Thiel Dumps Nvidia NVDA in Q3. Is This the AI Top?

Published on: Nov 17, 2025
Author: Maya Trent

Peter Thiel’s hedge fund sold its entire Nvidia stake in the third quarter, a $100 million exit revealed in a fresh 13F filing that landed after the bell. The move slices through a market still priced for an AI supercycle. Nvidia shares were up 1.8% to around 190 as of Monday afternoon, underscoring how crowded and resilient the trade remains even as one of Silicon Valley’s most famous investors quietly headed for the door.

Thiel exits Nvidia as stock stays bid

The filing shows Thiel Macro LLC offloaded roughly 538,000 Nvidia shares during Q3, closing a position that had ballooned in value as the chipmaker became the market’s purest AI hardware bet. He also cut Tesla by about 76% to 65,000 shares, while adding to Apple and Microsoft. The rotation is a clear signal: after one of the most spectacular momentum runs in tech history, at least one high-profile allocator is stepping back from the highest-beta AI names and moving into megacap balance sheets that can monetize AI without shoulder-to-shoulder cyclicality. The timing matters. Nvidia remains volatile but bid, still riding massive data center orders, while retail sentiment has rolled over. A popular sentiment gauge dropped to 41 out of 100 from 89 in late October, reflecting fragile risk appetite after months of parabolic gains. That disconnect — firm price, softening mood — is exactly when contrarian exits happen.

13F reality check and what it signals

A 13F is backward-looking. Positions are as of Sept. 30 and exclude shorts and many derivatives. Still, a full liquidation is unambiguous. Thiel has warned for months about AI exuberance, drawing lines to 1999 where narrative outran near-term cash flows. Today’s backdrop invites that lens: the S&P 500 trades near 23 times forward earnings while the FTSE sits around 14, a spread that has historically signaled US growth optimism bordering on complacency. Nvidia’s valuation remains a lightning rod. The bull case leans on supply scarcity, software lock-in and a multi-year capex wave from hyperscalers. The bear case focuses on concentration risk, margin mean reversion as competitors catch up, and the speed at which customers try to optimize AI spend. Thiel’s sell tells the market he sees better risk-adjusted returns elsewhere, or at minimum, the need to de-risk a crowded trade before the next volatility event.

From high beta to megacap ballast

The rest of the portfolio shuffle tracks that logic. Slashing Tesla while adding Apple and Microsoft points to a move up the AI stack toward diversified cash engines. Microsoft monetizes AI at the platform layer — cloud, enterprise software, copilots — with pricing power and recurring revenue. Apple, while later to visible AI features, offers defensiveness and a massive installed base that can be turned on with software updates. Tesla is exposed to auto cycle dynamics and a self-driving narrative that has been volatile in both timelines and regulatory friction. Nvidia sits in the capital equipment crosshairs, the chokepoint of AI compute where competition, export controls, and procurement digestion can swing quarterly numbers. Reallocating from NVDA and TSLA into MSFT and AAPL looks like classic late-cycle tech positioning: keep AI exposure, cut factor risk. It also reduces single-product dependency in a part of the market where expectations are unforgiving.

The new AI skepticism trade

Thiel’s exit arrives as AI euphoria collides with positioning fatigue. Retail participation has cooled, according to sentiment trackers, amid viral loss posts and a broader pullback in high-octane growth trades. That narrative shift matters because Nvidia’s rise has been a case study in positive feedback loops: blowout earnings, guidance upgrades, options activity, and persistent fear of missing out. When the marginal buyer hesitates, volatility replaces reflexive upside. The AI investment cycle is intact — hyperscalers are still writing huge checks — but investors are reassessing pacing and payback periods. The question is not whether AI is real, but whether 2026 earnings embedded in today’s prices can arrive on schedule. In that debate, an early and publicized exit by a marquee technology investor becomes a bookmark, if not a timing signal, for funds already trimming exposure.

Price action meets positioning risk

Despite the Thiel headline, NVDA ticked higher into the afternoon, a reminder that the marginal price setter is still institutional demand tied to data center capex and model training cycles. Traders are watching options skew, where dealers’ hedging flows can dampen or amplify swings. A single 13F does not break a thesis, but it can bend the narrative. If other whales are also rotating, the stock’s momentum could fray at the edges, especially into any hiccup in supply, export approvals, or order visibility. Nvidia’s dominance in AI accelerators remains intact, but the market has started to handicap the next phase: customer digestion, competitive responses from AMD and in-house silicon, and whether software and services can sustain margins as hardware mix evolves.

Reading the Tesla cut

Tesla’s 76% reduction is its own tell. The EV leader has rallied on autonomy hopes, energy storage, and margin recovery stories, but it remains sensitive to unit demand, pricing, and input costs. If Thiel’s book is a proxy for how growth-focused funds are recalibrating risk, Tesla may be caught between narratives while AI incumbents with stronger cash generation take flows. That rotation — away from aspirational AI exposure in autos toward incumbents monetizing AI now — lines up with the Apple and Microsoft adds. It also reflects a world where capital is getting more selective after a long period of multiple expansion doing the heavy lifting.

Valuation gravity is back

Macro also intrudes. With the S&P’s multiple stretched versus global peers, US tech must keep delivering outsized earnings to justify its premium. Any deceleration in AI spending, even temporary, could have an outsized impact on leaders like Nvidia, which are priced for near-perfect execution. That is the essence of bubble chatter: not that AI disappears, but that the curve of adoption and monetization wobbles relative to consensus. Thiel’s long-standing view that narratives can race ahead of cash flows resonates in this tape. Rotating into MSFT and AAPL keeps AI upside on the table while lowering the penalty if time-to-value slips.

What Nvidia bulls will argue

The counter to the Thiel read is straightforward. Nvidia’s order book remains deep, its software ecosystem widens the moat, and competitors face both technical and supply constraints. Even with sentiment swings, the company’s leverage to the buildout of AI infrastructure is unique. Bulls will say portfolio construction is not a referendum on fundamentals, and that profit-taking after a massive run is just discipline. Monday’s green print despite the filing supports that view, at least for now. The test comes if more filings confirm similar exits, or if upcoming customer commentary hints at a digestion phase that outlasts current expectations.

What to watch next for NVDA and AI stocks

Watch for corroborating moves in other 13Fs as managers show their Q3 cards, and for any Q4 updates that point to continued de-risking. Track hyperscaler capex commentary from Microsoft, Alphabet, Amazon, and Meta for clues on 2025 accelerator demand. Monitor retail flows and options positioning in NVDA, where swings in short-dated call activity have often preceded sharp moves. Keep an eye on export policy headlines and competitive roadmaps that could complicate supply and pricing. The Thiel sale does not end the AI trade. It marks a regime where selectivity, duration of cash flows, and balance-sheet strength reassert themselves. For Nvidia, that means the bar stays high, the scrutiny sharper, and every guide more consequential than the last.

AI Copper