By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel generally agrees that while AI-driven power demand is real, investing in VPU as an 'AI play' is misleading. The main risk lies in the regulatory and capital expenditure challenges faced by utilities, including potential compression of margins due to capex inflation and the risk of hyperscalers bypassing utilities. The key opportunity is the attractive income provided by VPU's 2.7% yield and low fees.
Risk: Hyperscalers bypassing utilities and shifting them to a secondary partner role, potentially compressing transmission margins and turning VPU into a standard, low-growth dividend play with zero AI upside.
Opportunity: VPU's 2.7% yield and low fees, which are attractive for income.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
The Vanguard Utilities ETF is a conservative play on the AI boom.
Led by AI demand, commercial electricity consumption is soaring.
This sector ETF is also appealing to dividend investors.
If a poll asked investors to rank the 11 S&P 500 sectors in order of excitement, there's an excellent chance that technology would take one of the top spots, while utilities would be toward the bottom, if not in the cellar.
That's an old-school way of looking at things. Thanks to the evolution of artificial intelligence (AI), utility stocks have more pizazz than market participants typically ascribe to the sector, providing an interesting refresh for exchange-traded funds (ETFs) such as the Vanguard Utilities ETF (NYSEMKT: VPU).
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
This $9 billion ETF should not be confused with a traditional tech fund or one dedicated to the broader AI investment thesis. Still, with this disruptive technology stoking exponential increases in power demand, the Vanguard fund is an interesting option for investors seeking some non-tech AI exposure.
As one of the largest ETFs in the category, this Vanguard fund allocates 62.4% of its portfolio to electric utility stocks, which is relevant because saying AI has power demands is an understatement. Data centers consume massive amounts of power.
Quantifying the power demands created by AI and cloud computing data centers isn't difficult. Last year, these facilities consumed a staggering 448 trillion watt-hours of electricity worldwide, exceeding the power consumption of all but 10 countries. That's a lot of power, and the portion consumed in the U.S. points to an opportunity with sector funds, including this Vanguard ETF.
A May report published by the Energy Information Administration (EIA) underscores the AI-related opportunity with this ETF. The EIA notes that 2026 power consumption will top last year's level, and that feat will be repeated next year. Of note to investors considering utilities exposure is that, in 2027, commercial power consumption will top residential consumption for the first time. Yes, that's an "AI thing."
Whether that makes this fund one of the best Vanguard ETFs remains to be seen. But with AI fostering new demand and with residential electricity prices unfortunately expected to rise this year and in 2027, this utilities ETF has the opportunity to deliver more growth than investors are expecting.
These days, there's plenty of AI enthusiasm to go around, but it can also create some confusion. Let's break down the AI hoopla as it relates to this Vanguard ETF. In culinary terms, the utilities fund is more of an AI nibble than a buffet.
At the end of the day, it's still a utilities ETF. That means its 2.7% dividend yield and its potential to rally if the Federal Reserve gets around to cutting interest rates are appealing traits for income investors looking for a sector ETF to hold for the long term.
Speaking of the long term, this ETF lives up to its Vanguard DNA, charging just 0.09% in management fees per year, or $9 on a $10,000 stake. That's well below the average expense ratio of 0.97% found on comparable funds.
Before you buy stock in Vanguard Utilities ETF, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Vanguard Utilities ETF wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $439,038! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,277,804!
Now, it’s worth noting Stock Advisor’s total average return is 942% — a market-crushing outperformance compared to 206% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of June 10, 2026. *
Todd Shriber has no position in any of the stocks mentioned. The Motley Fool has no position in any of the stocks mentioned. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"Utility stocks are constrained by regulatory rate-setting, meaning they capture volume growth far less efficiently than the tech companies driving the demand."
The article frames VPU as a 'conservative' play on AI, but this ignores the regulatory and capital expenditure (CapEx) reality. While AI-driven load growth is real, utility stocks are heavily rate-regulated. Unlike tech firms, utilities cannot simply raise prices to capture excess demand; they must navigate state public utility commissions (PUCs) to recoup the massive grid-hardening and generation costs required for data centers. Investors buying VPU are betting on a 'rate base' expansion, but if inflation or interest rates remain sticky, the cost of debt for these capital-intensive projects could erode dividend yields and compress P/E multiples. VPU is a play on infrastructure necessity, not AI margin expansion.
