Spending Policy and the Long Game for Endowments

Your investment policy statement almost certainly names a spending rate. Here's the harder question: where did that number come from?

For most foundations, the honest answer is that someone chose it a long time ago — from the tax code, from what a peer organization was doing, or from an advisor who has since moved on. It went into the document, and the document went into a drawer.

I'd like to take that number back out of the drawer, because the arithmetic underneath it has moved, and the direction it moved is not comfortable.

This piece is written for a particular reader: an organization with somewhere between a few million and a few tens of millions invested, a committee of three or four board members, and no investment consultant on retainer. If that's you, it's very likely nobody has ever run this arithmetic on your behalf. It takes about an hour, and you can do most of it yourself.

First, which rules actually apply to you

Two different legal frameworks govern charitable investment pools, and they impose very different obligations. Some boards aren't completely certain which one applies to them.

If you're not sure, your tax return usually settles it. A private foundation files Form 990-PF. A public charity files Form 990 or 990-EZ. That sounds like a filing technicality. It determines whether you have a legal spending floor, which is the single most consequential fact about your spending policy.

The 5% rule is a floor, and it was never built to be sustainable

If you're a private foundation, federal law requires you to distribute an amount equal to roughly 5% of the average fair market value of your investment assets each year. The amount computed for one tax year has until the end of the following year to go out the door, which gives you more timing flexibility than most boards use.

Three details inside that rule matter more than they look:

Grants count, and so does the reasonable cost of making them — due diligence, grantmaking staff time, the audit. It is not a 5% grants requirement; it is a 5% charitable-purpose requirement.

Investment management fees do not count — they're treated as a cost of managing the money, not as a charitable distribution. Your fees sit on top of the 5%, not inside it.

A separate excise tax applies — 1.39% of net investment income. It's a modest drag in asset terms, but it is one more thing the portfolio has to earn back.

Now add it up. If you intend to exist in perpetuity, the portfolio has to earn your 5% payout, plus the fees that don't count toward it, plus inflation, before the fund has held its ground in real terms. Call that somewhere between 8.5% and 9.5%.

No published forward-return forecast for a diversified portfolio is anywhere near that. Which produces a conclusion worth stating plainly:

A private foundation that distributes its legal minimum and calls itself permanent is, on current return expectations, shrinking in real terms every year.

That is not a governance failure, and I want to be clear that it isn't a criticism of anyone's stewardship. It is the design of the rule. Congress set the rate at 5% in 1976, in a very different return environment, and never indexed it to anything.

It's also worth saying that for a great many family foundations, gradual spend-down is the right answer and often the one the donor intended. The difficulty isn't distributing 5%. The difficulty is distributing 5% while assuming the fund will hold its purchasing power forever. Those two things have stopped being compatible, and the board that notices first is in much better shape than the board that notices in fifteen years.

If you're a public charity, you have more room than you may think

No federal minimum payout applies to a community foundation, church, school, or hospital endowment. What governs instead is state law — the Uniform Prudent Management of Institutional Funds Act, adopted in 51 jurisdictions including North Carolina — and UPMIFA changed in a way that a lot of old gift documents and board habits haven't caught up with.

Under the prior act, historic dollar value was a hard floor: the original gift plus every later contribution, each counted at the time it was made. If the fund fell below that number, you couldn't spend from it. UPMIFA removed the floor and replaced it with a standard — you may appropriate as much as you determine is prudent for the purposes and duration of the fund, acting in good faith and with the care an ordinarily prudent person in a like position would use.

Two consequences follow, and both tend to surprise boards.

The boilerplate in your older gift letters may not mean what it used to. Phrases like "income only" or "preserve the principal intact" still establish that a fund is perpetual. They no longer create a spending floor. Unless a gift instrument states a limitation specifically, that language doesn't displace your own prudence determination.

Because the mechanical test is gone, the deliberation itself now carries the weight. The act lists seven factors to weigh where relevant: duration and preservation of the fund; the purposes of the institution and of the endowment fund; general economic conditions; the possible effect of inflation or deflation; expected total return; the institution's other resources; and its investment policy. For a three-person committee, the practical version is that a short, dated memo in the minutes showing you considered these is worth more than a sophisticated portfolio nobody documented.

