Peter Thiel’s Nvidia Exit: Signal of an AI Bubble at the Tipping Point?

When FinTelegram’s recent investor briefing outlined a tail-risk “AI bubble crash” scenario, it was a macro thought experiment. Peter Thiel’s Thiel Macro LLC now adds a very real data point: according to SEC filings, the fund sold its entire Nvidia stake – around 537,742 shares worth roughly $100 million – during Q3 2025 (Source: Reuters).

The move comes just as central banks, from the ECB to the Bank of England, and even tech leaders like Google’s Sundar Pichai, warn of bubble-like dynamics and “irrational” AI spending (NYPost.com) At the same time, SoftBank and other large investors have also exited or reduced Nvidia positions, intensifying the debate over whether AI valuations are peaking (Source bitget.com).

Read more about the AI Bubble fears here.

How Much Does Thiel’s Move Matter?

Nvidia is the key equity barometer for the AI trade. Its chips power data centres, model training and much of the AI arms race. A complete exit by an insider-level tech investor like Thiel inevitably looks like a vote of no confidence in current pricing, even if his fund has not publicly framed it that way (Source: finextra.com).

However, portfolio mechanics matter. Thiel Macro had a highly concentrated Nvidia position; taking profits and reducing single-name risk near all-time highs is rational risk management. Read purely as a timing call, the sale is a warning signal, not definitive proof that the AI cycle is over.

Revisiting the Doomsday Scenario

In our FinTelegram briefing, we described a doomsday variant where:

• AI megacaps fall 60–75%,
• broad equity indices drop 30–40%, and
• crypto suffers a 60–80% drawdown in BTC and ETH, with many altcoins effectively wiped out.

Thiel’s exit and other large divestments slightly increase our estimated probability that such an extreme unwind could occur over the next 2–3 years – from roughly 10–15% to the upper end of that range. They indicate that sophisticated capital is starting to de-risk at the margin.

Our view: this is not a call to panic, but a clear reminder to investors that AI and crypto exposures are tightly linked high-beta trades. Concentration, leverage and liquidity risk should now be treated as primary portfolio variables, not footnotes.