The Silver Wave

I. The Only Variable That Is Already Known

Almost everything in investing is a forecast. What inflation will be next year. Whether central banks will raise rates. Whether spending on artificial intelligence will continue. Whether Taiwan will stay quiet. Every one of those things is an assumption dressed up in a model.

Demographics is not an assumption.

The people who will be 65 in 2040 are 51 today. They already exist. They have already been counted. Nothing, not rate policy, not a technological breakthrough, not an election, can change the fact that in fourteen years they will be 65. Demographics is the only major economic variable that is not forecast but read.

And what it says about the United States is unambiguous.

By 2030, every member of the baby boomer generation, born between 1946 and 1964, will have turned 65. That is roughly one fifth of the entire population of the country. In 2025 America passed what is known as Peak 65, the year with the most people turning 65: approximately 4.2 million.

The next threshold is 2034. Census Bureau projections show that for the first time in American history, people aged 65 and over will outnumber children under 18: 77.0 million against 76.5 million. The number of Americans aged 65 and over will grow from 58 million in 2022 to 82 million by 2050, an increase of 42%, with their share of the population rising from 17% to 23%.

There is a second, sharper layer inside those numbers. It is not simply that the total number of older people is growing. The oldest group is growing fastest. The population over 80 is expected to expand at an average annual rate of roughly 5.4% between 2026 and 2030, against 1.8% a year between 2010 and 2025. That is not a smooth curve. That is an acceleration.

The ratio between workers and retirees is shifting too. In 2010 there were 22 retirees for every 100 people of working age. By 2030 that number is expected to reach 35. After that the curve levels off, reaching 37 by 2050, but the move from 22 to 35 happens inside a single twenty-year window.

II. Japan as a Warning, Not an Analogy

If you want to see what an aging economy looks like, you do not need models. You just need to look at Japan.

As of 2022, 29.9% of the Japanese population was 65 or older, nearly double the American share. By 2030 that is expected to reach 31.4%, and by 2050 fully 37.5%. Japan is roughly twenty years ahead of the United States on this measure, which makes it a natural laboratory.

The result? Japan's health spending reaches 10.6% of GDP, against an OECD average of 9.3%. But the more important number is a different one: in 2015 people aged 65 and over, who made up about 27% of the population, generated approximately 60% of national medical expenditure.

We should be honest about the limits of the analogy. Japan has a universal health system, strong government control over prices, and a completely different payment structure. The United States spends nearly twice as much of its GDP on healthcare with a younger population. So the Japanese case is not a template America will repeat. It is evidence of direction, not of scale.

The direction is this: when a population ages, healthcare consumes a steadily larger share of the economy. And it does so permanently, not cyclically.

III. What the Research Says

The thesis that an aging population means higher health spending sounds obvious. That is precisely why it deserves to be tested rather than accepted.

Spending really is concentrated in age

The data from CMS, the American agency that runs Medicare and Medicaid, is unambiguous. In 2020, per capita health spending for people aged 65 and over was $22,356 a year. That group was about 17% of the population but generated roughly 37% of all health spending in the country.

Widen the age band and the picture gets clearer still. People aged 55 and over made up 30% of the population in 2023 but accounted for 57% of health spending. At the other end, people under 35 were 44% of the population and generated just 21% of spending.

The curve is not a straight line. It bends upward more steeply with every decade of life.

The CMS projections

The forecasting specialists at CMS expect national health spending to grow from roughly $5.9 trillion in 2026 to $8.6 trillion by 2033. As a share of the economy: from 18.6% in 2026 to 20.3% by 2033. Per capita spending is expected to rise from $16,570 in 2024 to $24,200 by 2033.

CMS states explicitly what the main driver is: people shifting out of private health insurance and into Medicare as a result of the continued aging of the baby boom generation.

Dementia as a separate curve

There is one disease whose relationship to age is so steep that it deserves to be looked at on its own.

