On conviction, models that miss the moment, and the LP question almost nobody answers right.
If you knew you could walk away with $33,000 by betting $100, would you? Would you double or triple down on a leader who faces bankruptcy, but could earn you billions?
Last week, two of them paid off. One overnight. One 20 years in the making. Both tell us a lot about conviction.
If you bet on the Knicks when they were down 29 points with Game 4 of the NBA finals slipping away, just before OG Anunoby tipped in the winning basket with 1.2 seconds left.
Odds of that happening? Less than 1%. Biggest comeback in NBA Finals history.
SpaceX? Going public at an expected $1.7 trillion valuation after more than one near-death experience. The dedication of those who stuck with it, whether engineers or investors, is minting millions.
When Models Get It Right, and Still Miss Everything
Nobody eyeballed the Game 4 Knicks win. Odds come from models that learn from thousands of scenarios and made best guesses, basically the same way we model portfolio risk.
They look precise and they feel authoritative.
But the Spurs went from the best shooting half in Finals history to bricking everything in sight in about twelve minutes. No model built on normal behavior sees that coming.
In 2008, SpaceX struggled with three consecutive rocket launches. Tesla was days from bankruptcy, and many told Elon Musk to let one company die to save the other.
But Antonio Gracias, a leading investor, did the opposite. He lent Musk money, kept investing in SpaceX through the failures, joined both boards, and never stopped adding capital when everyone else was heading for the exits.
The LP’s Hardest Question: When Do You Add Capital
Gracias wasn’t just right once. He kept being right by doing the thing most investors cannot bring themselves to do: adding capital when everything looks broken.
Most LPs pull out when a manager has a bad year, if the thesis seems wrong and the numbers get ugly. If others are redeeming, it can feel like the responsible move.
Mark Spitznagel at Universa Investments spent years buying options that expired worthless whenever markets were calm. Some were skeptical through years one, two, and three. But for those who stayed, or added, got a 4,144% return in a single quarter when COVID hit in March 2020.
Spitznagel hadn’t predicted COVID, he had just spent years refusing to treat catastrophic risk as zero, absorbing the cost of being early, and waiting for the moment everyone else was unprepared for.
Gracias didn’t just invest in Musk once and wait. He kept going back through the Tesla near-bankruptcy, three SpaceX explosions and every moment the model said this thing is finished.
That’s the move many LPs cannot make. When a manager is down in year two or three, the instinct is to pull capital. The quarterly statement is ugly. The thesis looks broken. Everyone else is redeeming. So you think about redeeming.
The pattern is similar every time. The manager looks wrong. The bet looks stupid. Most people leave. A few stay, or add. That’s where the money is.
OG’s (Anunoby’s) tip-in. AG’s (Gracias’s) checks. Spitznagel’s years of patience.
None of them required predicting the future, but all of them required staying in the game when the model, the crowd, and the quarterly statement said get out.
A less than 1% doesn’t mean impossible, but it does mean almost nobody is prepared for it. That gap is where the money is.