Bristol-Myers Squibb: A Deep Value-Investing Study

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

• QHow high is its market ceiling? Is it enlarging an existing slice of pie, or creating a brand-new market?4/10

  A

  Conclusion first: BMS's ceiling is real but limited. The vast majority of its revenue comes from grabbing a slice of an existing pie in "already-existing large disease areas" (anticoagulation, immuno-oncology, hematologic tumors, cardiomyopathy, schizophrenia), rather than creating a zero-to-one brand-new market. The only thing currently at "open-a-new-market" scale is Cobenfy's first non-D2-receptor mechanism in schizophrenia, and it is still in early ramp, far from a size that can carry the whole book. This is a large pharma that holds position with innovative molecules within mature demand, not a frontier-opening growth machine.

  First, the composition of the pie. In 2025 total company revenue was about $48.2 billion, of which the Growth Portfolio was about $26.4 billion, +17% year over year, and the Legacy Portfolio about $21.8 billion. Within the two pillars, Eliquis was about $14.4 billion globally and Opdivo about $10 billion, together accounting for half of total revenue. Eliquis makes substitution inroads in the atrial-fibrillation/venous-thrombosis anticoagulation market long cultivated by warfarin and dabigatran, while Opdivo splits the pie head-to-head with Merck's Keytruda in the mature PD-1 immuno-oncology race. Both are textbook cases of "enlarging and deepening an existing big pie," not pioneering new ground.

  Broken down by Baillie Gifford's "market ceiling" ruler, almost all of BMS's projects fall on the "enlarging an existing pie" side:

  • Anticoagulation (Eliquis): a fight for share in an existing market, and its share is being nibbled from the top by generics. The U.S. list price has begun to be cut, and the European apixaban compound patent/SPC is nearing expiry with early-launch generics already appearing. This ceiling is not un-opened; it is closing.
  • Immuno-oncology (Opdivo / Opdualag / Opdivo Qvantig): deep cultivation of the installed base in a mature red ocean via new indications, subcutaneous formulations, and combination therapies; the TAM is large but crowded, and share is suppressed by Keytruda.
  • Hematologic tumors (Reblozyl, Breyanzi): Reblozyl has passed $2 billion and Breyanzi $1 billion, penetrating existing anemia/lymphoma demand with better solutions.
  • Hypertrophic cardiomyopathy (Camzyos): this is the closest to "pioneering a long-underestimated disease area"; as the first cardiac myosin inhibitor, it did educate the oHCM niche; but after Cytokinetics' aficamten (MYQORZO) received FDA approval in December 2025, this just-opened small market must immediately be split between two players.

  The only thing that can truly be labeled "creating a new market" is Cobenfy: as the first muscarinic-agonist mechanism in decades for schizophrenia treatment that does not rely on D2-receptor blockade, if it succeeds in indications beyond schizophrenia such as Alzheimer's psychosis (AD psychosis), it could in theory re-enlarge a disease area that has long lacked a new mechanism. But this remains "imagination" rather than "already realized": its scale is still small, and the report itself lists Cobenfy AD psychosis as a key readout requiring continuous tracking, meaning this "new market" narrative is today far from landing at a degree that could change the height of the company's ceiling.

  So the honest conclusion: BMS's market ceiling is determined by three things—how much share it can take in several large disease areas, how fast that share decays, and whether new molecules pick up the baton in sufficient measure—rather than by whether it can conjure a new continent of demand out of thin air. Its growth ceiling is held down by two things: one, core revenue is a share business in existing markets, and the share of the two largest products (Eliquis/Opdivo) is entering a countdown (the 10-K discloses both have U.S. exclusivity at the earliest to 2028); two, the company's own 2026 full-year revenue guidance of about $46–47.5 billion is below 2025 actuals, showing that in the near term even "holding on to the existing pie" still needs new drugs to keep filling in, let alone raising the ceiling by opening new markets. For a framework like Baillie Gifford's that seeks "5x in ten years," this is a clearly point-losing question.

  Jun 10, 2026
• QOver the next five years, can its revenue at least double? Is growth driven mainly by volume, price, or new businesses?3/10

  A

  Conclusion first: almost impossible. Doubling revenue in five years (from about $48.2 billion to $96 billion) implies a ~15% compound annual growth rate, whereas BMS's own 2026 full-year guidance is about $46–47.5 billion, lower than 2025. For a mature large pharma whose near-term revenue is expected to decline, doubling revenue in five years would require a whole string of pipeline all beating expectations plus multiple large acquisitions, which is a low-probability event. Structurally, what can be counted on is "new-drug ramp (volume)" offsetting "old-drug generic collapse (a double hit of volume and price)," rather than a large expansion of the whole book.

  First, let's lay out the doubling arithmetic. 2025 total revenue was about $48.2 billion; doubling requires reaching ~$96 billion. But the midpoint of the company's 2026 guidance is about $46.7 billion, and the Legacy Portfolio is expected to decline 12%–16% year over year. When even the starting point is contracting, the bar for doubling in five years is not "optimistic," it is "unrealistic." The report's judgment on this matches mine: under the conservative/neutral scenarios, the first-ten-year growth rate is only 0%–3%.

