The Only Winning Move

The AI boom might run another year or two. Let's say 10% in 2026, another 10% in 2027, then a 30% crash in 2028. Do the math: you end up roughly where you started — minus three years of your life you can't get back. A 30% drop requires a 43% gain just to break even. Ask anyone who bought the Nasdaq in 2000 how long that took. Thirteen years.

But let's steelman the bulls. Suppose we one-shot this revolution without pain — no correction, no shakeout, straight to the promised land. Even then, only a handful of winners survive. Look around: every major player is spending as if they've already secured a seat at the final table. That can't be true. Mathematically, financially, physically — the chairs are limited and the music will stop.

The B2B loop tells the real story. Microsoft pays OpenAI. OpenAI pays Nvidia. Nvidia pays TSMC. Everyone's "revenue" is everyone else's capex. Net it out and ask: where's the external cash entering this system? Consumer subscriptions at $20 a month? The denominator for all this spending doesn't exist yet.

Then there's physics. Power grids rejecting data centers. Arizona fighting over water for fabs. This isn't a software scaling problem anymore. It's atoms now, and atoms don't 10x on a product cycle.

Every valuation metric — Shiller CAPE, Buffett Indicator, margin debt, equity as a share of household wealth — screams the same warning. The defense is always identical: "But this time the growth justifies it." It never does. The growth gets arbitraged away or the multiple compresses. Usually both.

I don't short. Shorting means paying for your conviction — theta decay, squeeze risk, bleeding every day you're early. Being early isn't wrong, but paying to be early is expensive.

I just hold cash at the risk-free rate and wait. Same bet, zero bleed. If I'm wrong, I miss gains but lose nothing. If I'm right, I get to buy when others have to sell. (Plus 10% long — a hedge against my own certainty.)

Sometimes the only winning move is not to play.

PS. Stay 100% cash long enough and you need the world to get worse to justify your stance. That's how rational caution curdles into doomerism. The narrative starts owning you instead of the other way around.

So the 10% long isn't just "I might be wrong about the market." It's "I might be wrong about myself." It keeps you tethered — forces you to stay in the game enough that you can't drift into apocalypse-brain.

That's not just humility. That's structural sanity.