No longer optimistic. The Iran war is a distraction from the real story: AI's impact on the job market in the next several years.

AI will change everything. If the war ends, markets will rip, but that's not predictable. His focus is the long-term health of the US job market.

The software sell-off was a wake-up call. At leading tech companies, best AI users have replaced 7 colleagues, going from 8-person teams to 1 person with AI.

AI models are improving exponentially. In three years, with exponential improvement, the broader knowledge economy could see massive displacement (e.g., from 7 to 6 workers). He uses Claude for legal documents he used to pay for.

Points to Mike Green's theory on passive investing. Passive buying removes active sellers and squeezes floats, creating huge multipliers (e.g., $1 into NVDA adds $100 market cap).

If layoffs cause 401(k) outflows, those flows reverse and the multipliers still exist, leading to 'falling knives' with not enough active managers to catch them. AI will cause the cataclysm that intersects with this passive market structure risk.

Almost everything has room to the downside. He's not shorting software. The best way to play it is in credit, expecting spreads to widen.

If the index crashes due to outflows, Mag 7 will do poorly. He recommends convex trades like long-dated OTM put spreads on ETFs like HYG and LQD to cap downside while playing the thesis.

Markets will price this in before job displacements occur. This will be a GFC-type fallout, but on a faster fuse.

Credit ETFs (HYG, LQD) face a liquidity mismatch risk: the underlying can become illiquid in a displacement scenario, APs step back, people redeem ETFs but can't get liquidity. This gives extra downside juice if short.

Timing is hard. Volatility increased due to Iran war; once past that, vol will die down. Long puts on credit ETFs are cheap.

In February, smart money conversations were picking up on this theme, but the Iran war distracted people. Spreads have already widened a bit. Credit has been too cheap.

Numb to the big numbers. In the aftermath of a massive correction, some of these would be names you'd want to buy.

Would not be long or short pre-IPO. His firm uses a momentum strategy on S&P 500 constituents and would buy if they score, agnostic to valuation.