Healthcare: two markets, one value chain

Demand is structural, and it keeps growing

Healthcare’s share of economic output across developed economies has risen for decades. Two forces explain the rise: a shift in population composition toward older cohorts that consume more healthcare per capita, and a rise in real healthcare spending per person in all age cohorts. Both trends have held consistently across the developed world, and both are set to continue.

Adults aged 65+ continue to grow in number

The over-65 population is growing, and that group spends far more on healthcare. These are two separate forces, and both point the same way.

On the first, the over-65 share of the population has climbed steadily for decades and is set to keep climbing. In 1950 it was around 8% in both the United States and Europe and roughly 5% in Japan and China. By today it has reached about 18% in the United States, 21% in Europe, and 30% in Japan, with China at 15%. By 2040 the figures are set to reach roughly 22% in the United States, 26% in Europe, 35% in Japan, and 27% in China (Figure 1). Every major economy has at least doubled its over-65 share since 1950, with Japan going from 5% to a projected 35%, a sevenfold rise. China is the fastest mover from a standing start: almost all of its ageing has come since 2000 and continues at pace. In absolute terms, these four countries will add roughly 200 million people aged 65 and over between 2025 and 2040, about 146 million in China alone, 33 million in Europe, 18 million in the United States, and 3 million in Japan.

On the second, a cross-country comparison found that across eight high-income countries² per-capita health spending for the over-65 cohort was approximately 3.4 times the rate for working-age adults aged 20 to 64, with the gap widening further for the oldest age groups within the over-65 group³. The cohort's higher spending reflects both more disposable income and a disease mix concentrated in conditions requiring long-duration therapeutic management, such as cardiovascular disease, type 2 diabetes, cancer, and neurodegenerative conditions. The combination of more 65+ adults and higher spending by the group implies sustained upward momentum on aggregate healthcare spending from demographic composition alone.

Spending per person keeps rising

Even setting ageing aside, real healthcare spending per person has tended to rise over time, and this is a larger force. Health expenditure currently accounts for 9.3% of GDP across the OECD on average, on a total-spending basis including both public and private sources, up from roughly 5% in 1970, and is projected to keep climbing. On current OECD projections, public health spending alone rises by about 1.5 percentage points of GDP by 2045⁴.

What drives that growth matters to pinpoint where the investment opportunity sits. OECD's component-based projection model⁵ attributes this growth to four drivers (Figure 2). The largest is income: as countries get richer, they choose to spend proportionally more on health. Ageing comes after, explaining roughly a quarter of the growth, smaller than most assume. The rest comes from the rising cost of delivering care and, importantly for this research, from new medical technology, the new drugs and devices that treat conditions which previously could not be treated. That last point is the bridge to the rest of the research.

A large and rising share of healthcare spending goes to innovation that did not exist before, and the investment question is who funds that innovation, and who captures its value. The next section follows the money.

A value chain that crosses private and public capital markets

If the first section is about the size of the demand pool, this one is about how the supply of new therapies is financed and where the resulting commercial value sits. The analysis focuses on pharmaceuticals, where the dynamic is most clearly observable; related patterns in medtech and life sciences tools are noted but not examined in depth.

Drug development is funded in stages

Drug development is usually funded by different sources of capital as it moves from lab to market:

• Discovery and preclinical research are funded primarily by public-sector grants (the NIH in the United States, EU framework programmes in Europe), academic institutions, and early-stage venture capital. This stage carries the highest scientific risk and the lowest capital intensity per programme.
• Phase I and II clinical development is funded primarily by venture capital, crossover funds, and partnership capital from listed pharma. Capital intensity rises sharply: mean costs per clinical phase are approximately $36 million at Phase I and $88 million at Phase II⁶.
• Phase III clinical development is the largest single per-trial capital commitment in development, at approximately $250 million per Phase III study. It is funded by public equity markets for companies that have listed, by listed pharma for assets acquired or partnered earlier, or by specialist late-stage private capital.
• Approval, launch, and commercialisation rest predominantly with listed pharma, reflecting the strong scale economies in the global regulatory, manufacturing, sales, and reimbursement infrastructure required to bring a drug to market. This is visible in the concentration of approved-drug revenues at the top 15 to 20 listed companies.

The headline cost of a drug is dominated by failure, not by any single trial. These per-phase figures are the costs of conducting each phase, not the all-in cost of producing an approved drug. Counting the failures as well as the successes, according to research the all-in cost per approved medicine ranges from approximately $3 billion to $4 billion, depending on the methodology applied⁷. That cost of failure is the central reason the financing chain exists: it is too large and too risky for any one owner to carry from start to finish, so the risk is passed along and repriced at each stage.

