U.S. Power Demand Could Set a Record in 2026. 1 Vanguard ETF to Buy to Cash in on the AI Power Surge

Vanguard Utilities ETF (VPU) is presented as an indirect AI beneficiary, with data-center-driven electricity demand supporting utilities exposure. The fund has a 2.7% dividend yield and a very low 0.09% expense ratio, making it attractive for income-focused investors, especially if interest rates decline. The piece is mostly commentary rather than new market-moving information, so the likely price impact is limited.

Analysis

The market is likely underestimating how “utilities” become a capacity-constrained infrastructure trade rather than a pure bond-proxy. The first-order beneficiary is regulated electric utilities with near-term load growth and exposed rate-base expansion; the second-order winners are grid equipment, transformers, switchgear, cooling, and gas-fired peaker capacity because AI load is lumpy, regional, and power-quality sensitive. That favors companies with balance-sheet access and interconnection queues already in hand, while weaker municipal-heavy or leveraged utilities can actually get squeezed if capex inflation outruns allowed returns.

The key timing nuance is that the equity story lags the demand story by quarters to years. Data-center load growth can be announced now, but revenue uplift usually arrives only after rate cases, transmission approvals, and multi-year capex cycles, so the immediate catalyst is not usage itself but forward guidance on capex and regulator acceptance. The bigger risk to the bullish thesis is not demand falling, but policy friction: if residential bills keep rising, regulators may cap pass-throughs, compress allowed ROEs, or force delayed project recovery, which would punish utilities even as volumes rise.

From a factor perspective, lower rates help utilities twice: they reduce discount rates and improve financing economics for capex-heavy balance sheets. But if rates stay elevated, the sector can still work selectively as a long-duration earnings grower rather than a yield substitute. The more interesting contrarian view is that the ETF may be the wrong instrument for the AI trade because it dilutes the true beneficiaries with slow-growth incumbents; the better expression is a barbell of grid/infrastructure winners plus the AI power chain, not broad utilities beta.

Net: this is a modestly bullish, not euphoric, setup. The trade is in the second derivative of AI electricity demand, where the market often misprices how much of the upside accrues to rate-base growth versus how much gets absorbed by higher capex and regulatory lag.

Key Decisions for Investors

• Long VPU or XLU on a 3-6 month horizon as a lower-volatility AI power exposure, but size modestly; upside is capped by ETF dilution, while drawdown protection comes from dividend support and rate-cut optionality.
• Prefer a barbell: long utility infrastructure enablers such as ETN/PH/ABB-style power equipment exposure versus long-only broad utilities, because the equipment names monetize the capex cycle immediately rather than waiting for rate cases.
• Pair trade: long regulated utilities with visible load growth, short high-duration bond proxies or levered utilities with weak rate-base visibility; this captures the spread between real AI-driven capex beneficiaries and crowded income substitutes.
• Consider buying call spreads on selected utilities names with heavy data-center footprints into the next earnings season; the best upside comes from guidance revisions, not from the current headline narrative.
• Avoid chasing after a rate-cut rally; if 10-year yields back up or regulators push back on rate increases, utilities can underperform despite stronger electricity demand.