January 1, 1970

How University Endowments Work: A Clear, Honest Explanation

Aerial view of a university campus representing the enduring nature of an endowment

Harvard's endowment traces to 1638, when a minister named John Harvard bequeathed his library and half his estate to a fledgling Massachusetts college. That gift started something. Nearly four centuries later, $944.3 billion sits in endowment funds across 657 American universities — a sum larger than the GDP of most countries. Yet most people have no idea how it actually works, or why Harvard can't just flip a switch and make tuition free.

What a University Endowment Actually Is

The word gets thrown around like it simply means "a school has money." It means something much more specific. An endowment is a collection of donated assets — cash, securities, real estate, art — invested permanently, where only the returns get spent and the principal stays intact.

Think of it like a fruit tree. You plant it, and every year you harvest what grows. You never cut the tree down. The goal is that your grandchildren can still harvest from it a century from now.

A $1 billion endowment generating 5% annually gives a university $50 million per year, forever, without touching the original gift. That's the logic — not a savings account to draw down, but a permanent income engine. The concept is built on perpetuity.

Harvard's endowment now stands at $56.9 billion (fiscal year 2025), spread across roughly 14,765 individual funds. About 80% of that total is legally restricted to purposes specified by donors. A scholarship endowed in 1962 still pays for a student's tuition today. The gift doesn't expire.

The Four Types of Endowment Funds

Not all endowment money behaves the same way. There are four main categories, and the differences matter.

True endowments (also called permanent endowments) are the core: donated funds that must be held in perpetuity. The principal is legally untouchable. Only investment returns flow out for spending.

Quasi-endowments are funds the university board has chosen to treat like endowments, but without the legal permanence. The school could, in theory, spend the principal if the board voted to. Few do, because donor relationships depend on good faith.

Term endowments are the rare exception — gifts specifying a time limit, after which the principal can be spent. A donor might give $5 million to fund a research program for 20 years. When the term ends, the fund is gone.

The final distinction cuts across all three categories: restricted versus unrestricted funds. A restricted fund can only be spent on whatever the donor designated — cancer research, need-based aid for engineering students, music instruction. An unrestricted fund gives the university flexibility to deploy money wherever it needs to go.

The catch: if a donor restricted a fund to something that becomes obsolete or impossible, the university can't just redirect it. Courts get involved through a legal doctrine called cy-pres. It's slow and expensive, and it rarely results in the broad reallocation critics imagine.

How the Money Gets Invested: The Yale Model

Before 1985, most university endowments looked like dull pension funds — roughly 60% stocks, 40% bonds, focused on U.S. markets. Then David Swensen took over the Yale endowment with about $1.3 billion and changed everything.

Swensen and his longtime collaborator Dean Takahashi bet heavily on alternatives: private equity, venture capital, real assets, and hedge funds. The thesis was that illiquid assets (those you can't sell tomorrow) command higher returns precisely because most investors avoid them. Universities, with their long time horizons and no need for monthly liquidity, were perfectly positioned to collect that premium.

The Yale Model proved that diversification isn't just about owning different stocks — it's about owning asset classes that behave differently under stress, across decades.

Under Swensen's leadership, Yale's endowment grew from $1.3 billion to $41.9 billion, compounding at roughly 13.1% annually over 35 years. The rest of the institutional world followed his lead.

By fiscal year 2025, according to NACUBO's study of 657 institutions, alternatives represented 55.7% of the average university endowment portfolio. Here's how it breaks down:

Asset Class	Average Allocation (FY25)
Private equity	17.1%
Marketable alternatives (hedge funds)	16.1%
Venture capital	11.7%
Real assets	10.8%
U.S. public equities	13.7%
Fixed income	10.7%
Other / international equities	~19.9%

Smaller endowments can't fully replicate this model. A $30 million college fund doesn't get invited into top-tier private equity deals with seven-figure minimums and decade-long lockups. That access gap creates a real performance gap, and it compounds over time.

The Spending Formula: Why Universities Can't Just Write Checks

This is the part that trips people up most. If Harvard has $56.9 billion, why aren't all students there for free?

The spending rule is the answer — and it's more constrained than you'd expect. Most universities spend between 4% and 5% of their endowment per year, applied to a moving average of the endowment's market value over the past several years (often 12 or 20 quarters). The Uniform Prudent Management of Institutional Funds Act (UPMIFA) governs this at the federal level, requiring institutions to balance current spending against preserving purchasing power for future generations.

The averaging matters enormously. Rather than spending 5% of whatever the endowment is worth today, schools apply the rate to a historical average. If markets drop 30% in a given year, the formula prevents a sudden collapse in university funding. If markets surge, it prevents an overspending binge that leads to painful cuts when reality catches up.

NACUBO's FY25 data shows the average effective spending rate was 4.9%, up slightly from 4.8% in FY24. Applied across $944.3 billion in total endowment assets, that generated $33.4 billion in total withdrawals during fiscal year 2025 — an 11% jump from the prior year.

Here's the non-obvious part: that spending rate has to clear inflation too. If a university spends 5% but inflation runs at 3%, the endowment grows by only 2% in real terms. Over 50 years, that slowly erodes purchasing power. The 10-year average return across participating institutions was 7.7%, which leaves room for spending and inflation — but not a lot of room for error.

Where the Money Actually Goes

The public image of endowments is elite universities sitting on hoards of cash while students drown in debt. The reality is more specific, and the numbers tell a different story.

