Billionaire Tech Mogul Peter Thiel Bought 8 New Stocks in Q2. They All Have a Common Theme.

Peter Thiel’s Thiel Macro restarted stock buying in Q2, investing more than $418M across eight new positions, all tied to power demand from the AI buildout. The largest stakes were Amazon (~$118M, 28% of the fund) and Vista Energy (~$76M, 18%), while regulated utilities American Electric Power (~$42.2M), DTE Energy (~$40.3M), FirstEnergy (~$40.0M), and CMS Energy (~$39.6M) were also included. The article highlights AEP’s $78B five-year capex plan (plus a potential additional $10B) and forecast load growth to 69GW by 2030 from 63GW, implying multi-year grid investment tailwinds, though risks include AI overbuild or demand slowdown.

Analysis

This reads less like a pure AI call and more like a validation of the “picks-and-shovels” trade where the bottleneck is grid capacity, not model demand. The first-order winners are the regulated utilities with visible load growth and rate-base expansion, but the second-order beneficiaries are transmission and gas-fired generation assets that can get built faster than new data-center load can be served. That makes the best near-term expression less about chasing AI software beta and more about owning names where incremental capex can be turned into approved earnings with limited commodity exposure.

The risk is that the market is already crowding into the same scarcity narrative while underestimating the financing drag. For AEP, CMS, DTE, and FE, the upside case depends on regulators accepting faster capital recovery and on long rates not rising enough to compress the equity multiple; if 10Y yields back up materially, the same capex that looks growthy can become a dilution story. For VST and VIST, the thesis is more convex but also more cyclical: if AI-driven power demand proves lumpy or hyperscaler contracts slow, merchant power pricing can retrace quickly.

Time horizon matters: over days, this is mostly a sentiment/positioning catalyst; over 1-3 months, watch for utility guidance revisions, interconnection backlog commentary, and data-center contract disclosures; over 6-18 months, the real test is whether load growth translates into allowed ROE uplift rather than just larger balance sheets. The contrarian view is that the market may be overpaying for visible capex while underpricing execution and regulatory latency. X-Energy is a long-dated call option, not a near-term earnings driver, so it should be treated as a watch item rather than a core trade until financing and commercialization milestones de-risk.

Key Decisions for Investors

• Long AEP / short XLU for 1-3 months: express the view that utility-specific AI load growth and capex visibility will outperform the broad utility basket; invalidate if AEP fails to maintain incremental load guidance or if 10Y yields spike >25 bps from current levels.
• Pair trade: long VST vs. short a rate-sensitive regulated utility basket (AEP/CMS/DTE/FE) into the next 1-2 earnings cycles; VST has more direct merchant pricing convexity if AI load tightens power markets, while the regulated names can be capped by slower rate recovery.
• Buy AMZN on pullbacks only, with a 3-6 month horizon: treat it as an AI infrastructure financing proxy, but size modestly because the stock is already a crowded AI beneficiary and the incremental edge is in data-center capex durability, not valuation re-rating.
• Avoid chasing XE as a core position; if used at all, treat it as a small call-option-sized sleeve with a 12-24 month horizon and only after evidence of financing/prototype milestones, since the commercialization path is too binary for a meaningful risk budget.
• Set an alert on AEP/CMS/DTE/FE next filings: if capex plans rise without corresponding load or allowed-return visibility, fade the move; if guided load and rate-base growth both step up, add on weakness as a structural 6-18 month rate-base compounding story.