Notes of Paul Tudor Jones (PTJ) on AI bubbles

Paul Tudor Jones on the AI Bubble Debate by Bloomberg

The only way to reduce debt to GDP is to have obviously nominal growth exceed your interest rate.

– Paul Tudor Jones

Here are notes for Paul’s interview and my opinions

  • Today feels like Oct 1999, but if this is a bubble, it’s a small one. Past bubbles ran 400–600%. Nasdaq is “only” ~200% off the bottom. Blow-off possible, not inevitable. [I agree; see my previous post Is it like internet bubble? in October]
  • Key bull case: rates. If Fed funds fall toward ~2.25–2.75%, that’s powerful fuel for equities. Markets look 6–9 months ahead, not at today’s data. [Sure]
  • Difference vs 1999: companies are profitable. [I don’t agree; I believe AI model companies like OpenAI etc. are losing a lot of money; let’s see when they publish numbers for IPOs]
  • Risk isn’t traditional leverage like in margin accounts — it’s derivative leverage: options, leveraged ETFs (up 250% from 2022 bottom), and trader-driven equity flows. [Very real]
  • Jones stays a trend follower. Recently, gold & silver > Bitcoin despite massive crypto inflows. He now expects precious metals to outperform crypto into year-end. [I wouldn’t agree back then; but I would be very wrong, so far]
  • Bond vigilantes were held in back; money debasement happened in gold and bitcoin instead. [True]
  • Biggest risk: concentration everywhere — stocks, investors, and policy power. [Agree]
  • Bottom line: short-term cautious, but Paul believes markets can be substantially higher by year-end. Likely long: Nasdaq. Short: Bonds.