billion of debt. Within months the shares halved. I do not know the average price paid and I will not guess at one, but a great deal of that money is simply gone.
I want to be fair: nobody times these things, and a board that refuses to buy its own stock at any price is not being prudent either. But this is a management team that began the year expecting to grow earnings nine to eleven per cent and now expects them to fall five to nine — a sixteen-point swing in eight months — having spent a record sum on its own equity at the top. A securities class action now alleges they described a business as growing while it was being taken apart. I express no view on the merits; such suits follow nearly every large decline in America. But I will discount this team's forecasts until they have earned the discount back, and I note that a new chief financial officer arrived eight days ago wearing a chief operating officer's hat as well.
There was a signal, incidentally, and it was published. In December 2025 Zoetis raised its dividend by six per cent, after four straight years of fifteen or better. That was five months before the first guidance cut. The board knew. It was written down. Nobody read it.
So what do I actually think?
I think the moat is narrower than it was two years ago and that this is permanent. I think the growth story that justified thirty-five times earnings is over for some years, and possibly for good. I think management has damaged its credibility and will need three or four quarters to begin repairing it. None of that is comforting.
But I also think this. Zoetis today earns a seventy-one per cent gross margin and a twenty-one per cent return on invested capital, carries debt at two times earnings before interest and depreciation with interest covered fourteen times, generates a seven per cent free cash flow yield, pays a dividend covered nearly three times over, and sells medicine to people who love their animals in a world where more people acquire animals every year and spend more on them. Its international business grew six per cent last quarter while the American one fell eleven, which tells you the products are fine and the customer is stretched.
At thirty-five times earnings you needed the growth story to be right. At twelve and a half times you need it merely not to be catastrophically wrong. That is a far better bet, and it is the sort of bet I have generally been rewarded for making — not because I could see the recovery, but because I was not required to.
I would begin buying, and I would do it in thirds. A third here, because the price now pays you for the uncertainty rather than asking you to pay for optimism. A third if it reaches seventy dollars, which is below the twelve-month low and would mean the market has decided there is another leg down. And the last third only when organic revenue growth turns positive again and stays there for two quarters — not when it is forecast to, when it does.
The reason I would stage it rather than commit is simple and I want to say it plainly: nobody knows where the earnings floor is, including the company, which has now missed its own estimate of it twice. Anyone who tells you they have identified the bottom in a business whose management has cut guidance twice in fourteen weeks is guessing. Staging is what you do when you are confident about the value and honest about the timing.
A wonderful business, growing backwards, at a price that finally reflects it. I have paid a great deal more for a great deal less.