Some years ago I sat on the finance committee of a small nonprofit. Our job was to look after its endowment, which is really just a pot of invested money that sends the charity a check every year and is supposed to keep doing it forever.

We spent one meeting on the question every committee like ours eventually asks. The market feels expensive and we are nervous, so what should we do about it?

The adviser who managed the money had an answer ready. He wanted us to move about five percent of the portfolio into alternatives, the industry's word for anything that is not a plain stock or bond fund. Hedge funds, private equity, private credit. Some of it we could have sold any day. Some of it would have locked our money up for years.

The pitch was reasonable and well presented. These strategies are supposed to make money in ways that do not depend on the stock market cooperating, which is exactly what you want when you are worried about the stock market. I asked to see the analysis before we voted, and what it showed is the reason I am writing this.

The Score Improved. The Goal Got Worse.

The change would have lowered our expected return a little, and lowered how much the portfolio bounced around by proportionally more.

There is a standard scorecard for that trade called the Sharpe ratio. It compares how much you earned against how much the value bounced around, which the industry calls volatility. Higher is better. Ours would have gone from 0.38 to 0.39. That is an improvement, and it was the number the pitch rested on.

Then I looked at the number that actually mattered to us. The endowment exists to send money to the organization every year, and to keep doing it long after everyone in that room has rotated off the board. When I ran the change through that lens, the odds of making those payments over the long term went down. So one number said we were better off. The number we cared about said we were worse off.

The reason is not complicated. A portfolio that owes a payment every year needs growth to survive the payment, and giving up return to smooth the ride makes the ride smoother and the destination less likely. The Sharpe ratio does not know we have a mission. It just compares return to bumpiness, and you can take less of both and still score better.

This is not only an endowment problem. The industry measures the thing that looks like the goal instead of the goal. You do not have a Sharpe ratio target either. You have a retirement to pay for.

The fees made it worse. The alternatives charged several times what we paid on everything else, and the private funds took a cut of any profits on top. So we were being asked to pay considerably more, gain a hundredth of a point on a ratio, and be less likely to fund the charity. I have written before about how complexity quietly eats returns. This was a clearer case than most, because even before you count the fees, the complexity did not do what it was supposed to do.

There Are Only So Many Things You Can Actually Buy

Step back and ask what you are really buying. At the bottom of every investment product in the world is a short list of ingredients. Stocks. Bonds. Cash. Real things like property and commodities. And contracts written on top of those. Options and futures are bets on what those things will do. That is close to the whole list.

A complicated fund is a recipe, not a new ingredient. What you pay for is the arrangement. How much of each thing to hold, how much borrowed money to use, when to get out. Sometimes that is worth the money. But look through to what the fund owns and you are still mostly holding stocks and bonds. What you own is not as different as the presentation makes it look.

What is different is the machinery, and every layer of it is another way to be wrong. A stock fund owns stocks. Add borrowed money, a hedge that needs constant adjusting, and a rule that sells automatically at a certain point, and each one is another system that has to work. Those systems also interact, which is hard to see until they are all under stress at the same time.

Leopold Aschenbrenner is a recent example. He is a former OpenAI researcher who wrote a widely read essay about how fast artificial intelligence would reshape the economy, then turned the argument into a hedge fund. For a while he was spectacularly right. The fund was up more than 400 percent in the first half of this year and ran something like $45 billion at its peak.

Getting returns like that required borrowing. At times the fund held as much as four times its own capital in AI-related stocks. Then in July the trade turned. Lenders started asking for their money back, the borrowed money multiplied the losses just as much as it had multiplied the gains, and by the end of the month the fund had sold essentially its entire public stock portfolio. Assets fell from about $45 billion to about $10 billion.

That story is not over. The fund is still running and still investing, the thesis may well prove right, and Aschenbrenner is young and clearly talented, so we will probably be hearing his name for a long time. This is not a story about anyone behaving badly. The point is smaller than that. Once you add borrowed money and the rules that come with it, things can go wrong for reasons those additions created. You have to be able to survive being wrong for a while, and that much leverage does not leave you long.

None of this means these strategies do not work. They work a lot of the time, which is why they are easy to sell. It means a simple mix of stocks and bonds has almost no way to surprise you. It will go down sometimes. But it will go down for reasons you already understood when you bought it.

What Kelly Johnson Handed His Engineers

I am not an engineer, but there is an idea from engineering I keep coming back to.

Kelly Johnson ran Lockheed's Skunk Works, the group that built the U-2 and the SR-71, and he is generally credited with the KISS principle. He taught it by handing his engineers a small set of tools and telling them the jet they were designing had to be repairable by an average mechanic, in the field, under combat conditions, with only those tools. The test was not whether it worked in the hangar. The test was whether a tired mechanic could fix it at three in the morning without ever having met the people who designed it.

An endowment has the same problem. Board members turn over. The committee ten years out will not include me or the adviser who made the pitch, and every product we add is one more thing they have to understand well enough to know when to sell it. Whoever understood the product will be gone long before the product is, and that cost never shows up in the analysis.

The Boring Answer Nobody Pitches

I recommended something else. If we were really worried about the stock market, sell a few percent of our stocks and buy bonds instead.

It costs almost nothing, and you can reverse it any day you want. Any board member could explain it to a donor in one sentence. And it cuts the risk we said we were worried about by more than a five percent slice of anything could, because a five percent position cannot move a portfolio much no matter how clever that position is. The stocks and bonds were always going to do the heavy lifting.

Why Everyone Keeps Reaching for the Complicated Answer

I do not think that adviser was acting in bad faith.

Finance people find this stuff interesting, and I include myself. Reading about what a clever hedge fund manager is doing is more fun than reading about an index fund, and it is easy to confuse a strategy being interesting with a strategy being useful to you.

The other reason is structural. Investors have moved enormous sums into index funds that charge almost nothing, and that money pays nobody much. The revenue went to the products that still carry real fees. You do not need bad intent for incentives to shape advice, and the problem is worst when the simple answer would have worked.

That does not make the recommendation negligent. It was defensible, and a committee that took it would probably have been fine. My objection is narrower. The change would not have moved the portfolio much in either direction, the added cost was certain, and the new ways to go wrong were unknowable. Paying a certain cost for a small and uncertain benefit is a bad trade, no matter how good the pitch is.

The Question to Ask Before You Add Anything

Before you add a product to your portfolio, name the goal it improves, and say the improvement in terms of that goal. Not "it lowers volatility." Something closer to "it makes me more likely to retire at sixty-five."

If the honest answer is that it moves a ratio slightly, ask what the ratio is for. If nobody can connect the answer back to something you actually care about, the product is not solving your problem. It is solving someone else's.

We voted it down. It was the least impressive decision the committee made all year, and I would make it again.