According to NACUBO's FY25 data, 47.4% of endowment spending goes directly to student financial aid — by far the largest single category. Academic programs and research take 17.7%, endowed faculty positions get 10.8%, campus facilities receive 7.6%, and the remaining 16.6% covers everything else.

At Harvard, the endowment contributed more than one-third of the university's total annual revenue in fiscal year 2025, and funded $784 million in scholarships and financial aid. Students from families earning under $85,000 pay nothing to attend. Families earning up to $150,000 pay less than 10% of their income. That's endowment money working as intended.

The Association of American Universities reports that research universities provide six times more institutional financial aid than students receive from federal sources. That's not a rounding error — it reflects a completely different funding model from what most people assume.

The restrictions still apply. A university can't redirect a restricted fund for pediatric cancer research to cover utility bills, no matter how tight the budget gets. So when critics say schools should "just spend the endowment" to solve a specific problem, they're usually ignoring that most of the money is already earmarked. There's no free lunch in endowment accounting.

The Endowment Tax and What It Actually Means

Since 2017, the federal government has taxed net investment income at private universities with large endowments. Congress expanded this in 2025 into a tiered structure: the wealthiest institutions now face rates of 8% on investment returns. Princeton, Yale, and MIT fall into that top bracket. Harvard, Stanford, and others face 4%.

This is a meaningful reduction in capacity. An 8% excise tax applied to a year with 10% gross returns effectively knocks net returns down to roughly 9.2% after tax — money that would otherwise compound or be paid out as financial aid.

I'll say clearly what the data implies: if the stated goal is to help students afford college, taxing the primary fund that pays for scholarships is a structurally odd mechanism. The AAU points out that university endowments function similarly to private charitable foundations — entities not typically taxed on investment returns for the same reason. The tax raises real revenue but reduces the endowment's ability to do exactly what critics want it to do.

New gifts to endowments also declined in FY25, dropping 9.2% to $14 billion from $15.4 billion the prior year. Institutions with assets under $50 million saw a brutal 26.5% drop in donations. The public conversation focuses on elite schools, but the median endowment is $253.6 million, and more than a quarter of NACUBO participants have assets under $100 million. For those schools, the endowment model is far more fragile than the Harvard headlines suggest.

Bottom Line

University endowments are permanent investment pools, not bank accounts or rainy-day funds. The principal stays invested indefinitely; only the returns get spent, and even those are heavily directed by donor agreements.

The spending rate (typically 4–5% per year, applied to a multi-year moving average) is the mechanism that keeps endowments alive across generations — not an arbitrary policy preference, but a legal and actuarial discipline.
The Yale Model reshaped how large endowments invest, shifting the average portfolio to 55.7% alternatives by FY25. Access to top-tier private equity and venture capital remains a real advantage for large endowments that smaller schools can't match.
Most endowment spending is legally restricted by donor intent. Redirecting funds at will isn't possible — the restrictions are binding agreements, not suggestions.
The endowment tax creates a direct tradeoff: excise taxes paid to the federal government are dollars that don't go to scholarships or research. Whether that tradeoff is worthwhile is a policy question, but the mechanics are clear.
For most universities, endowments fund somewhere between 10% and 35% of operating expenses. Harvard sits at over a third. Smaller schools with modest endowments operate with far less cushion — and that's the part of the story that rarely makes the news.

Frequently Asked Questions

Can a university legally spend its entire endowment?

For true (permanent) endowments, no. The Uniform Prudent Management of Institutional Funds Act requires institutions to preserve the purchasing power of these funds in perpetuity. Spending down the principal would violate both the law and the donor agreement. Quasi-endowments are technically spendable by board decision, but doing so typically signals financial distress and damages donor trust permanently.

Why does Harvard still charge tuition if it has a $56.9 billion endowment?

Because roughly 80% of that endowment is legally restricted to specific purposes set by donors. A fund designated for medieval manuscript preservation can't be redirected to cover tuition. Harvard does use unrestricted and financial-aid-designated funds aggressively — families earning under $85,000 pay nothing — but "just spend the endowment" glosses over the binding nature of those donor restrictions.

What made the Yale Model so influential?

David Swensen's core insight was that illiquid assets — private equity, venture capital, real estate, hedge funds — earn a return premium over public markets because most investors can't tolerate being locked in for years. Universities, which have no obligation to return capital on demand, were structurally well-positioned to collect that premium. The 13.1% annualized return over his 35-year tenure made the argument impossible to ignore.

Is there a common misconception about endowment size and wealth?

Yes. People often assume a large headline endowment number means a school is sitting on freely spendable cash. In reality, most of that figure is restricted, illiquid (tied up in private equity or real estate funds), and already generating income committed to specific purposes years in advance. The usable "float" in any given year is the spending distribution — in FY25, that averaged 4.9% of assets.

How does the endowment tax affect financial aid?

Directly. The excise tax on net investment income reduces the return available for distribution. A school paying 4% or 8% of investment income to the federal government has that much less to either reinvest for growth or pay out as scholarships. The AAU estimates this reduces institutional capacity for exactly the student support that critics of large endowments say they want to see more of.

What happens to an endowment fund if the designated purpose becomes obsolete?

Universities can petition a court to modify a restricted fund's purpose through a legal doctrine called cy-pres, which essentially means "as close as possible." A court might allow a fund for "typewriter repair instruction" to be redirected to general technology education. The process requires legal filings, court approval, and often notifying the donor's estate — it's not fast or cheap, which is why many outdated restricted funds simply sit underused.