According to the Alzheimer's Association report for 2026, the prevalence of Alzheimer's disease by age group looks like this: 5.2% among people aged 65 to 74, 13.8% among those 75 to 84, and 35.8% among those over 85.

Today roughly 7.4 million Americans over 65 live with dementia of this type. Absent a medical breakthrough, that number is expected to reach 13.8 million by 2060. The cost of treatment and long-term care for people with dementia is projected at $409 billion in 2026 and close to one trillion dollars by 2050.

This is a single diagnosis. What it costs in 2026 already compares to roughly 1.3 to 1.4% of American GDP, and it is moving along a curve that accelerates.

And now the counter-current, because there is one

Here we need to introduce something most analyses skip. It is not complicated, but it changes the picture.

In 1980 the American physician James Fries proposed a hopeful idea. If prevention and healthy living delay the onset of disease more than they delay death, then people will live longer but be ill for a shorter time. The period of sickness compresses. Cost per person would fall, even as life extends. This is known as the compression of illness thesis.

The opposing idea was formulated even earlier, in 1977, by Ernst Gruenberg, and argues exactly the reverse. Modern medicine handles the consequences of disease well, but in doing so it allows people to survive into ever more advanced age while accumulating more and more conditions at once. The period of sickness does not compress, it stretches. This is the expansion of illness thesis.

The real-world evidence is mixed, and that is the important part. A 2022 study in the journal Demography, tracking American generations between 1998 and 2016, concluded that successive cohorts experience neither the compression Fries predicted nor a universal expansion of time spent in poor health. The answer depends on what you measure: severe disability shows more compression, chronic disease shows more expansion.

There is a third line, and it is more uncomfortable still. In 1999 the economists Zweifel, Felder and Meier published research that became known as the red herring hypothesis. The phrase is an English idiom for a misleading trail, something that leads you in the wrong direction. Their argument runs as follows: the observed link between age and health spending is driven less by age itself than by proximity to death. A large share of what is spent on any one person is concentrated in their final months, regardless of how old they are. If that is right, an aging population by itself does not raise cost per person nearly as much as assumed. It simply pushes the same spending further out in time.

Twenty-five years of debate have not settled the question. Newer research even suggests that proximity to death is itself somewhat of a misleading trail, because it is really just a stand-in for a person's actual state of health.

What does this mean for the investor? That the aging thesis is true at the level of total demand. More people over 65 means more doctor visits, more procedures, more prescriptions, more Medicare enrollees. But the thesis is weaker than it looks at the level of spending per patient. Demographics guarantees volume. It does not guarantee margin. That distinction becomes central in the section on risks.

IV. Anatomy of the Sector

Healthcare is perhaps the most heterogeneous of the eleven sectors in the S&P 500. Under one label live business models that have almost nothing in common. Here is how XLV breaks down by subsector, based on the fund's composition as of 23 July 2026:

Source: State Street holdings file, 23 July 2026. Classification by Liquidity Desk.

The first thing to see: this is not a balanced sector. More than half of XLV is pharmaceuticals. Hospitals, the institutions that will literally meet the silver wave at their front door, are 1.3%.

The logic of each layer is different and worth understanding, because aging does not act on them equally.

Pharmaceuticals is an intellectual property business. It earns while the patent holds and loses almost instantly when it expires. Aging increases the number of patients, but the patent calendar determines the profit. This is a sector where demand is predictable and earnings are not.

Medical devices have perhaps the cleanest link to demographics in the whole sector. An artificial joint is not a lifestyle choice, it is a necessity that arrives with age. Projections show hip replacements in the United States growing 71% to roughly 635,000 a year by 2030, and knee replacements 85% to 1.26 million. That is volume that comes directly out of the age pyramid. Yet medical devices are only 14.5% of XLV.

Lab and research equipment, companies like Thermo Fisher, Danaher, Agilent and Mettler-Toledo, is the equivalent of ASML and Applied Materials from our May analysis of semiconductors. They do not discover drugs. They sell the equipment others use to discover them. They earn from the level of research activity across the whole industry rather than from the success of any one molecule. That is a steadier profile, but it depends on the size of pharmaceutical research budgets.