About fifteen states also adopted an optional provision under which appropriating more than 7% of fund value creates a rebuttable presumption of imprudence — meaning you'd have to come forward with an explanation rather than a challenger having to build the case from nothing. It is neither a cap nor a safe harbor. The official commentary describes spending nothing for three years and then 20% in year four to fund a building as potentially prudent, and separately notes that 6% might be imprudently high. The drafting committee's own view was that few funds could sustain a rate above 5%.

One point in your favor that almost nobody uses: the valuation method you actually adopt is binding, so long as it clears the statutory minimums — valued at least quarterly, averaged over at least three years. Write your method into the policy and it can't be second-guessed afterward.

Your hurdle rate, and the term your policy probably leaves vague

The identity is simple enough to belong on a single page of every investment policy statement:

Required return = spending rate + inflation + fees

Three terms. Two you already know precisely. The third is where policies go soft — and "preserve purchasing power" without saying purchasing power of what isn't a measurable standard.

The right index is the one that tracks what your money actually buys, and for most organizations it is not the CPI:

A scholarship fund buys tuition, which has its own well-documented and unhappy history.

A church or school endowment buys salaries, benefits, and building maintenance.

A grantmaking foundation buys its grantees' costs, which are mostly salaries — so the relevant index is closer to a wage series than to a basket of household goods.

The best-documented example of that gap is the Higher Education Price Index — HEPI, a cost basket built around faculty and staff salaries, benefits, facilities, and utilities. It has run above CPI in nine of the past eleven years. Before you conclude the gap is large, though, look at the size of it: about 0.55 percentage points a year over twenty-five years, and under 0.15 over the past ten, because the 2022–23 inflation spike briefly pushed CPI well above HEPI. In fiscal 2025 the gap ran a full point the other way.

So the reason to name your index isn't that one runs reliably hotter. It's that an unnamed hurdle can't be measured, and a hurdle you can't measure is one you can't hold yourself to.

Here's what the choice costs, at a 5% draw and 0.75% in fees:

Core CPI is the friendliest of the three and also the least defensible, since it strips out food and energy — which your grantees and your buildings very much still buy. On twenty-five-year averages the historical hurdle has been roughly 8.7%, against a realized twenty-five-year return of 6.6%.

What the forecasts actually say

What follows are third-party capital market assumptions, published by their authors. They aren't my forecasts, and I'm citing them as published:

BlackRock puts its Global 60/40 assumption at 7.23% over ten years.

Horizon Actuarial's adviser survey puts the average expected twenty-year return for a hypothetical multiemployer pension portfolio at 7.03%.

J.P. Morgan projects 6.4% over ten to fifteen years for a 60/40, rising to 6.9% with 30% in diversified alternatives added.

The published range is roughly 6.4% to 7.2%. None of it clears an 8% hurdle, let alone 9%.

Work backward from the most optimistic end and the sustainable spending rate lands a little under 3% net of inflation and fees. I'd read that as roughly 2.5% to 3% rather than a point estimate — the inputs are rounded and get revised quarterly. But every way I construct it, the answer comes in below 5%.

For scale, here's the resourced end of the field. The NACUBO-Commonfund Study of Endowments for fiscal 2025 covers 657 institutions and $944 billion: average return 10.9% net of fees, average effective spending 4.9%, ten-year average annual return 7.7%. Set that decade against the hurdle and it comes out approximately breakeven — on portfolios that were 86% in equities and equity-like strategies. Those institutions have investment staffs and consultants on retainer. It is worth knowing that they also ran close to even.

One more point, and it's the one that's hardest to sit with. That 7.7% came largely from multiple expansion — investors paying a higher price for each dollar of earnings, rather than earnings growing into the price. Which is exactly why the providers now forecast lower returns ahead. Two more strong years made the forward picture worse, not better. The return you just earned is not evidence you can keep earning it.

The decision underneath the spending rate

If the arithmetic holds — and I've shown the work so you can check it — then most charitable funds are spending more than their portfolios can reasonably be expected to support, and have been carried through by an unusually strong decade in equities.

For your board, that resolves into a question that is not really about investing at all: are you perpetual, or are you spending down? A great many foundations have never answered it out loud, which means they are answering it by default, one budget at a time.

There are three honest responses, and none of them is wrong.

Lower the draw. Available to public charities immediately; available to private foundations only down to the 5% floor.

Accept gradual erosion, deliberately, and record that you chose it. For a private foundation at the minimum, this is the default whether or not anyone has said so.

