Primer

Enterprise Foundations: Commercial Ownership, Patient Capital, and Long-Horizon Research

David Hilmer Rex · June 2026

Summary

An enterprise foundation is a nonprofit that holds a controlling stake in a commercial company and directs the profits toward research and philanthropy. Unlike a donor-endowed foundation, it never cashes out: research funding flows from ongoing commercial success rather than a finite endowment pool. Denmark has the world's highest concentration of this model — Novo Nordisk, Carlsberg, LEGO, and A.P. Møller-Mærsk are all foundation-controlled — and the empirical evidence shows that foundation-owned firms invest more in R&D, survive significantly longer, and maintain environmental and social performance better under stress than comparable companies. This primer examines how the model works, what conditions produced it, how it varies internationally, and what its governance challenges reveal about the relationship between ownership structure and long-horizon research.

The ownership model ↑

An enterprise foundation is a nonprofit entity that holds a controlling stake in a commercial company and directs the profits toward research, philanthropy, or social purposes. There are no private shareholders who can extract the surplus. The profits from operating a pharmaceutical company, a brewery, an engineering firm, or a toymaker accumulate in a foundation and are distributed according to its charter.

This is structurally distinct from three things it is often confused with. A donor-endowed foundation — like the Gates Foundation — draws its capital from wealth a founder extracted from a company. Enterprise foundations don't cash out; they remain the owner. A research endowment holds diversified financial assets; enterprise foundations hold controlling stakes in operating businesses, with income tied to a specific company's performance. A public enterprise is owned by the state; enterprise foundations are privately governed, with no direct state involvement.

The structural consequence is significant: because the foundation owns the company rather than endowing it from extracted wealth, the commercial enterprise must remain competitive, and research funding flows from ongoing commercial success rather than a finite capital pool. The foundation's capacity to fund science rises and falls with the company's fortunes — connecting long-horizon research directly to the health of specific industries.

Conditions, not design ↑

Enterprise foundations did not emerge from deliberate policy. They are products of specific combinations of commercial law, tax regimes, business culture, and individual founders who wanted institutional continuity rather than personal extraction.

The Carlsberg Foundation was established in 1876 by J.C. Jacobsen, who wanted his brewery to remain Danish-owned and committed to scientific research after his death. He was not designing a research-funding model; he was solving an inheritance problem. Danish commercial law permitted a self-owning entity — a foundation with no private beneficiaries — to exist. Other legal systems would not have.

The Novo Nordisk story goes back further. Nordisk Insulinlaboratorium was established in 1923 by the Nobel laureate August Krogh, together with the diabetes physician H.C. Hagedorn and the pharmacist August Kongsted, and adopted foundation ownership in 1926 — a structure tied to the conditions under which Krogh had secured the Scandinavian rights to the original insulin patent: that insulin production should benefit diabetes patients, not private owners. Two former employees, the brothers Harald and Thorvald Pedersen, started a competing company, Novo Terapeutisk Laboratorium, in 1925; it was transferred to foundation ownership in 1951. The two foundations and companies were eventually merged in 1989 to form the Novo Nordisk Foundation and Novo Nordisk as they exist today.¹ In both founding cases, the foundation structure was a constraint chosen to protect access and mission, not a deliberate philanthropic programme. Research funding on the scale the foundation now undertakes came later, as the commercial enterprise grew and the surplus had nowhere else to go.

Steen Thomsen's foundational study of Danish industrial foundations emphasises this point: these structures were not programmed. They were the residue of legal frameworks that permitted them, of business cultures that valued institutional continuity over personal extraction, and of specific historical moments that made the foundation structure attractive to specific founders. High inheritance taxes made it financially rational for Danish founders to route ownership through foundations rather than pass wealth to heirs. Change any of these conditions and the structures do not appear.²

This matters because it constrains what we can learn. Enterprise foundations are not a technology that can be deployed at will. They are precipitates of conditions — and understanding those conditions is more useful than treating the structures themselves as models to copy.