Insurers are the only layer for which aging cuts both ways. More Medicare Advantage enrollees means more revenue. But older and sicker members mean higher claims paid out. An insurer does not profit from people being ill. It profits from having correctly estimated how ill they will be. This is an insurance business, not a healthcare business.

Distributors, McKesson, Cencora and Cardinal Health, work with enormous volumes and razor-thin margins. They are logistics, not medicine. Aging increases the number of prescriptions, which is directly positive for them. This is one of the most underappreciated links in the entire sector.

V. XLV: The Instrument

The State Street Health Care Select Sector SPDR ETF is the oldest and largest way to access American healthcare as a sector. The fund launched on 16 December 1998 and has lived through the dot-com bubble, the financial crisis, the Obamacare reform, COVID and the current cycle.

Source: State Street / Stock Analysis, 24 July 2026

Two numbers in that table deserve attention, because they only appear to contradict each other.

Beta 0.57. This is the technical way of saying that when the market moves 1%, XLV moves 0.57% on average. For comparison, SMH, the subject of our May analysis of semiconductors, has a beta of 1.36. Healthcare is among the calmest sectors, alongside consumer staples and utilities. The historical data confirms it: over the period from December 1998 to May 2026 the average drop from peak for XLV was 7.64% against 11.62% for SPY, the fund that tracks the whole S&P 500.

Price return over five years: 23.65%. That is a harsh number. Five years in which the S&P 500 did many times better. Healthcare was one of the weakest performing sectors between 2022 and 2025. Its weight in the S&P 500 reached almost 16% at the end of 2022, second only to technology, and fell to 10.36% in April 2026, the lowest level since September 2000.

So: the aging thesis was true throughout that entire period and the sector still lagged badly. That is a lesson worth remembering. Certain demand is not the same as good returns.

Composition and concentration

Top 10 of 63 holdings | Source: State Street, 23 July 2026

The top five positions are 46.7% of the fund. The top ten are 62.0%. For a sector regarded as defensive and well spread, that is remarkable concentration.

And it has not always been this way. In 2020 Eli Lilly was around 4% of XLV. Today it is 16.1%, more than one dollar in every six in the fund. This change is not the result of a change in index rules. It comes down to one thing: GLP-1, the class of diabetes and weight-loss drugs that includes Mounjaro, Zepbound and Ozempic.

The companies you need to know

Eli Lilly is the sector's centre of gravity right now, in the way NVIDIA is for semiconductors. For the first quarter of 2026 the company reported revenue of $19.8 billion, up 56% year on year. Mounjaro generated $8.7 billion for the quarter, up 125%, and Zepbound $4.2 billion, up 80%. Earnings per share came in at $8.55 against expectations of $6.66. The company raised its full-year 2026 guidance to between $82 and $85 billion in revenue. The quarter was also the first with orforglipron approved, the first drug in this class that is taken as a tablet rather than by injection.

It is worth noting, though, what Lilly is not: this is not a company whose growth comes from aging. GLP-1 drugs treat diabetes and obesity, and their market depends on lifestyle far more than on age. The largest position in the most demographically driven sector is powered by a theme that has almost nothing to do with demographics.

Johnson & Johnson is the opposite pole. Broadly spread pharmaceuticals and medical devices, decades of rising dividends, a business stretched across dozens of treatment areas. This is the position that makes XLV defensive.

AbbVie is a story about life after the patent cliff. The company lost patent protection on Humira, the best-selling drug in history, and survived because it had built Skyrizi and Rinvoq in advance. It is a textbook example of how a patent cycle is managed, and a reason AbbVie deserves attention as a model rather than merely as a holding.

