Something unusual is happening across medical device stocks. Boston Scientific has fallen more than 50% year to date while 27 of the 31 analysts covering it still rate it a buy, with a consensus target implying roughly 65% upside. Medtronic has underperformed the broad market over the past year. Danaher lost a substantial share of its market value in a single session this week on a timing disclosure. The iShares US Medical Devices ETF has lagged health care overall despite holding the sector's strongest franchises.
When multiple leading companies in one sector fall hard while the analyst community stays bullish on all of them simultaneously, that is not a story about execution at any single firm. It is a category being repriced, and the useful question is what the market thinks it knows that the models have not absorbed.
The premise underneath the premium
Medical device companies have historically commanded good multiples on a specific and reasonable logic. Populations age, aging generates procedures, procedures require devices, devices carry recurring consumable revenue, and none of it is discretionary because people do not decline necessary surgery. Demographics provided a growth floor and clinical necessity provided pricing power.
Every element of that chain except demographics is now being questioned, and two distinct forces are doing the questioning at the same time. Separating them is the analytical work, because they have opposite implications and the market may be conflating them.
The first force removes procedures permanently
GLP-1 drugs are the clearest example, and the effect is already measurable rather than theoretical. Intuitive Surgical reported bariatric case declines in the high single digits as obesity medications altered treatment patterns, which is not a deferral of surgery but a replacement of it. A patient who achieves durable weight reduction pharmacologically does not have a bariatric procedure pending. That operation is gone.
The reason this matters beyond bariatric surgery is that the drug class affects the disease burden upstream of several other device categories. Sustained metabolic improvement plausibly touches cardiovascular intervention, orthopedic load-bearing joint procedures, and sleep apnea treatment, all of which are meaningful device markets. Whether those second-order effects materialize at scale, and over what horizon, is genuinely unknown. The point is that a therapeutic class which reduces the incidence of conditions requiring devices is a structurally different threat than a competitor with a better catheter.
Medical device revenue is a function of how many procedures happen. Anything that reduces the number of procedures reduces the addressable market rather than the company's share of it, and no amount of execution recovers it.
The second force defers procedures temporarily
The other pressure is affordability and coverage, and it operates on a different mechanism with a different resolution.
Enhanced ACA subsidies expired at the end of 2025, marketplace enrollment fell, and elective and deferrable procedures are exactly the category people postpone when coverage becomes expensive. UnitedHealth's chief financial officer attributed membership declines largely to affordability pressures, and device makers see the same phenomenon from the other side as procedure volumes soften.
But a hernia does not resolve because someone lost their subsidy, and an arthritic knee does not improve because a deductible rose. That demand is stored rather than destroyed, and it returns when coverage or income conditions change, frequently in a more urgent and expensive form.
The distinction that explains the divergence
Here is why analysts and the market can look at identical numbers and reach opposite conclusions.
A discounted cash flow model handles deferral gracefully. Push procedures from this year into next, and the present value barely moves. Terminal growth assumptions stay intact, which is where most of the value in a med-tech model sits. That is a large part of why sell-side targets remain far above current prices, and it is not obviously wrong.
Substitution is different. It changes the terminal growth rate, and small changes to a terminal rate produce large changes in valuation. A market that suspects some portion of the current volume weakness is permanent will reprice the multiple immediately, well before any model formally revises its assumptions, because a price can move on suspicion while a model requires evidence.
So the divergence between a 50% decline and a 65% implied upside is not necessarily analysts being slow or the market being irrational. It is two different instruments responding to two different questions. The models are answering whether the current numbers justify the current price given historical growth. The price is answering whether the historical growth assumption still holds.
Which is right depends entirely on how much of the volume softness is substitution and how much is deferral, and that ratio is not knowable yet from published data.
The company-specific problem that makes it worse
Sector pressure explains part of the decline. Boston Scientific's situation adds something else, and it is worth isolating because it is a different kind of damage.
Management lowered full-year organic revenue guidance in April, then walked expectations back again in late May, with the chief executive citing declining usage of standalone Watchman procedures and forecasting roughly flat sequential dollar growth into the third quarter.
Revising guidance downward twice within two months is a forecasting failure distinct from the underlying miss, and markets punish it disproportionately for a specific reason. A single guidance cut says conditions changed. A second cut shortly afterward says management did not have visibility into its own business when it issued the first one. Investors can price a known deceleration. They cannot price a company whose own estimates are unreliable, so they apply a discount for the uncertainty itself, and that discount persists until several quarters of accurate forecasting rebuild the credibility.
The same dynamic applied to Danaher this week, where a comparatively small revenue timing shift produced a disproportionate decline. In both cases the market was not repricing the reported number. It was repricing confidence in the forecast.
The competitive layer
There is a third pressure that is more conventional and easier to analyze. Johnson & Johnson's entry into robotic surgery changes the competitive backdrop for Medtronic and others in a category that has been effectively a single-company market for two decades.
Well-capitalized entry into a high-margin category compresses multiples in advance of any actual share shift, because the market prices the probability of future price competition rather than waiting for it. That is ordinary and it is separable from the volume questions above. It also cuts against the incumbents specifically rather than against the sector, which is why it explains some of the dispersion within the group.
What would resolve it
The sector's valuation depends on a factual question with an eventual answer, and several observable things will supply it.
Whether procedure volumes decline or merely decelerate. Deceleration in a growing base is consistent with deferral. Absolute decline in categories with GLP-1 exposure is consistent with substitution, and the two produce very different terminal values.
Whether GLP-1 effects appear beyond bariatric surgery. If cardiovascular, orthopedic, and sleep apnea device volumes hold up as obesity drug penetration rises, the substitution thesis is contained to one category and the sector reaction was excessive. If they soften in a correlated way, it is not.
Whether deferred procedures return. Coverage-driven deferral should reverse as enrollment stabilizes, and a visible catch-up in elective volumes would confirm that a meaningful share of the weakness was temporary.
And whether guidance stabilizes. For the companies that have walked back forecasts repeatedly, several quarters of meeting their own numbers would do more for the multiple than any single strong result, because the discount being applied is for unreliability rather than for performance.
The sector's underlying thesis, that populations age and aging generates medical intervention, has not been refuted. What has been disrupted is the assumption that the relationship between demographics and procedure volume is stable. A drug class that reduces disease burden and a coverage environment that delays care are both wedges in that relationship, and the market has decided to price them before anyone can measure them. Whether that turns out to be foresight or overreaction is the entire question, and the procedure volume data over the next several quarters is where it gets answered.