  Breaking down the drivers: among volume, price, and new business, price is a negative contributor, volume is a structural give-and-take, and new business is the only positive but not enough:

  Price: clearly a subtraction. In the U.S., CMS's first-round Part D negotiated prices take effect on January 1, 2026, with Eliquis cut to $231, about a 56% reduction from its $521 list price; the company also disclosed in its Q1 2026 report that the U.S. Eliquis list-price cut affected net-price (GTN) adjustments. Net-price power is squeezed on three sides by the IRA, PBMs, and rebates, so no revenue increment can be had on this path.

  Volume: old-drug ramp offset by generic collapse. On one side, new-drug ramp is genuinely strong: the Growth Portfolio grew +17% in 2025 to about $26.4 billion, with Reblozyl passing $2 billion and Opdualag/Breyanzi/Camzyos each passing $1 billion. On the other side, old drugs fall off a cliff: Revlimid fell -63% year over year in Q1 2026 to $349 million (U.S. generics are no longer quota-limited), Pomalyst keeps being eroded, and the exclusivity of the larger Eliquis ($14.4 billion) and Opdivo ($10 billion) has also entered a visible countdown: BMS's 2025 10-K discloses that Eliquis and Opdivo both have U.S. exclusivity at the earliest to 2028. This means that over the next five years the Growth increment largely just fills the hole left by Legacy, rather than net growth.

  New business: the only positive engine, but its scale cannot support a doubling. Cobenfy (new schizophrenia mechanism), Camzyos (cardiomyopathy), cell therapy, and follow-on registration projects such as milvexian, iberdomide, and mezigdomide are the true growth sources. But these are either just ramping or still in the Phase III readout stage; even if all goes well, together they can hardly recreate an equivalent $48 billion book within five years. The report's optimistic scenario (5% growth over the first ten years) already implies "all new drugs pick up the baton smoothly," and 5% is still far below the 15% needed to double.

  The only path that mathematically approaches a doubling is "continuously buying revenue through large M&A," but that is precisely a point-loser rather than a point-gainer: in recent years BMS has paid out in succession for Karuna (about $14 billion), Mirati (about $4.8 billion), and RayzeBio (about $4.1 billion) (acquisition considerations per the report). Revenue stacked up through M&A both raises the bar of future returns and is not the "organic, compoundable" growth that Baillie Gifford admires.

  So the honest answer: on the five-year-doubling question, BMS basically fails. What it is more likely to show is a gear-shift pattern of "revenue oscillating in a narrow band between $46 billion and $50 billion, or even slipping slightly, with new-drug ramp barely offsetting old-drug collapse." The main growth driver is new-business volume, but it is severely offset by falling price and old-drug volume collapse. For a framework hunting 10x/5x growth stocks, this is an answer that lands squarely on the "failing" side.

  Jun 10, 2026
• QFive years from now, what will pick up the baton as the next growth engine? Does this "second curve" exist today?4/10

  A

  Conclusion first: the second curve does exist today, and there is more than one, but its nature is "shifting gears to survive" rather than "leaping upward": the mission of these new engines is to fill the hole left by the Eliquis/Opdivo decline, rather than to take the company to a higher tier. In other words, BMS's "second curve" is more like a spare tire that extends the life of the first curve; run well, it at best keeps revenue flat, and run poorly, it falls into a value trap. It is not, in Baillie Gifford's sense, the kind of disruptive new business that can independently remake the company.

  First, the substance of this second curve. The candidates to pick up the baton in five years are already on the table, in three tiers:

  Already ramping and verifiable (the most solid tier): the Growth Portfolio grew +17% in 2025 to about $26.4 billion, within which Reblozyl has passed $2 billion and Opdualag/Breyanzi/Camzyos each passed $1 billion. This tier is not a pie in the sky; it is hard-cash sales, and it is also where BMS's second curve is more solid than a pure cash-burning biotech.

  Just started, with large imaginative space but tiny scale: Cobenfy, the first non-D2-mechanism new drug for schizophrenia in decades, had 2025 full-year revenue of only about $155 million per BMS key facts, still far from "engine" scale; its real bet lies in whether new indications such as Alzheimer's psychosis (AD psychosis) can be made to work, and the report lists Cobenfy AD psychosis as one of the most critical readouts to track. There is also cell therapy (follow-on CAR-T beyond Breyanzi) and radioligand therapy (RayzeBio's RYZ101).

  Still in Phase III, pure option value: milvexian (a Factor XIa anticoagulant, the next generation benchmarked against Eliquis), iberdomide, mezigdomide (protein degraders taking over from Revlimid/Pomalyst), admilparant, and others. These have no revenue at all today; whether they can succeed depends on three gates: clinical trials, approval, and insurance access.

  So, "does it exist today"? It exists, and the tiers are clear. But to characterize it honestly, three point-losers keep this second curve from supporting a "leap upward" narrative:

  First, it is racing against the first curve rather than stacking on top of it. Over the same period, the Legacy Portfolio is expected to decline a further 12%–16% in 2026, and Revlimid was already -63% in Q1 to $349 million. The new engine's increment is first eaten into by the old engine's shrinkage; the net effect is exactly why the company's 2026 guidance of $46–47.5 billion is below 2025.

  Second, the most imaginative one (Cobenfy's new mechanism) is the smallest in scale and lowest in certainty, while the large ones (the Opdivo franchise, hematologic tumors) are essentially still working the installed base in a mature red ocean, hard-pressed to provide explosive secondary growth. In other words, "the big ones aren't sexy, and the sexy ones aren't big."