The Danish case ↑

Denmark has the world's highest concentration of enterprise foundations relative to its economy. Foundation-controlled companies — including three of Denmark's four largest, Novo Nordisk, A.P. Møller-Mærsk, and Carlsberg — account for an estimated 68–70% of total Copenhagen Stock Exchange market capitalisation. More than 1,300 industrial foundations are registered in Denmark today, of which roughly 400 own companies of real economic significance.

The scale of annual distribution is substantial. In 2024, the largest enterprise-foundation grantmakers distributed:

• Novo Nordisk Foundation: DKK 8.2 billion — health and life sciences research
• LEGO Foundation: DKK 1.5 billion — creativity and education (down from DKK 3.2 billion in 2022)
• A.P. Møller Foundation: DKK 1.5 billion — maritime, cultural, and educational
• Lundbeck Foundation: DKK 1.1 billion — brain research and health sciences (roughly doubled since 2022)
• Villum Fonden: DKK 1.0 billion — technical and natural sciences
• Carlsberg Foundation: DKK 845 million — scientific research and arts

In 2024, Denmark's enterprise foundations (erhvervsdrivende fonde) distributed DKK 19.6 billion to charitable purposes — 71% of the DKK 27.5 billion in total charitable grants made by all Danish foundations that year, a level on par with the two preceding years.³ Health was the largest category at DKK 5.1 billion — though down DKK 3.9 billion from 2023 — followed by nature and climate (DKK 4.1 billion, the largest increase of any category), culture (DKK 3.9 billion), and education (DKK 3.6 billion). By 2025, the Novo Nordisk Foundation alone was awarding DKK 11.8 billion annually — its largest total ever — with the share going outside Denmark climbing from 9% in 2023 to 13% in 2024 and 21% in 2025, as the Foundation expands across the Nordics and Europe, East Africa, South and Southeast Asia, and the USA.

These numbers are not simply large — they are structurally significant. The Novo Nordisk Foundation alone distributes more to biomedical research annually than most national research councils. Its investments fund dedicated research centres across Danish universities, international research partnerships, and basic science programmes with no direct commercial application — well beyond what Novo Nordisk the pharmaceutical company funds as R&D.

What makes this more than a story about one generous funder is the ecosystem effect. Danish science has a research-funding infrastructure substantially insulated from parliamentary budget cycles and the 3–5 year timelines of traditional grant funding bodies. That insulation is a product of the ownership model, not just the accumulated wealth.

That scale also creates a coordination question the foundations have only recently begun to address. A 2019 EU peer review of the Danish innovation system found that the foundations, for all their weight, were "not strategically integrated" into it — funding university research richly, but not "consistently at the table" when national innovation strategy is set, and not co-funding research infrastructure the way the Wellcome Trust does in the UK. The clearest move in the other direction is the Pioneer Centres, in which the public Danish National Research Foundation and four enterprise foundations — Carlsberg, Lundbeck, Novo Nordisk, and Villum — jointly fund world-class centres on transformative problems such as artificial intelligence and climate, the first opened in 2022. The episode shows both the latent power of coordinated foundation capital and how unusual, until very recently, such coordination has been.

International variation ↑

The Robert Bosch Stiftung in Germany holds roughly 94% of the share capital of Robert Bosch GmbH, one of the world's largest engineering companies — but almost none of the votes, which sit with a separate trustee entity, Robert Bosch Industrietreuhand KG. The foundation is the income recipient rather than the controller: nearly the inverse of the Novo structure, where the foundation holds modest capital but commanding votes. Annual distributions fund health, education, international understanding, civil society, culture, and science — mirroring the Danish model in the flow of funds, if not in the mechanics of control.

The Knut and Alice Wallenberg Foundation in Sweden, and the broader Wallenberg network, hold controlling stakes across Swedish industry — Ericsson, ABB, SKF, Atlas Copco, and others — and distribute income to Swedish universities and research institutes. The Wallenberg Foundations are Sweden's largest private research funder.

Tata Sons, studied alongside Novo Nordisk and Inter IKEA by McNulty and Thomsen (2026) in their analysis of foundation governance, is controlled by Tata Trusts — charitable trusts holding approximately 66% of Tata Sons' equity.⁴ The Tata model is rooted in Parsi charitable and civic traditions and is one of the clearest non-European examples of the enterprise foundation model operating at scale.

