This series has now read the books of more than thirty endowed fortunes, from Girard's 194-year-old trust to a Powerball winner's Magic Valley fund to a billion that landed last year. Every one of them answers the same question differently: once a fortune exists to be given away, how should it be invested while it waits? The filings and policy documents answer in public, and lined up together they sort into seven distinct styles, each matched, well or badly, to the job the money has to do. This post maps them, with the returns and the risks, everything cited, nothing invented, and a standing caution: this is reporting on how these organizations invest, not advice about how you should.
Style one: the fiduciary balanced pool, for money that must pay forever
The clearest written specimen is the oldest. The Girard estate's trustees publish an investment policy that targets 50 percent equities, 10 percent international, 30 percent fixed income, and 10 percent alternatives, each with a band around it, governed by the Prudent Investor Standard's own language: managing "not in regard to speculation, but in regard to the permanent disposition of the fund." Managers are reviewed quarterly, judged over three-to-five-year horizons against per-asset-class benchmarks, and dischargeable for failure. The results, from the audited statements: roughly 10 percent in fiscal 2024, a year the S&P 500 returned 24.6 percent and the college endowment universe averaged 11.2 percent, and, the number that justifies everything, a loss of about $360 thousand on half a billion dollars in fiscal 2022, when equity markets fell hard. The school got paid every year of the three.
That is the whole logic of the style. A portfolio that must write checks forever cannot ride an index that drops by a quarter, because the deep down year forces selling corpus at the bottom to pay the bills, and that sequence is what actually kills perpetuities. Who it fits: any fund with a permanent, non-negotiable obligation, a school, an operating charity, a pension. The benchmark is not the market. It is the payroll, met since 1848.
Style two: the endowment model, for large perpetuities with patient decades
Step up in size and the balanced pool grows teeth: partnership interests, private equity, and land, illiquid assets that pay a premium precisely because they cannot be sold quickly. The Gates Family Foundation's filings show the type, a book dense with partnerships and real assets behind $634.6 million, fitting a funder whose stated priorities include Colorado land, water, and forests, so the portfolio literally resembles the mission. The college endowments this style is named for averaged 11.2 percent in fiscal 2024 while private equity returned 5.8 percent and venture 1.7 that year, which is the style's honest fine print: the illiquidity premium is real over decades and absent over quarters. Who it fits: funds above a few hundred million with investment committees, long horizons, and no need to liquidate in any given decade. Who it doesn't: anyone who might need the money soon, which is why you will not find it in the small family funds below. A living founder's fund can also simply pick: the Anschutz Foundation's schedule lists individual positions, $11.2 million of Enterprise Products Partners, Enbridge, Under Armour, an energy-tilted stock picker's book, which works on the same logic as everything else in this taxonomy, because the founder's own fortune stands behind the mission either way.
Style three: the single-stock concentration, for founders who never let go
Milton Hershey's deed put the chocolate company itself into the school trust, and today the Milton Hershey School trust holds roughly $17.4 billion including 81 percent of The Hershey Company's class B shares. One school owns a controlling interest in one of the world's most famous companies, and the deed makes the position nearly impossible to unwind. The style's virtue is alignment: the fortune's engine keeps running for the mission, and a century of chocolate profits built America's richest school. The style's risk is the obvious one, concentration, a single company's fate carrying a single school's fate, and Pennsylvania courtrooms have hosted the argument for decades. The Albertson foundation shows the arrival version: Kathryn Albertson's 1997 gift of $660 million in grocery stock multiplied the foundation's giving fourteenfold in one stroke, and diversifying such a gift afterward is itself a years-long project. And the Laura Moore Cunningham Foundation shows the style at family scale, held for generations: its latest schedule carries $82,958,601 of US Bank stock, seventy percent of the fund, in the foundation descended from the family that co-founded Idaho's first bank in 1863. Who it fits, by design: nobody, which is why the style survives only where a founder's document forces it, a family holds by conviction, and the company cooperates by enduring.
Style four: Treasuries and funds, for the quiet family scale
Below about $50 million the filings stop showing exotic holdings, and for good reason. The Bews Foundation's return shows government obligations as visible line items behind its $22.5 million; the Seagraves foundation runs the same way behind $30.3 million. At this scale a fund has no investment committee, no staff, and a payout obligation near 5 percent that ordinary yields and a balanced fund can meet without heroics. The style is unglamorous and structurally correct: complexity has fixed costs, and on $20 million those costs eat the mission. Who it fits: nearly every family foundation that will ever exist, which is most of them. The quiet Idaho funds this series profiled, stocking school pantries and funding scholarships, do their work precisely because their portfolios are boring.
Style five: the ignition and pass-through postures, for money in motion
Two of the newest filings show what investing looks like when the money doesn't stay. The Hexagon Foundation received $503.8 million in contributions in 2024 and granted $89.6 million the same year, money arriving and departing at velocity, which means short duration and liquidity, not markets. The OneRoot Foundation holds $1.1 million but granted $6.2 million during the year, a pure conduit, refilled and emptied annually. And the Avenir Foundation is the ignition specimen at rest: a $1.24 billion excess of revenue in one fiscal year, the fortune landing all at once, with grants of just $2.7 million while the fund decides what it will be. Who these postures fit: funds where the investment question is secondary to the flow question. The risk isn't market risk; it's governance risk, whether the pipe's controllers keep pointing it at the mission.