The transition between private and public capital ownership of a drug asset typically occurs at or before Phase III, mediated by M&A, initial public offering (IPO), or partnership, though this is a dominant tendency rather than a fixed rule. Figure 3 summarises the predominant capital source, approximate cost, and probability of success at each stage. Some companies raise public equity at earlier clinical stages and fund development through public markets directly, and a smaller number develop, approve, and commercialise products independently without ever being acquired: Vertex, Gilead, Amgen and Regeneron are examples of fully integrated commercial businesses built from privately funded origins.

The stage at which the private-to-public transition occurs also moves with market conditions. In risk-on periods, most recently 2020 to 2021, the IPO window opens to assets at progressively earlier stages, extending in the most buoyant conditions to preclinical and platform-stage companies with no clinical data, and public equity can fund early clinical development directly. In risk-off periods, most recently 2022 to 2024, the window narrows and private capital extends further into late-stage development. Acquisition by listed pharma, by contrast, has continued across both kinds of period, a pattern examined in sub-section 2.3.

Most new drugs today are invented outside big pharma

Large pharma increasingly buys its innovation rather than producing it in-house. Of the 138 drugs approved for the top 20 biopharma companies between 2015 and 2021, 65% came from outside the company, mostly through acquisition; only 28% were invented in-house. Four of the 20 (Johnson & Johnson, Sanofi, Takeda, Biogen) sourced every one of their approvals externally⁸.

Smaller companies now carry the innovation pipeline. Emerging biopharma (companies with under $500 million of R&D spending and under $200 million of sales) accounted for 65% of the drug pipeline in 2021, up from 34% two decades earlier, and for 42% of FDA filings, up from 11% in 2012⁹. Figure 4 shows the shift.

The pattern is consistent with a structural decline in the productivity of in-house pharmaceutical R&D documented over several decades. The number of new FDA-approved drugs per billion US dollars of R&D spending has roughly halved every nine years since 1950, an 80-fold decline, a pattern researchers named “Eroom’s Law” (Moore’s Law spelled backwards)¹⁰. Externalisation of drug discovery represents one of the responses available to large pharma when the marginal productivity of in-house research has declined materially relative to the cost of acquiring de-risked assets from outside the firm.

Patent expiries fuel M&A capital flows

Patent expiry is a constant condition. Every year a wave of drugs loses exclusivity and the revenue collapses to generic competition.

Over the next five years, brands generating more than $300 billion in annual sales face loss of exclusivity across the industry, equivalent to 3-4% of the global market in each year, with a peak near 7% in 2028 driven by the anticipated loss of Merck's Keytruda, the single largest event. After Merck, the next four most exposed companies (AbbVie, Johnson & Johnson, Roche, and Bristol Myers Squibb) each face more than $30 billion in cumulative sales at risk by 2030, and the average top-25 pharma company has around $24 billion at risk by the end of the decade¹¹. Figure 5 lists the largest of these expirations.

Expirations of this scale create a revenue replacement problem that internal pipelines have not historically been able to solve at the top listed names. A company facing the loss of a multi-billion-dollar drug within a defined window, and unable to replace it from in-house development alone, has a direct incentive to acquire de-risked external assets. This is the mechanism that connects the externalisation pattern of sub-section 2.2 to the sustained flow of acquisition capital: the patent cliff sets the size of the gap, and the limited productivity of internal R&D determines that much of it will be filled from outside the firm. Recent transactions illustrate the pattern, including Johnson & Johnson’s $14.6 billion acquisition of Intra-Cellular Therapies (2025), Bristol Myers Squibb’s $14.0 billion acquisition of Karuna Therapeutics (2024), and AbbVie’s $8.7 billion acquisition of Cerevel Therapeutics (2024).

M&A has been the route to replace revenues from patent expirations. The flow of acquisition capital is not uniform across cycles. Activity softened in 2024 to its lowest level since 2021, then paused again early in 2025 amid uncertainty over US tariff and drug-pricing policy, with first-half dealmaking limited mainly to transactions below $5 billion. As that uncertainty eased through the year, the pace reasserted itself sharply: aggregate biopharma M&A reached approximately $133 billion in 2025, more than double the 2024 level and the second-highest annual total in five years, across roughly 50 transactions at an average size of $2.7 billion¹². The recovery was led by four deals of $10 billion or more, including Johnson & Johnson’s acquisition of Intra-Cellular Therapies and Novartis’s $12.7 billion acquisition of Avidity Biosciences. What has been consistent across this cycle is the structural presence of the flow rather than its level in any single year, and the approach of the 2025-2030 expiry wave has coincided with a renewed emphasis on external sourcing.

Implications for portfolio construction

The first two sections describe a single value chain: a demand pool that grows steadily, and a supply of innovation financed across a sequence of private and public hands. The questions for investors are what that structure has meant for diversification, returns, how listed and private healthcare have performed over long horizons, how closely they have moved together, and mainly how they complement each other if you want exposure to the sector.

