Bristol-Myers Squibb Company BMY

Key data

The business

BMS runs as a single segment: it discovers, makes, and sells prescription medicines. Two engines drive it. The legacy blockbusters, Eliquis (a blood thinner whose profits are shared with Pfizer), Opdivo (a cancer immunotherapy), and the fading Revlimid, still throw off most of the cash. The growth portfolio, Reblozyl for anemia, Camzyos for a heart condition, Breyanzi cell therapy, Sotyktu for psoriasis, and Cobenfy, the first genuinely new way to treat schizophrenia in decades, bought for $14 billion through Karuna in 2024, is what has to carry the next decade. The moat is patents plus the billion-dollar, ten-year trial machine that refills them. The concrete thing a patient holds: a Cobenfy capsule taken twice a day, one that does not blunt dopamine the way every antipsychotic since the 1950s has. Three wholesalers move most of the U.S. volume, and no single country outside the U.S. is more than 10% of revenue.

The numbers

The last five quarters show a business holding its top line and cleaning up its bottom line. Revenue rose 5.7% year over year in Q2 to a record $13.0 billion, and the first half grew 4.2%.

Read the quarters with care. Q2 net income of $3.3 billion more than doubled the $1.3 billion a year earlier, but most of that jump is accounting, not demand: last year's quarter carried a $1.5 billion acquired-research charge and $830 million of drug amortization, versus zero and $437 million this year. Underlying sales grew a single-digit 5.7%. On the adjusted numbers analysts track, BMS has now cleared estimates four quarters running, most recently $2.04 against $1.60. The multiple that looked cheap for a reason a quarter ago is still about 15 times even after the stock ran to a fresh 52-week high, so the market re-rated the earnings, not the story.

Across full years the shape is a plateau: revenue has sat near $48 billion since 2021, the 2024 loss was the Karuna writedown running through the income statement, and 2025 normalized to $3.46. R&D is 22% of revenue, a real number for a company whose survival depends on the pipeline. The variant is simple: the market prices BMS as a melting ice cube, with consensus revenue sliding from about $48 billion now toward $38 billion by 2029 as the patents lapse. What it does not fully price is the growth portfolio reaching escape velocity first. The print that settles it is the quarterly run-rate of the new drugs against the decline of the old.

Management

The capital story is deleveraging, not buybacks. After spending big on Karuna, Mirati, and RayzeBio, BMS repurchased zero stock in both 2024 and 2025, having bought back $6.3 billion, $8.0 billion, and $5.2 billion in the three years before. It is paying down debt instead: long-term debt fell from $49.5 billion in early 2025 to $42.9 billion, and net debt dropped $2.4 billion in the first half alone, funded by $4.5 billion of operating cash flow against $2.6 billion of dividends. Net debt sits at about 2.1 times EBITDA. Insider activity is quiet and one-directional: no buys, and three small sales this year, the largest by David Elkins totaling about $1.85 million on April 1, plan status not disclosed. The four straight adjusted beats suggest guidance set to be cleared, not stretched.

How it fails or surprises you

The 2028 cliff. Eliquis and Opdivo, together a large share of profit, lose U.S. exclusivity around 2028. Consensus already has revenue falling from about $42 billion in 2028 to $38 billion in 2029. If generics arrive earlier or bite harder, that hole widens fast. Watch the first confirmed generic Eliquis entry date and the 2027 guide.

Cobenfy breaks out (right tail). Cobenfy is approved for schizophrenia today, but trials are testing it in Alzheimer's-related psychosis and as an add-on therapy. Land either and a roughly $2 billion drug becomes a multi-billion franchise the 15x multiple does not pay for. Watch the quarterly Cobenfy run-rate and the adjunctive and Alzheimer's readouts.

The doubling was mostly accounting. The fact this read leans on least: Q2 profit doubled while sales grew 5.7%. Strip the lower charges and the growth portfolio still has to out-run the legacy decline in dollars. Watch the crossover quarter when new-drug revenue exceeds the drop in the old.

Closing thoughts

This is a stock the market has already judged, and mostly judged fairly. At the 8th percentile of its own twelve-year multiple, with consensus revenue openly modeled to fall through the cliff, the cheapness is not a secret waiting to be found, it is the price of a known 2028 problem. The people on the other side are holders who will not pay up until the newer drugs prove in dollars that they can replace Eliquis and Opdivo, and that proof is quarters away, not here. The fatter tail is probably still the downside, because 2028 is close and the growth portfolio is not yet large enough to cover the loss. But this is a re-rating question, not a survival one: the dividend is covered, debt is falling, and cash flow is heavy. Nothing here threatens the company's existence, only its multiple.

The bet is still that the newer medicines, Cobenfy and the cancer and blood drugs behind it, get big before Eliquis and Opdivo lose their patents around 2028. What breaks it is simple and datable: the generics arrive on schedule and the new portfolio is still too small to fill the gap. The one pair of numbers that tells you first is the combined quarterly revenue of the growth portfolio against the combined decline of Eliquis and Opdivo. When the first crosses the second, the bet is working. If 2028 arrives before it does, it is not.