UnitedHealth Group is the most interesting case at the moment. After a difficult stretch the company is recovering. The key measure here is what share of collected premiums goes out again to pay medical costs. In 2025 that share rose to 88.9%, which is very high. In the second quarter of 2026 it fell to 86.7% from 89.4% a year earlier, earnings per share reached $6.38, and full-year guidance was raised to between $19.50 and $20.00. How this is being achieved deserves to be named, though: the company is exiting unprofitable Medicare Advantage markets and expects to lose more than 3 million members in 2026. The margin is recovering through contraction, not growth.

Thermo Fisher and Intuitive Surgical represent the two infrastructure bets in the sector. One on research activity in general, the other on robotic surgery as a standard of care. Both are businesses where revenue comes from equipment already sold and the consumables that go with it, rather than from one-off sales.

VI. What XLV Does Not Capture

Honesty requires saying this too. XLV is not a healthcare fund. XLV is a fund of the healthcare companies inside the S&P 500. The difference matters, and it has four dimensions.

European pharmaceuticals are entirely absent. Novo Nordisk, AstraZeneca, Roche, Novartis, Sanofi, GSK. Not one of them is in XLV, because not one of them is in the S&P 500. This is especially sharp in the GLP-1 context: Novo Nordisk is Eli Lilly's direct competitor in the most important drug battle of the decade, and the XLV investor participates in only one side of that duel.

Small and mid-cap biotech is almost entirely missing. XLV holds 63 companies, all large, all proven, all profitable enough to be in the S&P 500. Innovation in biotech, however, happens mostly at companies that are not yet there. An investor who wants exposure to the early stage of discovery looks at funds like XBI or IBB, not at XLV.

Long-term care and senior housing are not in the sector. This is perhaps the most ironic gap for a demographic thesis. Welltower and Ventas, the two companies that literally own the buildings an aging America will live in, are classified as real estate and sit in XLRE, not XLV. And their results are exactly what the thesis predicts: occupancy in Welltower's operating portfolio reached 87.3% in the first quarter of 2026 against 85.1% a year earlier, and the company carries a market value above $165 billion, larger than almost any position in XLV.

Hospitals are nearly absent. HCA, Universal Health Services and DaVita together are 1.3% of the fund. If you believe the silver wave means more hospital admissions, more procedures and more strain on the system, XLV gives you almost no participation in the point where that physically happens.

Put differently: XLV captures the pharmaceutical profit from American healthcare extremely well. It captures medical devices partially. And it captures almost nothing of the care itself, the infrastructure, or the early-stage innovation.

VII. The Counter-Current

Demographics is a tailwind. It is not a guarantee. Here is what stands on the other side of the thesis, and this is the section that deserves the closest reading.

Prices are no longer free

The Inflation Reduction Act fundamentally changed the rules. For the first time, Medicare negotiates drug prices directly. The first negotiated prices took effect in 2026 for ten medicines. The third round, announced in January 2026, covers 15 drugs and for the first time includes ones administered in a clinic or hospital rather than dispensed at a pharmacy. Final prices will be published on 30 November 2026 and take effect on 1 January 2028.

The scale of the effect: negotiated prices reduce net spending on the selected drugs by an average of roughly 22%.

This is a permanent change, not a one-off event. The list widens every year. The mechanism that for decades allowed American pharmaceuticals to earn the highest margins in the world is being dismantled step by step. And the irony is precise: the same aging that increases volume also increases the political pressure on price. The more people are in Medicare, the bigger the budget problem becomes and the more inevitable regulatory intervention becomes.

The patent cliff to 2028

According to EY estimates, the twenty largest biopharmaceutical companies have roughly $180 billion of revenue exposed to patent expiry through 2028.

Two examples explain the scale. Merck's Keytruda generated around $29.5 billion in 2024, approximately 56% of the company's entire business, and the key patents on the intravenous version expire in 2028. Bristol Myers Squibb's Eliquis brings in about $13 billion a year, expiring in 2027 or 2028.

Merck is 5.52% of XLV. Bristol Myers Squibb is 2.15%. This is not an abstract risk. It has a date.