  Third, a substantial part of this second curve was bought rather than grown. Cobenfy came from the Karuna acquisition of about $14 billion, RayzeBio about $4.1 billion, Mirati about $4.8 billion (considerations per the report). Buying pipeline is not itself wrong, but it means the second curve's return bar has been raised by acquisition cost, and it depends on whether management can keep buying the right assets, and not too dear, next time.

  Overall: BMS's second curve is clearly visible today and backed by real revenue, which is better than many growth stocks that only have a slide deck; but its destiny is to "catch the decline and keep things flat," not to "double the company's size again." For Baillie Gifford's ruler seeking "full firepower from years three to ten, 5x in ten years," a second curve whose main tone is survival and whose increment is largely offset by old-drug shrinkage can only count as "exists but not strong enough." This is the fundamental reason this question gets a neutral-to-weak rating.

  Jun 10, 2026
• QWhat is its core competitive advantage? Will this moat widen or narrow over the next three to five years?5/10

  A

  Conclusion first: BMS's core moat is the three-piece set of "patents + regulatory barriers + global commercialization scale," the hardest being the legal moat of patents and FDA access. But it carries an inherent countdown attribute, and over the next three to five years it is most likely "stable-to-narrowing," rather than widening. The reason is not that the company has gotten worse; it is that the patent clock on the two most profitable products (Eliquis, Opdivo) is running toward zero, and while the new moat (Cobenfy, Camzyos, cell therapy) is being rebuilt, it is not yet thick enough to fill the gap left when the old wall collapses. This is a clearly point-losing question.

  First, what the moat is. BMS's competitive advantage does not rest on brand mindshare (prescription drugs lack the consumer loyalty of a Coke), on cost (the unit cost of a pill is irrelevant), or on network effects (drugs are not a social platform); it rests on three layers:

  • Patents and regulatory barriers (the most core, strong): a new molecule takes ten years and billions of dollars from discovery to FDA approval, and generics cannot enter during the exclusivity window. This is BMS's most real source of pricing power.
  • Global commercialization scale (strong): global market access, insurance negotiation, hospital and pharmacy channels, and the clinical-trial network cannot be replicated by small and mid-size companies. The reason Eliquis can reach $14.4 billion is half the quality of the molecule and half this global distribution machine.
  • Clinical data and indication expansion (medium): real-world data and continual indication expansion can extend product life cycles, but they do not constitute a non-replicable long-term barrier.

  The key judgment lies in "whether it widens or narrows over the next three to five years," and my answer is narrowing, with two-way evidence:

  The old moat is thinning (the main point-losing line). BMS's 2025 10-K discloses that Eliquis and Opdivo both have U.S. exclusivity at the earliest to 2028, while the European apixaban compound patent/SPC for Eliquis runs out earlier, in November 2026, and early-launch generics and IP litigation have already appeared in Europe. This is what is most distinctive about the pharma moat: replicating the whole of BMS is hard (years plus billions), but replicating the profit erosion of a single product is fast; once the patent window opens, generics/biosimilars can overwhelm the profit pool within a few quarters. Revlimid's -63% in Q1 is a live specimen. Worse, the moat has also been chiseled from the top by policy: CMS cut Eliquis's negotiated price to $231, about a 56% reduction from $521, effective January 1, 2026. Even within the exclusivity window, net-price power has been compressed by the government.

  The new moat is being rebuilt (but not yet thick enough). Camzyos has a U.S. patent term to 2036 and SPCs to 2038 in several European countries, a new wall with a long enough time span; the exclusivity of Sotyktu, Breyanzi, and Reblozyl also runs to around 2031–2036 (per the report's 10-K). The problem is: first, each of these new walls is far smaller in scale than Eliquis/Opdivo and cannot fill the gap; second, even a just-erected wall immediately has someone come to tear it down: in December 2025, Cytokinetics' aficamten (MYQORZO) received FDA approval, becoming a direct same-class competitor to Camzyos in hypertrophic cardiomyopathy, showing that even a new growth asset has no worry-free moat.

  By cross-comparison, in the main immuno-oncology race BMS's Opdivo has long been suppressed by Merck's Keytruda, which itself shows that on the largest mature battlefield, BMS is not the player with the widest moat.

  Historical metrics can only prove "how profitable the past moat was" (72% gross margin, 31.6% operating margin), but a moat is forward-looking: the marginal change looking 4–8 quarters ahead is "the old wall peeling off faster, another policy chisel, the new wall being patched but thin, and the new wall also facing competitors crashing in." So the honest conclusion: BMS has a real moat, and the legal-barrier layer is very hard, but it sits in a "net-narrowing" zone. On Baillie Gifford's core growth question of "the moat should widen over the next three to five years," BMS gives an answer in the opposite direction. This is the key to why it overall counts only as "decent quality, questionable growth" rather than a "great growth stock."

  Jun 10, 2026
• QIf its core business is disrupted, does it have the gene to reinvent itself? How does it treat mistakes and bad news?5/10

  A

  Conclusion first: BMS has the ability to "reinvent itself," but this reinvention is an institutionalized, money-bought, industrial-grade blood transfusion, rather than the in-the-genes disruptive impulse of a founder-type company. It has indeed replaced and rebuilt its core products multiple times in the past (from Plavix to Opdivo, from the Celgene legacy to Cobenfy), proving it will not be killed by the collapse of a single product; but its reinvention relies heavily on the two life-extension pipes of "R&D pipeline + large M&A," and each transfusion comes at the cost of raising the bar of future returns. On its attitude toward bad news, it is quite candid in written disclosure, but on "proactively pricing in bad news" it is passively reactive, rather than proactively self-revolutionizing.