Style six: the spend-down, where the portfolio shrinks on purpose
Greg Carr's foundation disbursed $11.9 million against $53.8 million in assets in its latest year, better than twenty percent, because his $100 million-plus pledge to Gorongosa National Park runs on the park's clock, not perpetuity's. Holland Ware's timber fortune pays out around fifteen percent. And the Julie Ann Wrigley Foundation's filing shows the style completed: $256,041 remaining after the gifts that built Arizona State's sustainability institute went out the door, final checks to Sun Valley arts organizations. Spend-down inverts every rule above: as the horizon shortens, the portfolio de-risks toward cash, exactly the duration matching this series teaches for households, run in reverse. Who it fits: mission-sprint funds whose problem cannot wait for compounding, and founders who would rather see the work than endow the institution. The risk is the honest trade: nothing outlives you, which is sometimes the point.
Style seven: the pooled machines, where donors choose and scale invests
The donor-advised platforms run a different question entirely. DAFgiving360, the renamed Schwab Charitable, reports $41.1 billion across accounts whose donors pick from investment pools while the money awaits their grant instructions; WaterStone runs the same machinery at $746 million for faith-aligned givers. The community foundations pool differently: the Denver Foundation's century-old trust holds $1.32 billion across thousands of donors' funds under one investment office. In all three, the individual giver outsources the investment style entirely, which is the style: professional pools, index-like breadth, administrative ease. Who it fits: donors who want giving without running an institution. The structural trade this series stated in Volume One stands: a foundation is a purpose that outlives you, enforceable in court; a donor-advised account is a wallet that survives only as long as someone keeps choosing.
The returns table, read honestly
Across every style, the fiscal 2024 comparators line up like this, all sourced above: the S&P 500 returned 24.6 percent; the endowment universe averaged 11.2; Girard's balanced trust made about 10; private equity returned 5.8 and venture 1.7. Read naively, everyone should have owned the index. Read structurally, the way this series reads everything, each return is the price of a different insurance policy. Girard's ten bought a flat 2022 and an uninterrupted payroll. The endowments' eleven bought decades of illiquidity premium still maturing. Hershey's return is whatever chocolate does, by covenant. The small funds' yield bought simplicity that keeps a two-person board honest. And the payout side runs its own spectrum, from Avenir idling near zero at ignition, through the 5 percent statutory cruise most of the shelf holds, to Ahimsa's seven, Ware's fifteen, Carr's twenty-plus, and the spend-downs' everything: velocity, like allocation, turns out to be a written choice, visible in every filing.
The method that works, then, isn't a portfolio. It's a sentence this series has now found in every durable set of books from 1831 forward: match the money to the job, in writing, and let strangers be able to check. Girard's trustees, Penrose's orchard, the Powerball fund in Twin Falls, and the Walton grandsons' mountain-town foundations disagree about nearly everything an investment committee could argue over, and agree completely about that.
Related reading
- The Richest American Died in 1831. His Money Is Still Following Orders.: the Girard policy and returns, with the household playbook.
- The Quiet Shelf, Volumes One and Two: the thirty filings behind this taxonomy.
- Bill Daniels Repaid Debts He Didn't Owe: the lineage from Franklin to Hershey to Daniels, and how to pull any filing yourself.
- The Ledger Lessons: the seven laws underneath all seven styles.
Fact-check notes and sources
- Girard (the 50/10/30/10 targets with bands, the quoted Prudent Investor Standard language, quarterly reviews and three-to-five-year horizons, the roughly 10 percent fiscal 2024 result as author's arithmetic from $51.0 million net investment income on about $501 million, and the approximately $360 thousand fiscal 2022 loss): the Board of Directors of City Trusts Investment Policy and the estate's audited financial statements.
- The fiscal 2024 comparators (the S&P 500's 24.6 percent, the 11.2 percent endowment average, private equity's 5.8 and venture's 1.7): the NACUBO-Commonfund Study of Endowments as summarized in its published coverage.
- Hershey (the roughly $17.4 billion, the 81 percent of class B shares, and the deed's constraints): Wikipedia, "Milton Hershey School" and Hershey Trust Company, attributed, with the spending litigation per ProPublica's investigation. Albertson's 1997 stock gift and fourteenfold jump: the foundation's history.
- All foundation figures (Gates $634.6 million and its partnership-and-land schedules; Bews $22.5 million and Seagraves $30.3 million with visible government obligations; Hexagon's $503.8 million in and $89.6 million out; OneRoot's $1.1 million against $6.2 million granted; Avenir's $1.24 billion revenue excess and $2.7 million granted; Carr's $11.9 million disbursed on $53.8 million; Ware's $55.4 million on $375.3 million; Wrigley's $256,041; DAFgiving360's $41.1 billion; WaterStone's $746.4 million; the Denver Foundation's $1.32 billion): read directly from each organization's most recent public Form 990 or 990-PF, as itemized with links in Volume One and Volume Two of the Quiet Shelf; payout percentages are the author's arithmetic from those figures and are labeled approximate.
- The Cunningham US Bank line ($82,958,601) and the Anschutz stock positions: reproduced from those foundations' most recent Form 990-PF schedules; no corporate-genealogy claims are made about the 1863 bank.
- Carr's Gorongosa pledge: Gorongosa's biography and Wikipedia, "Gregory C. Carr", attributed. Wrigley's ASU gifts: ASU's account. The DAFgiving360 rename: Schwab's announcement.
This post is informational and journalistic, describing how the named organizations report investing their assets. It is not investment, tax, legal, or philanthropic advice, and no style described here is a recommendation. All organizations and individuals are discussed from public records and their own publications as nominative fair use, with no affiliation implied and nothing endorsed by them.