The Kavli Trust in Norway, established in 1962 by Knut Kavli, owns the Kavli food group outright and distributes its entire surplus to research, culture, and humanitarian causes — a far smaller case than Novo Nordisk or Bosch, but among the structurally purest: the company exists for no purpose other than funding the mission. It should not be confused with the unrelated US Kavli Foundation, which funds science institutes in astrophysics, physics, and neuroscience from a conventional endowment — wealth Fred Kavli extracted by selling his sensor company Kavlico. The two Kavlis are, in miniature, the distinction this primer turns on: an enterprise foundation on one side, a donor-endowed research funder on the other.

The Wellcome Trust illustrates what happens when the enterprise foundation structure ends. Founded from the estate of Sir Henry Wellcome, who owned Burroughs Wellcome & Co., it functioned as a true enterprise foundation for decades before selling Wellcome plc to Glaxo in 1995. The sale endowed what is now a diversified portfolio of roughly £40 billion (£39.9 billion as of September 2025). The enterprise foundation model ended; the research endowment it generated is its legacy.

The United States has no direct equivalent. American tax law and corporate governance norms have not produced a settled legal framework for self-owning entities without private beneficiaries. Hershey, through the Hershey Trust, comes closest structurally, but the trust's management of the company's ownership has been contested repeatedly — reflecting the absence of the stable legal architecture that makes Danish enterprise foundations durable across generations.

A taxonomy of enterprise foundation models ↑

Enterprise foundations are not a single thing. Different structures produce different governance challenges and different relationships between commercial ownership and research funding.⁵

What enterprise foundations make possible ↑

Enterprise foundations enable a specific kind of research investment that other funding structures struggle to sustain: patient capital on long horizons without commercialisation requirements.

Government grants are typically awarded on 3–5 year timelines, require preliminary results to justify renewal, and are sensitive to shifting political priorities. Venture capital needs exits within a fund's lifecycle. University endowments face spending rules tied to portfolio returns. Individual philanthropy is subject to the donor's evolving interests. Enterprise foundations with stable commercial income can fund research on 10–20 year horizons without requiring a pathway to a specific commercial outcome — and without organisations dying when they don't produce publishable results on a grant calendar.

Thomsen et al. (2018), using a decade-long panel of Danish companies, document what the patient capital claim actually means in practice.⁶ Compared to other firms, foundation-owned companies have more stable ownership structures, lower management turnover, more conservative capital structures with higher equity ratios, and higher scores on composite long-termism indices covering investment, employment, and financing practices. The survival advantage is striking: the probability of surviving beyond 40 years is 30% for foundation-owned firms versus 10% for other firms. These effects hold relative to family-owned firms as well as all firms, suggesting foundation ownership captures something structurally distinct — not simply the general family-business long-termism effect.

Schröder and Thomsen (2025), examining listed foundation-owned companies between 2003 and 2020, find that foundation-owned firms outperform family-owned and investor-owned peers on environmental and social dimensions by approximately 16% relative to the sample mean. The effect is consistent across five independent ESG databases and holds up under propensity score matching, difference-in-differences, and instrumental variable tests. Foundation-owned firms have higher employee satisfaction (roughly 6% above the mean), lower workplace injury rates, and lower environmental damage costs than comparable non-foundation-owned companies. More female board representation too: about 3.4 percentage points higher, a 19.5% increase over the average.

The outperformance is sharpest for foundations with explicitly charitable missions. Foundations with specific environmental or social objectives encoded in their charters show environmental scores roughly 42% higher than family foundation-owned counterparts — the legal commitment does real work. Foundations with both environmental and social objectives show aggregate ESG scores 13 points (26% of sample mean) higher than peers. Ownership form alone is insufficient; what the charter says is what drives outcomes.

