Every investment process is a compressed theory about how the world works, and most of them are inherited rather than derived. That is not a criticism. Inheriting a process from people who survived several cycles is usually wiser than inventing one. The difficulty arrives when the conditions that made the inherited process work quietly stop holding, and the process keeps producing confident answers anyway.

Managers describe the experience consistently. Nothing breaks. The model still runs, the outputs are still plausible, the committee still reaches a decision. What changes is the hit rate, gradually, in a way that is statistically indistinguishable from bad luck for approximately as long as it takes to do real damage.

The assumptions nobody writes down

The most dangerous assumptions are the ones never stated, because a stated assumption gets revisited. Cheap and abundant capital was one. Predictable supply chains was another. So was the idea that a business could be built for one market and sold into all of them, which underwrote a great deal of the last two decades and is now qualified almost everywhere.

We wrote down every belief the process depended on. It took a week and there were forty-one. Nine of them were still true.

A fund partner

The managers who have handled the transition well share a specific and unglamorous habit: they maintain an explicit list of the conditions their strategy requires, review it on a schedule, and treat a broken condition as a trigger for work rather than a topic for discussion. It is closer to preventive maintenance than to insight.

Slower, smaller, more specific

  1. Narrow the universe. Broad mandates were a bet on legible global markets; that legibility is what has degraded.
  2. Underwrite the downside case in operational detail, not as a percentage haircut.
  3. Extend the holding period assumption before extending the return assumption.

None of this is novel. It is, in fact, conspicuously old — the practices being rediscovered would be familiar to an investor of the 1970s, which several of the people making the argument point out with some amusement. The novelty is in how thoroughly they had been forgotten, and how quickly the forgetting happened.