# Investing in the Unknown and Unknowable **Richard Zeckhauser**  --- _When nobody can assign probabilities, the crowd leaves. That's the opportunity._ Zeckhauser starts where most investment theory stops. Standard finance assumes you know the possible states of the world and can assign probabilities to them. Uncertainty, in this framework, is just risk with wider confidence bands. Zeckhauser calls the territory beyond this UU: unknown and unknowable. Not just wide bands, but states you cannot enumerate and probabilities you cannot estimate. Will the Vietnamese government let you sell insurance at scale? Will a friend's software program find a use nobody anticipated? Will a 300-home development ten miles west of the city attract buyers? These are not reducible to probability distributions. They are genuinely unknowable. His positive claim: UU situations, precisely because they repel sophisticated capital, offer disproportionate returns to investors with the right profile. Three conditions create the opportunity. The situation is genuinely unknowable. Complementary capabilities are required to act on it, so the general market can't participate. And there's no reason to believe the other side of the transaction knows more than you do. When all three hold, the price you pay is likely to be too low. --- **The most useful maxim is the sidecar principle.** When someone with genuine complementary skills, vision, or connections is investing their own capital in a UU situation, invest alongside them. Venture capitalists earn extraordinary returns not because their models are better but because early-stage companies need their networks and judgment. The return comes from the combination of scarce skills and wise selection, not from information processing. For the rest of us, the practical move is to identify people whose capabilities give them an edge in genuinely unknowable territory and attach capital to their decisions. This is where the serial acquirer model gets interesting. [[Constellation]] and [[Lifco]] don't try to predict which of their hundreds of acquisitions will outperform. They apply a repeatable process, P/WC thresholds, cultural fit, permanent ownership, and let the portfolio compound. The sidecar principle operates at the subsidiary level: the group is investing alongside founder-operators who know their niches intimately. The group provides capital and governance. The founder provides the complementary knowledge that makes an otherwise unknowable bet investable. --- **Zeckhauser's framework for sizing bets maps directly onto the [[Kelly Criterion]].** The greater your expected edge, the more capital you should commit. But the corollary is just as important: when you can't estimate the edge, size down or find someone whose complementary capabilities let them estimate it better than you can. Kelly's formula requires inputs, win probability and payoff ratio, that UU situations refuse to provide. Zeckhauser's contribution is to explain what you do instead: assess your knowledge relative to the other side of the transaction, discount for the possibility that someone else has information you lack, and resist the natural tendency to over-discount ambiguity. Just because you don't know the risk doesn't mean others do. Sometimes it pays to act on the unknown when the odds are extremely favourable, even though there is some chance you're on the wrong end. The behavioural trap is symmetrical. Overconfident investors pile into UU territory assuming their models capture the unknowable. Ambiguity-averse investors flee from it, discounting far more than the actual risk warrants. Both errors are systematic. The first destroys capital. The second leaves it on the table. Zeckhauser's prescription is unsentimental self-assessment: review your past decisions, honest about the ratio of skill to luck, and build a working sense of where you have genuine [[Confidence]] in your own judgment versus where you're rationalising. --- **The essay [[Unknown and unknowable]] explores this through the lens of a customer health dashboard where the numbers are green and the churn is double.** Zeckhauser formalises the same insight from the investment side. The metrics measured one layer. The decisions that mattered, champion departures, competitive relationships, board-level mergers, sat in a layer the dashboard could never reach. Bridge, which Zeckhauser plays at a world-class level, is his analogy: hundreds of decisions per session with incomplete information, combined with the discipline of accepting good decisions that produce bad outcomes. That peacemaking skill, separating decision quality from outcome quality, is the temperament required for UU investing. The limitation of the paper is that it leans heavily on individual brilliance: Bill Gates, Warren Buffett, people with vision or complementary capabilities that make unknowable bets rational for them. The serial acquirer evidence suggests a different path. Process discipline, a small headquarters, permanent ownership, and [[Priors]] built from hundreds of transactions can substitute for individual genius. You don't need to predict which bet will work. You need a system that survives the ones that don't and compounds the ones that do. ---