  First, whether the implicit premise of a "self-reinvention gene" holds at all. Judging whether a company can survive when its core business is disrupted comes down to whether it has genuinely replaced its core engine in the past. On this point BMS passes: it is long past being a "few old Plavix/Celgene blockbusters" company. In 2025, the Growth Portfolio already made up about 55% of total revenue (about $26.4 billion, +17%), meaning the company's main body of revenue has been relay-replaced by generation after generation of new molecules. Being able to repeatedly complete the "old-drug collapse → new-drug replacement" gear shift is itself a kind of industrialized self-reinvention capability, stronger than the small pharma that cannot carry on once a core patent expires.

  But to characterize this reinvention honestly, there are three clear qualifications:

  First, its reinvention is "bought," not "grown." Cobenfy came from the Karuna acquisition of about $14 billion, and for accounting reasons 2024 recognized about $12.1 billion of Acquired IPRD expense, RayzeBio about $4.1 billion, Mirati about $4.8 billion (considerations per the report). This shows that when its core business is threatened, BMS's instinctive reaction is to "pull out cash and buy a new life outside," rather than incubating a disruptor internally. Buy right and the life-extension succeeds; buy too dear or buy wrong and it is a double blow. It is a kind of reinvention that is effective but expensive, and that stakes success or failure on management's M&A eye.

  Second, it is passively reactive, not proactively self-disrupting. What Baillie Gifford values most is the "revolutionize yourself" gene (proactively incubating a replacement while the core business is still making money). BMS's gear shift is a defensive action forced by the patent cliff: only when the old drug is about to expire is it compelled to find a new drug to take over, rather than proactively cultivating something to disrupt Eliquis while Eliquis is at its peak. This kind of reinvention can avoid sudden death, but it can hardly bring a leap.

  Now to "how it treats mistakes and bad news": this is the key to judging the honesty of the reinvention, and BMS's performance is "candid in writing, transparent in framing, but passive in posture":

  • Positive: in the 10-K the company writes clearly about the patent cliff, IRA government negotiation, generic erosion, and future price-renegotiation risk, without dodging. On accounting, it gives relatively full disclosure of large non-cash items, CVR fair value, and IPRD impairment. In 2024 the GAAP loss of about $8.948 billion (mainly from M&A-related IPRD and intangible-asset amortization) was also laid out faithfully, while operating cash flow still reached about $15.190 billion, with no cosmetic cover-up of the loss's source.
  • Reservation: but "candidly acknowledging risk" and "proactively pricing in bad news, self-revolutionizing in advance" are two different things. BMS only responds passively after generics are already at the gates and insurers have already started cutting prices, rather than restructuring the portfolio into place years in advance. The report gives management candor "medium-to-high" and capital allocation only "rational but not exceptional," consistent with this.

  By cross-comparison, this "life-extension via M&A" reinvention path also has a side effect: it makes the GAAP statements harder and harder to read (heavy IPRD, CVR, intangible amortization), so external investors find it harder to judge true earning power. One cost of the reinvention is that transparency is diluted by acquisition accounting.

  Overall: BMS will not fall just because Eliquis or Opdivo expires; it has a proven, industrialized transfusion mechanism and does not hide bad news, which lets it pass on the "survive" dimension. But its reinvention is defensive, bought, and passive, lacking the founder-style gene Baillie Gifford prizes: "proactive self-disruption, willing to revolutionize itself for the long term." On this question it can score a medium, but not a high mark.

  Jun 10, 2026
• QDoes management (especially the founder) have long-term vision, with interests deeply tied to the company? Is it willing to sacrifice current profit for five to ten years out?4/10

  A

  Conclusion first: BMS's governance structure is compliant, professional, and disciplined, but it is fundamentally not the "deep founder alignment" type of company Baillie Gifford prefers: this is a century-old operation run by professional managers, with no founder, and the ownership stakes of the CEO and executives are extremely low, with interest alignment relying mainly on the compensation system rather than personal net worth. It is willing to bet for the long term (continuously pouring into R&D, paying enormous M&A fees to buy future pipeline), but this long-term investment is more a survival response forced out by the patent cliff, rather than the proactive foresight of "willingly sacrificing current profit for five to ten years out." On this question centered on "founder long-term vision + interest alignment," BMS structurally lands on the weak side.

  First, let's be clear about the "founder alignment" premise: BMS has no founder. It is a mature giant formed from the 1989 merger of Bristol-Myers and Squibb, and the current CEO, Christopher Boerner, is a professional manager. So the structure Baillie Gifford values most, "a founder staking his own net worth on the company and running it with an owner's mindset for decades," is naturally absent at BMS.

  Just how weak is the alignment? Look at the holdings: citing the proxy statement, CEO Boerner held only about 146,890 shares as of March 2026, neither any individual nor management and the board combined holds more than 1% of outstanding shares, and third-party insider ownership is about 0.05%. Against the company's roughly 2.04 billion shares outstanding and about $114 billion market cap, executive holdings amount to a drop in the bucket. This means management and shareholders are "not unaligned, but certainly not of an owner's mindset": their wealth comes mainly from the pay package, rather than equity that puts them in the same boat as minority shareholders.