The tariffs that are not really tariffs

The logical first impression is that tariffs on imported medicines ought to be good news for XLV. The fund holds only American companies. The tariff hits imports. It looks like straightforward protection in favour of exactly what you own.

The answer is "partly yes, but not for the reason you expect." And the difference matters.

First, the legal basis. On 20 February 2026 the US Supreme Court struck down, by 6 votes to 3, the tariffs imposed under the emergency IEEPA statute. The ruling, however, touched only that particular route. Tariffs under other statutes, including Section 232 of the Trade Expansion Act of 1962, were left intact.

The pharmaceutical tariffs were signed on 2 April 2026, six weeks after the ruling, and precisely under Section 232. The administration did not use the route that was challenged. It chose the one the court left standing. The conclusion is clear: do not count on these being struck down in court.

Second, the rates themselves. The order imposes tariffs on imported patented medicines and on the active ingredients from which they are made. The headline rate is 100%. It falls to 20% for companies with approved plans to move production to the United States, reverting to 100% after four years, and to 15% for products from the EU, Japan, South Korea and Switzerland. Generic drugs, biosimilars (copies of biological medicines) and drugs for rare diseases are excluded. The measures take effect on 31 July 2026 for the largest companies and 29 September for the rest.

Third, and this is the substance. By the time the order was being signed, most of the large manufacturers had already stepped out from under it.

Fourteen of the seventeen companies the administration approached signed pricing agreements in exchange for a three-year exemption from tariffs. The list spans both sides of the Atlantic: Pfizer, Eli Lilly, Amgen, Bristol Myers Squibb, Gilead and Merck on the American side, and AstraZeneca, Novo Nordisk, Novartis, Sanofi, GSK and Roche through Genentech on the European.

This is where the most obvious bullish hypothesis collapses. If you expected the tariff to punish Novo Nordisk in favour of Eli Lilly in the fight for the GLP-1 market, it does not happen. Both signed. Both received exemptions. The competitive asymmetry the tariff appears to create was negotiated away before it took effect.

Fourth, what the exemption cost. It was not a gift. The companies paid for it in two currencies.

The first is price. The agreements require lower prices for Medicaid, direct discounts to consumers through the new TrumpRx platform, and a commitment to launch new medicines in the United States first. The second is capital. Pfizer committed to $70 billion of new US investment. Merck put up $1 billion for its first American Keytruda facility. Eli Lilly has already committed more than $50 billion to US manufacturing since 2020, including four new plants.

Look at that exchange carefully. A three-year reprieve from a tariff, paid for with permanent price concessions and tens of billions in capital expenditure. The reprieve expires. The concessions do not.

Which gives us the real conclusion: tariffs are not a separate risk to the sector. They are a continuation of the same one. The headline is trade policy, the mechanism is pricing policy. This is the same downward pressure on prices we saw with the Inflation Reduction Act, only negotiated instead of legislated.

So where do tariffs genuinely help XLV? In one direction, and it is real. Buying an exemption requires scale: lawyers, lobbying, a balance sheet that can absorb multi-billion-dollar investment, and a portfolio large enough for the administration to bother negotiating with you. The companies in XLV are precisely the ones that can. The mid-cap pharmaceutical companies and the biotechs, which as we saw are largely absent from the fund, cannot. In that sense the tariff is a barrier to entry rather than protection of a market. It favours the large at the expense of the small, and XLV holds only the large.

And one detail worth checking. Three companies have still not finalised agreements: Johnson & Johnson, AbbVie and Regeneron. Together they weigh 19.6% of XLV. Almost one dollar in every five in the fund sits in companies that have not yet paid for an exemption. Whether they sign, on what terms and at what cost, is an open question with a specific date attached: 29 September 2026.

Medicare's solvency

According to the trustees' annual report for 2026, the part of Medicare that pays for hospital treatment will exhaust its reserves by 2033. That is three years earlier than the 2024 report projected, which pointed to 2036.