The performance held under stress. In a difference-in-differences analysis around the 2008 financial crisis — an exogenous shock that prompted most firms to cut ESG spending — foundation-owned firms maintained their environmental performance by 16% of the mean and social performance by 16% of the mean relative to the control group, even when severely financially distressed. The non-distribution constraint that defines foundation ownership appears to function as a credible commitment to stakeholders precisely when commercial pressure is highest.⁷

One finding cuts against simple optimism: Schröder and Thomsen find no significant governance advantage for foundation-owned firms compared to other ownership types, attributing this partly to investor unfamiliarity with the dual-class share structures common in foundation ownership. Purpose and commercial resilience are real advantages of the model. Governance exceptionalism is not automatic.

The patient capital argument, taken alone, understates what enterprise foundations represent to the metascience conversation. Nielsen and Qiu's Science++ project frames the core question as: what social processes make different kinds of science possible? Funding cycles, career structures, incentive systems, and accountability norms don't just select which science gets done; they determine what kinds of science are possible at all. By that standard, enterprise foundations are not simply a well-resourced version of the same model — they restructure the social processes of research in ways that extend beyond time horizon.

A researcher working inside an enterprise-foundation-funded institution faces a different incentive structure from an academic grant-holder. There is no grant renewal at stake. Publication timing is not driven by a funding cycle. Institutional identity is not dependent on accumulating the reputation markers legible to a hiring committee. Multi-year coordinated work — tool-building, platform development, sustained effort on a single defined problem — is structurally supported rather than penalised by the career system. The research culture that results is different, not just longer-funded. This is what connects enterprise foundations to the current institutional reform conversation: they are not a novel design, but they are the most durable existing evidence that a different set of social processes for science can be maintained at scale, across changing conditions, for over a century.

Governing at a distance ↑

McNulty and Thomsen (2026) examine five foundation-owned enterprises — Novo Nordisk, Inter IKEA, Tata Sons, Rambøll, and the Norwegian food group Kavli — drawing on 35 interviews, foundation documents, and archival data.⁸ Their analysis centres on a structural paradox: foundation boards must be close enough to the enterprise to protect its purpose, but distant enough not to become its operating managers. Resolving this tension is the core governance work of the enterprise foundation model.

Their framework identifies three complementary elements of effective foundation governance.

Managerial distance. Foundation boards delegate business judgment to company boards rather than directing management themselves. The foundation holds the mission; the company board runs the enterprise. This separation allows commercial competitiveness without the foundation becoming a bureaucratic principal in day-to-day decisions. The danger is distance becoming disengagement — a foundation board so removed from operations that it can no longer evaluate whether the mission is being pursued.

Purpose commitment. Shared values between foundation and company, encoded in legal documents and in governance culture. McNulty and Thomsen describe this as "psychological as well as material ownership" — trustees who understand themselves as guardians of a purpose, not managers of an asset.⁹ Legal entrenchment of mission prevents drift when commercial pressures intensify. This is also what Schröder and Thomsen find drives the ESG outperformance: not ownership structure in the abstract, but the specificity of what the charter says.

Board competence. Trustees who understand both the business and the mission. Foundation boards cannot protect a purpose they don't understand, and they cannot maintain appropriate distance from operations without the competence to evaluate whether operations are actually pursuing the mission. Getting both dimensions right — business knowledge and mission knowledge — is the selection problem every enterprise foundation board faces across successive generations of trustees.

The governance model that emerges is neither passive stewardship nor direct management. It is purposeful ownership exercised at a distance, sustained by legal entrenchment and cultural transmission. Whether a given foundation achieves this balance is a function of board design, charter specificity, and institutional culture — none of which are automatic consequences of holding shares in a company.

These mechanisms are best understood through specific decisions. McNulty and Thomsen document four cases where foundation boards exercised their ownership role at critical moments:

Refusing the merger: Novo Nordisk and Serono (2004). Executive management at Novo Nordisk proposed a merger with the Swiss pharmaceutical company Serono that would have relocated headquarters to Geneva and diluted the foundation's controlling ownership. The Novo Nordisk Foundation declined — the business case was not found sufficiently convincing, and the merger would have endangered the stable ownership base specified in the foundation charter. Prescient: Novo Nordisk is now one of the most valuable European companies and among the world's largest pharmaceutical firms.

