The method, step by step

1. Agree on return expectations for each asset class. Use a long data history. He used 20 years.
2. Build four or five model portfolios that could reach the return you need.
3. Measure the risk of each one. Use standard measures such as value at risk, and run each through past crises.
4. Pick the risk level the board can live with. That choice drives every later decision.

What return does an endowment need?

For an endowment, the target is usually spending plus inflation. Dartmouth's FY2025 report states the goal directly. A spending distribution of about 5% plus inflation of 2% to 3% works out to a long-term return goal of at least 7%. Brown's FY2025 report uses a higher number, combining a 5.5% high-end payout with a 2.9% inflation measure for a hurdle of 8.4%.

Your number will depend on your own spending policy. Write it down and put it at the top of every committee packet. When a manager pitch promises returns far above that hurdle, the committee has an easy first question. What risk are we taking to get there?

Stress tests that boards understand

Abstract risk numbers put trustees to sleep. Named scenarios don't. Ask staff to show what each model portfolio would have done in a sharp equity crash, a long flat market and a rate spike. Then ask the harder question. Could we still meet our spending policy?

The hardest part

The CEO I interviewed said the toughest job is killing the "hypnotic focus on promised performance." Products get sold on hope, not risk. A committee that maps risk first is much harder to sell to.