Lazy Trader AIIn early May 2026, Michael Burry told his 200,000-plus Substack subscribers that “the market has jumped the shark” and that “the end of… this… is nigh.” The backdrop: a significant leveraged short expressed through January 2027 put options on SOXX, the semiconductor ETF.
Within hours, the takes were everywhere. “Burry is shorting the AI trade.” “The Big Short guy is back.” Some of that is fair. Most of it skips the harder, more useful question: what does a disclosed put position actually tell you, and what should you do with it?
A big disclosed put is not automatically a directional bet on a crash. It can be conviction, a hedge against a long book, or cheap tail insurance. The disclosure rarely tells you which — and the difference is everything.
He's not alone — and that matters
Burry is the loudest, but he's not the only one expressing caution. Stanley Druckenmiller and David Einhorn have voiced similar concerns about AI-era valuations. And a paper from Man Group — the world's largest publicly traded hedge fund — argued that the AI boom is real, but the financial structure built around it appears to be expanding faster than any credible adoption curve can justify.
At the same time, plenty of Wall Street isn't calling a bubble at all — they're just adding portfolio protection in case. Bank of America's read on past bubbles suggested tech wasn't yet in bubble territory. So the “smart money” is genuinely split, which is itself the most honest signal available: nobody actually knows, and the sophisticated response is to hedge, not to bet the farm either way.
What a disclosed put actually is
Two technical points that headlines almost always butcher:
- 13F put positions are reported by notional, not by risk. A filing shows the value of the underlying the puts control — not how much capital is actually at risk, which is just the premium. A “$1 billion put position” might represent a fraction of that in actual dollars committed.
- You can't see the rest of the book. A 13F shows long equity and listed options. It does not show short stock, swaps, or the broader portfolio those puts might be hedging. The puts could be offsetting a much larger long position you can't see.
A put on a 13F is a silhouette. You can see the shape of the position, not the body it's attached to.
Conviction, hedge, or insurance?
Every large put position is one of three things, and they demand completely different responses:
- Conviction short. The investor genuinely expects the underlying to fall and is positioned to profit. This is what the headlines assume.
- Hedge. The investor is long the theme (or the market) and bought puts to cap downside. Net exposure may still be bullish. Copying just the puts would put you in a position the original never actually held.
- Tail insurance. Cheap, far-out-of-the-money puts bought as a low-cost hedge against a low-probability crash. The investor fully expects to lose the premium most years. Following this as a “bet” is a guaranteed slow bleed.
The leverage and the long-dated January 2027 expiry on Burry's SOXX puts suggest something more committed than passive tail insurance — but without the rest of the book, you can't be certain it isn't hedging a long AI position elsewhere.
Burry's track record with shorts
This matters for calibration. Burry's famous 2008 housing short was career-defining and correct. But his subsequent high-profile bearish positions have a more mixed history — disclosed shorts against Tesla and ARKK in the early 2020s, several of which were covered relatively quickly or were partly hedges rather than naked directional bets.
The lesson isn't “ignore Burry.” It's that being early is indistinguishable from being wrong until it isn't — and a disclosed snapshot doesn't tell you whether the position is still on a week later. He has covered fast before.
How to read it without copying it
- Treat it as a thesis to investigate, not a trade to mirror. The valuable part is “a serious investor thinks the AI-infra leg is fragile” — go pressure-test that yourself.
- Separate the notional from the risk. Don't read a headline notional as the amount someone is “betting.”
- Look for confirmation in faster data. Is unusual put flow building in SOXX, NVDA, or the neo-clouds? Are dark pool prints leaning sell-side? Disclosed positioning plus live flow is a real stack; disclosed positioning alone is a silhouette.
- Respect the disagreement. The fact that equally sophisticated investors are on both sides is the signal. The honest position is “hedge your exposure,” not “pick a side and max out.”
The discipline
The reflex to copy a famous investor's disclosed position is the same reflex that makes people copy a viral options-flow screenshot. In both cases you're acting on a partial picture, after a delay, without the context that produced the position. Sometimes it works. As a process, it's a way to lose money with conviction.
Read the position. Investigate the thesis. Confirm with live data. Size for the real possibility you're wrong. That's the difference between using smart-money signals and cosplaying them.
A disclosed put is conviction, hedge, or insurance — you usually can't tell which · 13F puts are notional, not risk · You can't see the rest of the book · Burry has covered fast before · Use the disclosure as a thesis to test, confirm with live flow, and size for being wrong.
This article is for educational purposes and is not financial advice. Nothing here is a recommendation to buy or sell any security or option. Options carry substantial risk of loss. Lazy Trader AI is a market-monitoring tool, not a brokerage.