“There are three ways a smart man can go broke: liquor, ladies, and leverage.” — Charlie Munger
The first two are slow. Leverage is fast.
Was reading about the AI trade this week and came across a story that’s basically a live case study in why leverage destroys smart people faster than dumb ones.
Leopold Aschenbrenner. 25 years old. Columbia valedictorian at 19. Ex OpenAI researcher. Wrote a 165 page essay on AI that made him a Silicon Valley celebrity. Started a hedge fund called Situational Awareness in 2024 betting on the AI infrastructure trade, chips, data centers, power. By mid this year the fund had returned over 1,500% since inception and was up 439% in the first six months of 2026 alone. Assets peaked around 45 billion dollars.
Then July happened.
He was running roughly 4x leverage on his positions. When AI stocks were flying, that leverage was the whole story, it turned good calls into extraordinary returns. But leverage doesn’t care which direction you’re right or wrong in, it just multiplies whatever happens. When his top holdings started falling 30 to 50%, the same leverage that built the fund started tearing it apart. Margin calls came in. He had to sell to raise cash, not because his thesis was wrong, but because the lenders don’t wait for you to be proven right.
In one month, the fund lost 67% of its value. He was forced to offload almost his entire public portfolio in a fire sale to Citadel, at a discount, just to survive. A 45 billion dollar fund got cut down to around 10 billion in weeks.
And here’s the twist that makes this whole story sting even more.
The moment the Citadel deal closed, the exact same stocks he was forced to dump started ripping higher. Nebius rallied over 30% off its low. Bloom Energy jumped as much as 40%. Chip heavy indices in South Korea surged as much as 18% the very next session. US tech stocks logged their best single day in months, simply because the market’s biggest forced seller was finally out of the way.
Read that again. He wasn’t wrong about the direction. He was forced to sell at the bottom because of leverage, and the stocks recovered almost immediately after he was gone. The thesis survived. The leverage didn’t let him survive long enough to be right.
Here’s the part that should sit with every trader reading this. Aschenbrenner wasn’t wrong about AI. His long term thesis might still play out exactly as he predicted. What killed him wasn’t the idea, it was the structure. Leverage removed his ability to be patient. A position you believe in can survive being wrong for a while. A leveraged position can’t, because the lender decides the timeline, not you.
This is exactly why I keep saying leverage and borrowed money in trading is not a strategy, it’s a countdown timer. Taking a loan to trade, running high leverage on options, doubling position size because you’re “sure” this time, it all works fine until the one week it doesn’t, and that one week erases years of gains.
The market doesn’t punish being wrong. It punishes being wrong with no time to be right.