  The alignment rests on institutions rather than personal net worth, and on this side it is up to standard: the 2026 proxy statement shows 10 of 11 director nominees are independent, and the audit/compensation/governance committees are all independent directors; the CEO must meet a 6x-salary ownership requirement and other core executives 3x, all compliant in 2025, with clawback and trade pre-clearance systems in place. This is the standard governance setup of a large-cap blue chip: it can prevent obvious agent overreach, but "compliant" does not equal "owner's mindset"; the ownership threshold is a required floor, rather than a signal of proactive heavy buying.

  Now to "whether it is willing to sacrifice current profit for five to ten years out": this is the substance of the question, and BMS's answer is "willing to invest, but the motive is defense rather than foresight":

  • Willing to spend for the long term, and the evidence is real: the company pours about $9.951 billion into R&D each year, and paid out in succession for the successor pipeline: Karuna about $14 billion, with about $12.1 billion of Acquired IPRD expense recognized in 2024. This expense directly turned 2024 into a huge GAAP loss, and management did it anyway, knowing it would look bad, which shows it truly staked money on the future rather than dressing up current profit. This deserves credit.
  • But this is survival forced by the patent cliff, not composed foresight: BMS spends to buy pipeline essentially because Eliquis/Opdivo are set to expire around 2028 and the whole thing collapses if the holes are not filled. This is a different nature from what Baillie Gifford has in mind: "the core business still at its peak, yet proactively sacrificing the present to cultivate a disruptor ten years out."

  On capital-allocation discipline, BMS by contrast has a side worth praising, distinct from the "serve the short-term stock price" bad archetype: it did not dress up EPS through aggressive buybacks; instead it put its focus on deleveraging: in 2025 it repurchased/redeemed about $8.739 billion of principal debt to strengthen the balance sheet (per the report), while maintaining a dividend of about 4.5%. The report gives management candor "medium-to-high" and capital allocation "rational but not exceptional," matching my view: it does not run wild, but it also does not amount to a textbook-excellent capital allocator that continuously creates excess returns, and "needing M&A to extend the future's life" is itself a nagging concern about the return bar.

  Overall: BMS's management is trustworthy, governance is compliant, dividends are reliable, and the deleveraging direction is right; as a blue chip, its management is "entrustable." But it has no founder, interest alignment is extremely weak (insider about 0.05%), and the base tone of its long-term investment is defensive survival rather than proactive foresight. Measured by Baillie Gifford's ruler of "founder long-term vision + deep interest alignment + willingness to sacrifice the present for the long term," BMS is clearly weak on the alignment item, one of the dimensions that drags down the growth score across the whole piece.

  Jun 10, 2026
• QIf it disappeared tomorrow, how much would customers miss it? Is its way of growing sustainable, not reliant on harming society and regulation?5/10

  A

  Conclusion first: look at it in two parts. "Would customers miss it": yes, and at an essential-need level, because BMS holds multiple life-saving drugs in anticoagulation, immuno-oncology, hematologic tumors, and hypertrophic cardiomyopathy, where a patient stopping the drug pays a cost in health or even life; this layer of indispensability is hard. "Is growth sustainable and non-harmful to society and regulation": basically sustainable, with positive social utility (it sells drugs that genuinely treat disease), but its profit model has a structural tension with the public-payer system, and its pricing is being pushed back by regulation. It is "society (the government payer) thinks it earns too much and is starting to cut," rather than "earning money by harming society." Overall this is a question of "strong indispensability, but regulatory sustainability under pressure," a mixed bag for Baillie Gifford.

  First, indispensability. If BMS disappeared tomorrow, who would miss it most? Doctors, hospitals, payers, and patients, not ordinary consumers; and the missing is highly sticky:

  • Patient level (the hardest tier): Eliquis's $14.4 billion globally corresponds to stroke prevention for millions of atrial-fibrillation/thrombosis patients, Opdivo's $10 billion is the immunotherapy backbone for a large number of cancer patients, Camzyos is a breakthrough therapy for hypertrophic cardiomyopathy, and Cobenfy provides schizophrenia patients with the first new mechanism in decades. Once these drugs are gone, for patients it is not as simple as "switching brands"; it means a treatment interruption and health harm. This essential need, where "stopping the drug has real medical consequences," is an indispensability that consumer-goods companies cannot provide.
  • But with an important discount: a drug's indispensability is at the "molecule level," not the "company level." What patients cannot do without is the apixaban molecule and the nivolumab molecule, rather than the company BMS: once the patent expires, generics/biosimilars can provide the same treatment, and early-launch generics of Eliquis have already appeared in Europe. So "customers cannot do without BMS's drugs" holds within the exclusivity window; once exclusivity passes, "what cannot be done without is the ingredient, not BMS." This is precisely the other side of its countdown-bearing moat.

  Now to "whether growth is sustainable and whether it relies on harming society and regulation": this is the more discerning half of the question:

  The social-utility direction is positive; it does not earn by harming society. BMS sells drugs that are rigorously clinically validated and FDA-approved, life-saving or quality-of-life-improving; what it creates is real medical value, not gambling, not induced consumption, not environmental pollution. On "whether this business's existence is a net good or net bad for society," it is a net good.

  But the profit model has a structural tension with the public-payer system, and regulation is pushing back. This is where the most honesty is needed: a large chunk of pharma's high profit (BMS's gross margin about 72%, operating margin about 31.6%) comes from high pricing to payers (Medicare, commercial insurance) within the exclusivity window, and that money ultimately comes from taxpayers and the insured. Society's pushback has already been institutionalized: CMS's first-round negotiation cut Eliquis to $231, about a 56% reduction from its $521 list price, effective January 1, 2026, and each subsequent year brings more drugs into the fold. In other words, BMS's growth is not "by harming society," but its high margin has been judged by society, through the government-negotiation mechanism, as "needing to give some back," which will keep compressing the profit pool: the regulation is not out to destroy it, it is out to make it earn less. This is the biggest external constraint on its growth sustainability.

