🔔 Subscribe to The Trend Catcher / The Value Anchor — unlock the full historical archive and never miss a weekly pick again.

After a six-month hiatus from public equities, billionaire investor Peter Thiel is back in the market. However, rather than chasing high-flying semiconductor stocks like Nvidia or buying back familiar mega-caps, Thiel is betting heavily on the next critical frontier of the artificial intelligence boom: electricity.

According to the latest 13F regulatory filings, Thiel Macro has ended a two-quarter streak of holding zero public market assets. The fund recently deployed approximately $419 million across just eight stocks. Notably absent from the portfolio are former staples such as Apple Inc.(AAPL.US), Microsoft Corporation(MSFT.US), and Tesla Motors, Inc.(TSLA.US).

Instead, the filing reveals a highly concentrated, thematic portfolio designed to capitalize on the massive energy demands of AI data centers, signaling a strategic pivot from "computing power" to "power generation."

Amazon: The Strategic Anchor

The fund’s largest new position is Amazon.com, Inc.(AMZN.US), with 495,000 shares valued at roughly $118 million at the end of the quarter, representing 28.18% of the portfolio.

In Thiel’s strategy, Amazon serves a dual purpose. As the parent of AWS, it is a primary beneficiary of surging AI compute demand. Simultaneously, Amazon has emerged as one of the world's largest institutional buyers of data center power and nuclear energy. The tech giant recently committed $500 million to advanced nuclear reactor developer X-Energy—a company that also appears in Thiel Macro's new portfolio.

By anchoring with Amazon, Thiel is essentially buying the demand side of the AI equation while using the rest of his capital to acquire its upstream energy suppliers.

A 71% Allocation to the Power Supply Chain

The remaining seven stocks in the portfolio make up nearly 72% of Thiel Macro's total holdings. This capital is exclusively allocated to companies spanning the entire electricity and energy value chain:

• Power Generation & Nuclear: Vistra Corp.(VST.US) and X-Energy, Inc. Class A(XE.US).
• Grid & Utilities: American Electric Power Company, Inc.(AEP.US), DTE Energy Company(DTE.US), FirstEnergy Corp.(FE.US), and CMS Energy Corporation(CMS.US).
• Traditional Oil & Gas: Vista Oil & Gas(VIST.US), an Argentina-focused shale operator, which serves as the portfolio’s traditional baseload energy hedge.

Thiel’s stock selection closely mirrors the recent infrastructure moves of Big Tech. For instance, Vistra Energy recently signed a 20-year power purchase agreement to supply AWS with up to 1,200 megawatts of nuclear power. Meanwhile, DTE Energy is slated to power a planned 1-gigawatt data center for Alphabet Inc. Class A(GOOGL.US).

Trading the AI Bottleneck

Industry analysts note that this aggressive reallocation does not indicate Thiel is turning bearish on artificial intelligence. Rather, it highlights a sophisticated shift in his investment thesis: moving away from the crowded trade of "who can build the most GPUs" toward "what will restrict AI's expansion."

While silicon manufacturing can be scaled up relatively quickly, the infrastructure required to power it—power plants, nuclear reactors, transmission lines, and grid connections—takes years, if not decades, to build. As tech behemoths race to construct mega-data centers, the limiting factor for AI growth is rapidly shifting from chip availability to grid capacity.

While a 13F filing only provides a quarterly snapshot of U.S. equity holdings and doesn't reflect Thiel’s broader private investments, the directional signal is undeniable. Rather than abandoning the AI narrative, Thiel is evolving it—stepping away from AI "valuation trades" and positioning himself firmly in "bottleneck trades."

Ultimately, regardless of which tech giant wins the foundational model war or which chipmaker dominates the hardware market, they all share one non-negotiable requirement: massive, uninterrupted electricity. And Peter Thiel has positioned his portfolio to supply exactly that.