Patient capital in practice: Kavli and the Norwegian dairy market (1998–2007). When Norway deregulated its dairy market in 1996, Kavli entered via a joint venture to challenge the incumbent cooperative monopoly. The monopolist fought back fiercely; both original partners gave up. Kavli's foundation owner, the Kavli Trust, sustained 117 consecutive months of losses — nearly ten years — before Q-Dairies turned profitable. The foundation's purpose (benefit to society, long-termism as an explicit company value) provided the institutional backing to hold course. A company with impatient owners could not have absorbed a decade of deficits.

Purpose under pressure: Rambøll exits oil and gas (2022–2025). With the backing of the Rambøll Foundation, Rambøll launched a sustainability strategy in 2022 that involved exiting from oil and gas — one of its most lucrative business areas — and retraining a substantial share of its workforce. The decision was immediately tested by Russia's invasion of Ukraine and rising gas prices. The foundation held the course regardless. Rambøll stopped tendering for oil projects in 2022, gas in 2023, and committed to a full exit by 2025. "We want to decarbonise the world, and that's not compatible with supporting upstream oil and gas," the managing director for energy stated.

Protecting purpose: Tata dismisses Cyrus Mistry (2016). When Tata Sons' chair Cyrus Mistry began restructuring the group in ways the Tata Trusts viewed as prioritising short-term financial targets over the company's civic mission — selling off businesses, replacing seasoned executives, acquiring without consulting the majority shareholder — the Trusts voted to remove him. The dismissal was controversial in parts of the Indian business community. An event study showed no significant market reaction, suggesting shareholders priced in the purpose-protection function of foundation ownership.¹⁰

Critiques and open questions ↑

Conflict of interest. The Novo Nordisk Foundation owns one of the world's largest pharmaceutical companies specialising in diabetes and obesity treatments, and funds an extraordinary volume of research into metabolic disease. The question is not whether the foundation manipulates research — the evidence does not suggest this, and its scientific governance is generally well-regarded. The structural question is whether the research ecosystem it funds inevitably tilts toward problems commercially relevant to its enterprise, and whether its scale in Danish life sciences creates dependencies that make it difficult to fund work in adverse directions. The concentration of funding power shapes what gets studied regardless of intent, in ways that are difficult to audit from outside.

Mission drift at the foundation level. Eldar and Ørberg identify the primary risk for income-generating foundations as drift at the foundation level rather than in research design: boards become more focused on managing the commercial enterprise than on distributing its proceeds to the charitable mission. The legal permanence of foundation charters protects purpose from commercial pressure but does not protect against institutional inertia. A foundation that has grown into a sophisticated financial institution may lose sight of the scientific mission it was built to serve.¹¹

Non-replicability by design. Enterprise foundations emerged from conditions that are difficult to reconstruct through deliberate policy. Jurisdictions without legal frameworks permitting self-owning entities, cultures where founders expect to extract personal wealth from companies they build, and tax regimes without incentives for foundation ownership will not generate enterprise foundations by wishing for them. What may be transferable is the governance principle — ensuring commercial profit flows toward mission-driven purposes rather than private extraction — through IPO lockups, royalty structures, or foundation spinout mechanisms. But these are imperfect analogues that lack the structural permanence and legal clarity of the Danish model.

Governance opacity. Enterprise foundations face no private shareholders and often limited public accountability. Foundation boards can become self-perpetuating, closed institutions. The 1984 Danish Foundations Act and its subsequent revisions addressed some of this, but governance standards vary significantly internationally. The accountability mechanisms that exist for publicly traded companies or government-funded research bodies do not straightforwardly apply, and the foundations themselves have little incentive to invite external scrutiny of their decisions.

Geographic and sectoral concentration. Danish industrial foundations are heavily concentrated in pharmaceuticals, brewing, shipping, and industrial engineering. The research they fund reflects these industries' adjacencies. Research that doesn't map to any foundation's industrial history — fundamental physics, social science, humanities, certain areas of mathematics — may be systematically underfunded not because of active exclusion, but because of the cumulative effect of where the incentive structures point. Enterprise foundations diversify the research funding landscape relative to government grants and venture capital, but they diversify it in the direction of whatever industries happened to produce foundations, not in the direction of what science most needs.