  There is another layer of channel-concentration fragility: the company discloses that the three big wholesalers account for about 87% of its U.S. gross revenue (per the report); end demand is dispersed, but upstream receivables concentration is high. This does not count as "harming society," but it is indeed a concentration risk in the growth chain.

  Overall: BMS passes the "can customers do without it" gate: its drugs are essential, stopping them has real medical consequences, social utility is positive, and it does not earn by harming society, which is a solid plus. But it does not pass the full test of the "regulatory sustainability" gate: the high-margin model is being systematically pushed back by government negotiation, and what patients truly cannot do without is the molecule rather than the company, with stickiness going to zero once exclusivity passes. So on this question it is "strong indispensability (molecule level, limited to the exclusivity window) + positive social utility + but regulation imposes substantial suppression on its profit sustainability." A mixed bag, which precisely reflects BMS's overall picture of "a genuine good-business character, but doubly constrained by policy and the patent clock."

  Jun 10, 2026
• QWhat are this business's unit economics (gross margin, incremental returns)? Do they get better or worse as it scales? Where does the money it earns go?7/10

  A

  Conclusion first: BMS's unit economics are very pretty at the "money-making efficiency" layer: about 72% gross margin, about 31.6% operating margin, and about 24.6% FCF margin, a classic high-gross-margin, capex-light, strong-cash-flow model. But it has a fatal asymmetry: scaling up does not necessarily make unit economics better, because its real "reinvestment tax" is a never-fillable R&D-and-M&A black hole, rather than plants; and the incremental returns of its most profitable installed-base products are being eroded by genericization and insurance price cuts. The money it earns goes mainly to three places: R&D, buying pipeline via M&A, and debt repayment plus dividends. Overall this is a question of "excellent unit economics but uncertain incremental returns."

  First, unit economics themselves, a layer that is BMS's real strength:

  • Gross margin: on a trailing basis, gross margin is about 72.0%. Each additional box of drug sold has extremely low marginal cost, so gross profit is almost pure increment.
  • Operating margin: about 31.6%; even after heavy R&D and SG&A, about 30% still drops to operating profit.
  • Extremely capex-light: 2025 capex was only about $1.311 billion, against about $48.2 billion of revenue, less than 3% of revenue. This means growth barely relies on spending on plants and equipment.
  • Strong cash conversion: 2025 operating cash flow was about $14.156 billion and reported free cash flow about $12.8 billion, an FCF margin of about 24.6%.

  Looking at this set of numbers alone, the unit economics are first-class. But the follow-up "do they get better or worse as it scales" is precisely BMS's soft spot, to be honestly unpacked:

  First, its "incremental return" depends on whether the next molecule works, not on scale effects. For an ordinary manufacturing/software company, the larger the scale, the more fixed costs are spread and the better unit economics get. But the bulk of BMS's cost is R&D of about $9.951 billion and M&A: this is a "perpetual reinvestment tax." Every molecule has a patent countdown, and the company must keep spending on R&D plus buying pipeline to replace expiring products, or revenue collapses. So no matter how large it gets, it cannot stop this spending, and unit economics will not naturally improve just because it "got bigger"; what truly determines incremental return is the success rate of R&D/M&A, rather than production volume.

  Second, the marginal returns of installed-base products are deteriorating. The most profitable Eliquis was cut about 56% by CMS, and Revlimid was -63% in Q1: these are the company's highest-margin mature products, and their shrinkage will drag down overall marginal economics. This is also why the company's 2026 gross-margin guidance drops to about 69%–70%, below the current ~72%: scale has not shrunk, but the portfolio gear-shift is lowering unit profitability.

  Third, reported FCF cannot all be counted as incremental return. The report's reminder is right: much of BMS's out-licensing, asset acquisition, and BD spending lands in investing activities or Acquired IPRD rather than capex. So within the roughly $12.8 billion of reported FCF, what can truly be "painlessly distributed" must be discounted: the report's conservative Owner Earnings range is $10.5–11.5 billion, midpoint about $11 billion, after deducting the quasi-maintenance BD investment needed to sustain competitive position. This is exactly where unit economics "look better than they actually are and must be discounted."

  "Where does the money it earns go": three destinations, structurally clear:

  1. R&D (about $9.951 billion/year): the first line of defense against the patent cliff.
  2. Buying pipeline via M&A: Karuna about $14 billion, Mirati about $4.8 billion, RayzeBio about $4.1 billion (considerations per the report): this is "buying the future," but it also raises the return bar.
  3. Debt repayment plus shareholder returns: in 2025 it repurchased/redeemed about $8.739 billion of principal debt (per the report) to strengthen the balance sheet, while maintaining a dividend of about 4.5% and keeping about $5 billion of buyback authorization. This part reflects capital discipline: no dressing up EPS through aggressive buybacks.

