What I have seen go wrong

In the book I wrote about institutional investors and family offices that trailed their peers and didn't realize it. Their CIO or outside consultant was choosing which numbers to show. Even with good intentions, investment teams can fall prey to their own biases and hide real risk or weak results.

It gets worse when pay is tied to performance and the team picks the yardstick.

Lock the yardsticks in the policy

The investment policy statement should name:

1. The return objective. Usually spending plus inflation over a long period.
2. The policy benchmark. A blend of indexes that matches your target allocation.
3. A simple market benchmark. Many institutions use a global stock and bond mix. Brown, for example, compares itself with a 70% stock and 30% bond portfolio. Dartmouth uses a 60/40 mix.
4. A peer comparison. Such as a published endowment universe for your size band.
5. The time periods that count. One year is noise. Five, ten and twenty years tell the story.

Ask for the same table every time

Each quarter, the committee should see the same table in the same order. Actual return, policy benchmark, simple benchmark and peer median, across the same periods. If anything changes, staff should explain why in writing.

The cheap alternative

The simple stock and bond mix is the low-cost choice the board could always make. If a complex portfolio with high fees trails it over ten years, the board should know. Some years it will. Dartmouth's FY2025 report, for example, shows its 10.8% return below a 60/40 benchmark's 12.0% for that year, and well ahead over ten years. That is exactly the kind of honest table a board should get.