Back

Everyone’s focusing on Leopold Aschenbrenner’s latest 13F.

Massive hedges on SMH, NVDA, AMD, AVGO, ORCL.

This is a relative trade:
- Long second-order AI winners (power, data centers, storage, re-purposed miners)
- Short / hedge the crowded, fully-priced mega-cap names
Translation:He’s not betting against AI. He’s betting the obvious winners are overpriced.

If AI capex keeps exploding:
Power demand spikes --> BE wins
Storage demand rises --> SNDK wins
Compute shifts to infra layers --> IREN / CORZ / CRWV win

But if expectations on hyperscalers / chip leaders compress even slightly:
- NVDA / SMH downside funds the trade

Do you think NVDA, ORCL, AVGO are becoming overpriced?

In my opinion, I’m not relying on his track record, it’s too short.

But look at the setup:
NVDA trading 26.5x forward earnings, AVGO 38x+, ORCL pushing multi-year high multiples while AI capex expectations are already priced for years of perfect execution.

You’re effectively paying today for growth that has to materialise flawlessly.

It doesn’t take a collapse,  even a 5–10% multiple compression wipes out a big chunk of expected returns.

If he’s hedging here, it suggests the market is pricing close to a best-case scenario with very little margin for error.

**The Image doesn't show the full portfolio**