  Overall: BMS's unit economics are "statically" excellent (high gross margin, capex-light, strong cash flow), a real plus; but it has a hard flaw "dynamically": scaling up brings no unit-economics improvement, the marginal returns of the most profitable products are being eroded by price cuts and genericization, and reported FCF must be discounted for perpetual BD investment. Most of the money earned is drawn off by the two "life-extension pipes" of R&D and M&A, leaving shareholders with a steady but non-fast-growing return flow. Against Baillie Gifford's ideal model of "the larger the scale, the better the unit economics, with increasing incremental returns," BMS is a counter-example. It is a "high-quality but incremental-return-diminishing under pressure" business.

  Jun 10, 2026
• QFor it to rise five-fold in ten years, which conditions must all hold at once? Are these conditions realistic? What expectation does today's share price imply?3/10

  A

  Conclusion first: for BMS to rise five-fold in ten years, a string of high-difficulty conditions must hold "simultaneously," and these conditions also contradict one another: this is almost an impossible mission. Rising five-fold from about $56, about $114 billion market cap means the share price reaching about $280 and market cap about $570 billion; working backward from a reasonable pharma P/FCF of 12–15x, free cash flow would have to rise from today's about $11.9 billion to about $45–50 billion, i.e., a ~15% FCF compound growth rate over ten years. For a mature large pharma whose near-term revenue is still declining, this is unrealistic. What today's share price implies is exactly the opposite: a low-expectation value-stock pricing of "valuation recovery + slow growth + high dividend," rather than a high-growth expectation.

  First, let's lay out "which conditions must all hold for 5x in ten years" one by one, tagging their realism:

  1. The shrinkage of the Eliquis/Opdivo profit pool must be significantly slower than expected. But the fact is both expire from U.S. exclusivity in 2028, Eliquis already faces generics in Europe, and CMS has already cut Eliquis's price by about 56%. This is a contract/legally-fixed countdown, not a "might be a bit slower" probabilistic event. Realism: low.

  2. The new product group must not just fill holes but net-grow a re-created book. The Growth Portfolio is now about $26.4 billion, +17%, but most of its current increment is eaten up by Legacy shrinkage, and the company's 2026 guidance is $46–47.5 billion, below 2025. For a 5x, Cobenfy, Camzyos, cell therapy, milvexian, iberdomide, mezigdomide, and nearly all the rest would have to become big blockbusters and not be split by competitors (yet Camzyos already faces head-to-head competition from aficamten). Passing all three gates of clinical, approval, and insurance access at once, realism: very low.

  3. Net-price power is not further eroded by policy. But IRA negotiation brings more drugs in each year, and net price under long-term pressure is a structural direction. Realism: low.

  4. Management does not make a single value-destroying large acquisition in ten years, and every pipeline it buys delivers. BMS's growth depends heavily on M&A such as Karuna at about $14 billion to extend its life, and zero M&A misses for ten straight years is an extremely high bar. Realism: low.

  5. The market must deliver a large valuation re-rating (P/FCF rising from about 9.6x to a growth stock's 20x-plus). But a re-rating happens only when the first four conditions hold and the company truly becomes a growth stock. Realism: contingent on the first four.

  The key contradiction is: these conditions are not only individually low-probability, they also fight each other. Condition 1 needs old drugs to hold steady, condition 2 needs new drugs to explode: but BMS's cash is limited, and pouring into R&D plus M&A to bet on condition 2 makes it harder to simultaneously maintain the defensive investment condition 1 requires; while policy (condition 3) is an exogenous variable the company cannot control. The report's optimistic scenario (5% growth over the first ten years, corresponding to intrinsic value of $90–105) already implies "all new drugs pick up the baton smoothly + no big mistakes," and 5% growth corresponds to only about one-fold of upside, an order of magnitude away from 5x (which needs ~15% FCF growth). So the honest conclusion: the conditions needed for 5x in ten years are unrealistic; this is not a reasonable expectation for a mature large pharma like BMS.

  Now to "what expectation today's share price implies": this is the most valuable part of the question, and the answer is "the market is not pricing in growth for it at all":

  • BMS's current price corresponds to about 9.6x P/FCF, about 7.9x EV/EBITDA, about 9.2x forward P/E, and about a 4.5% dividend yield. This is classic "low-expectation value stock" pricing, not "high-growth" pricing. Against Merck's EV/EBITDA of about 11.0 and AbbVie's about 15.9, the market is clearly using a lower multiple to discount BMS's "patent cliff + execution risk."
  • In other words, the expectation today's share price implies is: old drugs decline, new drugs offset only part, Owner Earnings stay roughly flat or slowly decline, and valuation gets no premium for the long haul. The report's neutral scenario (3% growth, intrinsic value $60–76) roughly matches the current about $56: the market pricing sits at "the lower bound of fair value to the low-to-mid end."
  • This precisely shows: BMS's investment logic is "betting the market overdid the transition-period discount, and future single-digit-to-low-double-digit annualized returns come from dividend + slow growth + partial valuation recovery," rather than "betting it rises five-fold." The report's expected annualized return of about 8%–10% neutral and about 11%–13% optimistic is consistent with this.

  Overall: for BMS to 5x in ten years, five low-probability, mutually contradictory conditions must hold at once, which is unrealistic. It is not a "5x in ten years" candidate in Baillie Gifford's sense. And today's about $56 share price also honestly reflects this: the market gives it a value stock's low multiple, not a growth stock's high expectation. The real dispute is not "can it 5x" (basically no), but "is this low valuation an opportunity or a value trap": if new drugs fail to pick up the baton, about 9.6x P/FCF could be merely the reasonable price of a continually shrinking cash flow. This is the fundamental reason this growth question gets a low score.