This may be the single most important risk in the whole analysis, because it connects both sides of the thesis. The silver wave is good news for the volume of healthcare services and bad news for whoever pays for them. And in the United States the largest payer for older people is the federal government. When the arithmetic stops working, and it has a deadline, the response will be some combination of lower prices paid to hospitals and doctors, tighter eligibility criteria, and more pressure on manufacturers. None of those is good for the sector's margins.

This is exactly where we return to the red herring. If a significant share of spending is determined by proximity to death rather than by age itself, then budget pressure will build more slowly than assumed. But by the same logic, so will the sector's revenue. Both sides of the equation are connected.

Concentration in GLP-1

Eli Lilly is 16.1% of XLV. A large part of the entire sector's performance over the past year comes down to one class of drugs.

Forecasts for this market keep moving. Goldman Sachs raised its estimate for the global anti-obesity drug market to roughly $114 billion by 2030, up from $101 billion, with tablet forms expected to account for about 40%. Morgan Stanley projects the wider GLP-1 market reaching around $190 billion by 2035. But the same Goldman had previously cut its forecast, from $130 billion to $95 billion, in 2025.

Forecasts that swing by tens of billions in both directions inside two years are not forecasts. They are a consensus still forming. And the XLV investor has one dollar in every six riding on the outcome.

VIII. Where the Sector Stands Now

For context, here is XLV against the other ten sectors of the American economy:

Price return, excluding dividends | Source: Stock Analysis, 24 July 2026

The picture is clear. Over three years healthcare is among the weakest sectors, 19.33% against 99.72% for technology. Over one year and three months it is among the strongest. The sector is recovering from a multi-year period of lagging.

Valuation reflects that. On expected earnings healthcare trades at roughly 17 to 19 times, against about 22 to 23 times for the S&P 500. That is a discount, and it is the reverse of the historical norm. In the late 1990s the sector traded at around 1.4 times the index multiple; today it is at about 0.8. XLV's own P/E of 24.64 is higher because it reflects earnings already delivered rather than those expected. The gap between the two numbers is the expected growth.

Whether that discount is an opportunity or a fair price for regulatory risk is exactly the question the market has not yet answered.

IX. The Bottom Line

Healthcare is a rare case in investing: a sector whose long-term demand is mathematically known. The people who will fill the waiting rooms in 2040 have already been born, already been counted, and are already a specific age. There is no scenario in which this wave does not arrive.

But the last five years are a lesson in themselves. The thesis was true the whole time and the sector still lagged badly, with its weight in the S&P 500 falling to the lowest level since 2000. Known future demand is not the same as future returns. Between the two stand prices, regulation, patents, and the question of how much has already been paid.

The three things this analysis shows:

First, demographics guarantees volume but does not guarantee margin. More patients means more procedures and more prescriptions. Whether that turns into profit depends on prices. And prices in American healthcare are no longer free, and will not become freer. The same force that creates the demand also creates the budget pressure against it.

Second, XLV is not a pure demographic bet. The largest position in the fund, at 16%, is driven by a theme about metabolism rather than about age. The layers that would benefit most directly from aging, hospitals, long-term care and senior housing, are either minimal or entirely outside the fund. If you are buying XLV because of the silver wave, you are buying something that only partly overlaps with your thesis.

Third, this is a defensive position with average returns. Beta 0.57. An average drop from peak of 7.64% against 11.62% for SPY. On a comparable basis from the fund's inception to May 2026, XLV's annual return is about 8.2% against roughly 8.7% for SPY. A difference of around half a percentage point for considerably less turbulence. That is the trade. For some investors it is exactly what they want. For others it is not enough.

The silver wave will arrive. That is the only certain thing in this entire analysis. The question every investor has to ask is duller and more important: how much of it is already in the price, and where exactly along the chain will the profit stick.

Understanding a sector does not tell you when to buy. But it tells you something more important: what exactly you are buying. And in investing, clarity about the thesis is half the work.