The two markets are shaped very differently

The listed healthcare universe is a few hundred companies large. Private healthcare is tens of thousands of small ones. The S&P Global 1200 Health Care index¹³, used here as the listed proxy, holds 110 to 120 constituents; broader global healthcare indices run to roughly 200. The private healthcare universe is roughly 30,000 venture- and private-equity-backed healthcare companies globally¹⁴, the majority of them venture and growth stage. The flow of capital into this universe is substantial and has been growing: global healthcare private equity deal value reached a record of more than $190 billion in 2025, surpassing the prior 2021 high¹⁵.

The shapes differ because the two markets sit at different points on the value chain: listed pharma provides concentrated exposure to a relatively small number of large companies; private healthcare provides exposure to a much wider field of smaller companies, mostly at earlier stages of commercial development.

Healthcare has lagged the broad market in recent years, but private has beaten public over the long run

Public healthcare has trailed broad equities over five and 10 years. Listed healthcare returned about 7.1% a year over five years and 8.7% over 10 years, against roughly 12.8% for the broad S&P Global 1200 (Figure 6). The sector has been out of favour. An allocation to healthcare is a view that this reverses, supported by the structural demand in Section 1.

However, within healthcare, private capital has beaten public over the long run, mainly in buyout. Over 10 and 20 years, healthcare buyout returned about 16.8% and 15.3% a year, well ahead of listed healthcare’s 8.7% and 9.3%. Venture sits in between. Only the most recent five years break the pattern, and for a known reason: the 2022-2024 repricing of growth and venture assets pulled five-year private returns (about 6.5% combined) below listed healthcare’s 7.1%. As section 3.5 discusses, the outperformance depends heavily on manager selection.

The two markets have similar, but not identical, characteristics

Private and listed healthcare move together moderately. Measured on reported returns over the past 20 years (2006 to 2025), the correlation between listed healthcare (S&P Global 1200 Health Care) and private healthcare is about 0.48 for buyout, 0.58 for venture, and 0.59 for the combined exposure. When the private series are unsmoothed¹⁶, the correlations barely move: over 20 years they become 0.48, 0.56, and 0.60. Because unsmoothing changes so little, the moderate correlation appears to be a real feature of the two markets, not an artefact of how private assets are valued.

A correlation around 0.5 to 0.6 is in line with what the value-chain relationship predicts: it is high enough to confirm the two markets are linked, but well short of the near-unity that would make one a substitute for the other. Roughly half of private healthcare return variation is not explained by listed healthcare. The two are connected ends of one chain, not two routes to the same exposure. Figure 7 sets the two side by side.

The two markets are weighted toward different parts of healthcare

Beyond stage and return profile, listed and private healthcare differ in which sub-industries they actually contain. Measured on a consistent taxonomy (Figure 8), the two markets are visibly complementary as each gives material access to areas the other misses.

The private universe is far wider than the listed one (section 3.1), and that width concentrates in areas that listed markets structurally underweight or miss entirely. Healthcare technology and early-stage biotechnology are the clearest cases. Biotechnology by count is the single largest category in private healthcare, comprising thousands of preclinical- and clinical-stage companies versus a handful of large commercial names in the listed index. Health care services and facilities follow the same pattern: material in private, marginal in listed. The reverse also holds: large-cap pharmaceuticals and health care distributors carry substantial weight in listed indices but are a small fraction of the private opportunity set.

This matters for implementation. Even within the subset used for this analysis (Pictet Alternative Advisors' healthcare exposure, over 1,000 companies), the sub-sector complementarity is already visible. An allocator has more than enough surface to act on sub-sector convictions that listed indices cannot express.

The composition data also reinforces the return and correlation findings of the preceding sections. An investor who holds only the listed market is, to a first approximation, holding large pharma and large medtech, and is barely exposed to the healthcare technology and services innovation that sits predominantly in private hands. Holding both is what captures the full sector.

Manager selection decides the private outcome

In private healthcare, picking the manager matters significantly. Across healthcare-focused private capital funds measured over rolling 10-year horizons, the spread between top-quartile and bottom-quartile performers is wide: in healthcare buyout, the 75th and 25th percentile 10-year internal rates of return sit at approximately 21.5% and 7.0% respectively, a spread of roughly 14 percentage points; in healthcare venture capital the spread is wider, with the 75th percentile near 14.6% and the 25th percentile below zero¹⁷. Dispersion of this magnitude means the choice of manager, not merely the choice to allocate, drives realised outcomes in the private sleeve. The listed sleeve, by contrast, is more straightforwardly accessed through low-cost index or active strategies, where manager dispersion is far narrower.