Notes from the Road: Bayes and the Boreen

Investors have spent decades watching patterns hold. Long-term stocks rise. Bonds move opposite equities. Certain correlations persist. Yet these consistencies may not be laws of nature, but artifacts of a specific set of world conditions that nobody actively questions. What happens when those conditions change and investors discover they've been building portfolios on unstated assumptions rather than timeless truths?

• Historical consistency can mask hidden assumptions. A pattern that repeats for decades feels permanent, but it's often held together by beliefs about how the world works that are never examined.

• Experience becomes a trap when the world changes. Investors shaped by a major market crash often spend careers fighting the last war, treating singular events as permanent features of markets.

• When priors break down, updating becomes impossible. New information so far outside expectations creates paralysis instead of learning. The stakes feel too high to test assumptions experimentally.

• Comfort with a model prevents questioning what's beneath it. We rely on correlations that "work," but we rarely ask which unstated facts about the economy or policy or human behavior those correlations actually depend on.

• Owning less of things with unstable foundations might matter more than owning less volatility. If the assumptions supporting a prediction model are fragile, the real risk isn't daily price swings but the day the model breaks entirely.