Convert to an explicit spend-down with a horizon — twenty years, twenty-five, a generation. Increasingly common, and it lets you spend far more on the mission while the people who care about it are still in the room.

There is a fourth that gets reached for, and it deserves caution: pursuing return the forecasts don't capture, which in practice means illiquidity or manager risk. It can be the right call. For a fund this size, though, the added fees, oversight burden, and liquidity constraints deserve hard scrutiny — they rarely solve the underlying mismatch, and this is the response most likely to be sold to you by someone who benefits from it.

What's hard to defend is a policy asserting permanence alongside a spending rate no published forecast supports, with nothing in the record showing anyone compared the two. UPMIFA imposes no record-keeping duty of its own; some states do. But in practice, the record of your deliberation is how prudence gets proved.

Four things worth doing, in order of effect

1. Name what your money has to buy

Pick the index deliberately, write it into the policy, and accept whatever hurdle it implies. A higher honest hurdle is far more useful than a lower vague one. This is a fifteen-minute decision that makes every subsequent one measurable.

2. Smooth on purpose, not by accident

About three-quarters of institutions — 74.3% in fiscal 2025 — spend a percentage of a moving average of market values, most often a three-year average. Independent simulation work finds smoothed approaches come closer to clearing the hurdle than spot-value methods, for a reason that's intuitive once you see it: you can't retroactively un-spend money you've already withdrawn, so the cost of an unsmoothed rule is asymmetric.

Your policy may borrow the Yale-style rule instead — last year's spending increased for inflation, blended with a target percentage of current market value. It's a sound rule. Note that Yale and Stanford each target 5.25% long-term, run it inside collars that cap how far the payout can move in a year, and employ investment staffs numbering in the dozens. Copying the formula doesn't copy the machinery behind it.

And be clear-eyed about what smoothing does. It stabilizes your budget. It does not raise returns. A long lag will keep distributing at yesterday's levels well into a decline, which is exactly when you can least afford it.

3. Watch the withdrawals that aren't in your policy

If I could put one number in front of your committee, this would be it. In fiscal 2025, only 83.5% of endowment withdrawals were distributions made within the institution's own spending policy. Roughly one dollar in six left outside the normal rule, and special appropriations above policy accounted for 5.3% of the total.

For an organization your size this rarely looks like recklessness. It looks like a one-time grant to a capital campaign, or covering a bad year in the operating budget, approved by good people for good reasons. But this is where spending discipline earns its keep: a carefully built policy stops being one the moment exceptions become routine. The problem is almost never the formula — it's the exception process. Your minutes will tell you quickly. If special appropriations appear year after year, your effective rate is your real policy, and the written one is a statement of intent.

4. Give the investment policy a review date and a named owner

This is the recommendation most specific to organizations without staff, and the one most likely to still be working in ten years. An IPS with no scheduled review is a document that gets read when something has already gone wrong. Put an annual review on the calendar, name the person responsible for putting it on the agenda, and write both into the policy itself.

While you're there, one benchmark for context: across the NACUBO institutions the endowment funded a median of 6.1% of operating expenses. If your fund covers substantially more than that, a spending cut is a budget event and needs to be planned as one. If it covers much less, you have more freedom to lower the draw than you may have assumed.

Five questions for your next meeting

If you take nothing else from this, take these:

Are we a private foundation or a public charity — and does everyone on this committee know which?

What inflation index does our policy name, and if it doesn't name one, what are we implicitly assuming?

What is our hurdle rate as a single number: spending plus that index plus fees?

Over the last five years, what share of our withdrawals fell outside the spending policy?

Have we ever decided, out loud and in the minutes, whether this fund is perpetual or spending down?

If you can answer all five from memory, you're ahead of most committees. If you can't answer three of them, that's your agenda for the next meeting.

If it would help to run your own numbers

None of this is a prediction, and none of it is specific to your organization — it's arithmetic built on other people's published assumptions, and your fund may look quite different.

If your committee would find it useful to see this run against your own portfolio, spending policy, and cost inflation, I'd welcome that conversation. It's an afternoon's work, and the board comes out of it with a number tied to your own circumstances rather than a sector average.

Thank you for reading this far. Questions about any part of it are always welcome.

---

Reid Culp
President & Founder, TAGStone Capital, Inc.