  Jun 10, 2026
• QWhy has the market not yet realized all this? Is it that it can't understand, looks down on it, or can't see far? What would become the "narrative inflection point"?3/10

  A

  Conclusion first: on this question, BMS first requires an honest reversal: it is a stock that "the market has already fully realized the problem with and discounted accordingly," rather than an "underappreciated treasure the market has not realized." The low multiple of about 9.2x forward P/E, 9.6x P/FCF, and a 4.5% dividend yield is not the market failing to understand; it is a discount the market proactively assigns after understanding the risk set of "patent cliff + net-price cuts + M&A life-extension." So the real question is "whether this discount is overdone (opportunity) or just right or even insufficient (value trap)," rather than "why the market has not yet realized its good." Among the three explanations, BMS's situation is mainly "can't see far (unable to price the long-dated uncertainty of pipeline delivery)," rather than "can't understand" or "looks down on it."

  First, correct the premise: Baillie Gifford's classic framing of this question (market can't understand / looks down / can't see far) defaults to the stock being undervalued. But BMS's market pricing has no obvious "un-realized" element: it is a large-cap blue chip, well-covered, and information-transparent, and the valuation discount is the market's rational result based on known risks. So we must honestly examine the three possibilities one by one:

  "Can't understand": basically does not hold. BMS is a century-old blue chip thoroughly researched by sell-side and institutions, and the 10-K spells out the patent cliff, IRA, generic erosion, and price renegotiation in full, so there is no real information asymmetry. The only slightly "hard-to-understand" thing is its accounting: heavy Acquired IPRD, CVR fair value, and intangible amortization distort the GAAP figures (in 2024, a huge GAAP loss of about $8.948 billion, yet operating cash flow reached about $15.190 billion). Those who look only at P/E will misjudge it as "now expensive, now loss-making," but any professional investor who looks a bit at cash flow understands, so this does not constitute a systematic "can't understand."

  "Looks down on it": partly holds, but for legitimate reasons. In a market where "the growth narrative is king," capital is more willing to award a high premium to pharma with a clear growth story: Merck's EV/EBITDA is about 11.0, AbbVie's about 15.9, while BMS is only about 7.9. The market indeed "looks down on" BMS's transition period, but this disdain is grounded: the company's 2026 guidance has revenue declining further to $46–47.5 billion, and the most profitable Eliquis was cut about 56%. This is not irrational prejudice; it is pricing of a real negative.

  "Can't see far": the most fitting explanation. The market's core disagreement on BMS lies in "whether the unknown good can be delivered," not in the known bad (the patent cliff is common knowledge): whether Cobenfy works in AD psychosis, whether Phase III projects like milvexian/iberdomide/mezigdomide can form large enough sales replacement, and whether the new moat can fill the gap in the old wall. These are long-dated variables that will not be settled until 5–10 years out; the market naturally cannot give them high-certainty pricing, so it uses a low multiple to fully discount "execution risk." The report listing this long string of key readouts as items to track is in essence an admission: the flip side of the low valuation is that the market is unwilling to pay for unproven pipeline delivery.

  But here we must add a Baillie-Gifford-style honest cut: "can't see far" could be an opportunity, or the market could be quite right. Cheap for over two years and barely rising in a broad rally: by value-investing discipline, the default is that it is more likely a "structural problem" rather than a "temporary mispricing"; BMS's discount has persisted, and it faces structural patent expiry and policy price cuts (certain negatives), rather than one-off event-type negatives. The "value trap" risk the report repeatedly stresses lies precisely here: if new drugs fail to pick up the baton, about 9.6x P/FCF is not undervaluation; it is the reasonable price of a continually shrinking cash flow, and the share price could even fall toward $35–40, corresponding to a 35%–45% permanent loss (per the report). So the "market hasn't realized it" premise must be used cautiously for BMS: more likely the market has realized it, and realized it correctly.

  What would become the "narrative inflection point"? The inflection point will not come from valuation itself, only from fundamentals falsifying one side:

  • Upside inflection (discount recovered): ① the Growth Portfolio's growth rate continuously covers the Legacy decline and revenue returns to positive growth, proving "the gear-shift succeeded"; ② Cobenfy gets strongly positive Phase III data in new indications such as AD psychosis, opening a large new-mechanism market; ③ next-generation products like milvexian get approved and ramp, proving the pipeline connects; ④ Owner Earnings stabilize above $11 billion and stop shrinking. If any one is realized, the market may repair the "patent-cliff discount" toward a "normal value-stock multiple."
  • Downside inflection (trap confirmed): ① Eliquis/Opdivo net price falls faster than expected; ② key Phase III trials fail in succession; ③ management does another high-priced, low-return large acquisition; ④ Owner Earnings fall below $9 billion for two consecutive years. If any one occurs, the market will re-price it as a "chronically declining large pharma," at a lower multiple of 7–8x Owner Earnings.

  Overall: BMS is a stock that "the market has seen the risk of and discounted with a low multiple," rather than "a cheap item the market has not seen," the most important element being "can't see far": unable to price long-dated pipeline delivery. The narrative inflection point hinges on the empirical results of the old-to-new product handover, rather than sentiment. On Baillie Gifford's capstone question of seeking "a great growth stock the market has not yet realized," BMS's answer is precisely a counter-example: there is no obvious cognitive-gap bonus to be had here, and low valuation and high risk are two sides of the same coin. This also echoes the report's overall tone of "Watch, insufficient margin of safety for the conservative."

  Jun 10, 2026