Losing Ninety Percent of the Time

Thirty-two losses out of forty-five trades. That is a failure rate above seventy percent, and it belonged to a portfolio that returned more than most people earn in a decade. The number does not add up — until it does, and then it rewrites what you thought you were measuring.

Outcomes are not distributed evenly. They pile up at the extremes. A few events carry almost all the weight, and everything else is noise wearing the costume of data. Most of your decisions, financial and otherwise, will resolve as small losses or smaller wins. Most of the time you are just paying tuition on the one decision that was going to matter. The mistake is not losing often. The mistake is treating all losses as equivalent — as if the score is what the game is about.

Jessica Livingston and her co-founders at Y Combinator funded hundreds of companies across early batches, most of which returned little or nothing. Airbnb alone accounted for returns that dwarfed the rest of the portfolio combined. The math was never about win rate. It was always about what happened when something went right at scale. One position, unheld by anyone with a normal appetite for cutting losses, did the work of a hundred ordinary bets.

The hour before you close out a position — the moment you are staring at a number that has not moved the way you expected — is where this either lands or it doesn't. The question is not whether you should tolerate being wrong. You will be wrong. The question is whether this particular loss is tuition or whether it is the thing itself, the one that could run. You cannot know for certain. But you can ask whether you are selling because the thesis broke or because the discomfort did. Those are different reasons that feel identical at three in the morning.

A portfolio full of right answers, each one modest, will almost never change your life. The thirty-two losses in that record were real losses. They just were not the point.