Utilities are effectively 'toll booths' for the AI revolution; if data centers must be built, the utilities are guaranteed a regulated return on every dollar of mandated infrastructure investment.
"VPU benefits from AI-driven power demand but captures minimal upside because utilities are regulated pass-throughs; the real margin expansion goes to unregulated power producers and renewable developers, not legacy utilities."
The article conflates two separate dynamics: structural AI-driven power demand (real) with utilities as an AI play (misleading). Yes, data centers will consume massive electricity, but utilities are regulated monopolies with capped returns—they don't capture AI upside, they pass costs through to ratepayers. VPU's 2.7% yield and 0.09% fee are attractive for income, but the 'AI pizazz' framing is marketing. The real risk: if commercial power demand spikes faster than grid capacity, utilities face capex inflation that regulators may not fully allow them to recover, compressing margins. The article also omits that private power producers and renewable developers—not legacy utilities—are capturing the highest-margin data center contracts.
If commercial demand truly exceeds residential by 2027 and utilities must invest heavily in grid infrastructure, regulated rate-of-return models could actually deliver steady 8-10% returns as capex gets embedded in rate bases—potentially outperforming a flat-yielding ETF in a rising-rate environment.
"VPU offers limited upside to AI-driven power demand because regulated returns and rate-case lags blunt earnings leverage."
The article correctly flags surging U.S. commercial power demand from data centers, with EIA data showing commercial use overtaking residential by 2027. Yet VPU's 62% electric-utility weighting exposes investors to rate-regulated entities whose allowed ROEs are capped by state commissions and whose capex recovery lags multi-year. Low 0.09% fees and 2.7% yield are attractive for income, but they do not offset the sector's historical beta to 10-year yields or the risk that hyperscalers bypass utilities via behind-the-meter generation. Forward multiples already price in modest growth; any delay in rate-case approvals would compress returns.
Regulators have repeatedly fast-tracked data-center interconnection tariffs and allowed above-average ROEs in high-growth states, potentially lifting realized earnings well above the article's implied baseline.
"AI-driven load growth could lift VPU, but the upside depends on durable commercial electricity demand and a supportive rate environment; without those, high-rate risk and regulatory headwinds could cap returns."
The article frames AI-driven power demand as a secular tailwind for utilities and suggests VPU as a low-cost, dividend-friendly way to play it. That thesis could hold if AI deployments translate into durable commercial electricity growth and data-center load expands meaningfully. However, risk lies in VPU’s concentration in regulated electric utilities, which makes earnings highly sensitive to interest rates and policy decisions; grid-capital investments can compress near-term margins; energy prices and regulatory shifts could cap upside; and real-world AI demand may be slower or more lop-sided than implied. The bullish case rests on sustained load growth and a favorable rate environment.
Against this view, AI-driven load growth may prove volatile or smaller than anticipated, as efficiency and shifting workloads dampen incremental demand; and if rates stay high, VPU’s income appeal and valuation upside could underwhelm compared with growth stocks.
"The rise of behind-the-meter, direct-to-generation power deals threatens the utility business model by stripping away the transmission and distribution margins that justify the sector's regulated rate-of-return."
Grok, you mention hyperscalers bypassing utilities, but you're missing the 'co-location' risk. Hyperscalers aren't just bypassing; they are increasingly demanding direct access to generation assets—like Amazon's Talen Energy deal. This shifts utilities from being the sole 'toll booth' to being a secondary partner. If utilities lose the ability to capture the transmission margin because data centers go 'behind-the-meter' to avoid grid fees, the VPU thesis collapses into a standard, low-growth dividend play with zero AI upside.
"Behind-the-meter bypass is a margin risk, not a thesis killer—unless adoption exceeds 25-30% of incremental data-center load by 2027."
Gemini's co-location risk is real, but the math doesn't kill VPU yet. Amazon-Talen is a *single* hyperscaler deal; most data centers still grid-connected. The transmission margin compression is a valid concern, but utilities retain rate-base recovery on *distribution* infrastructure—the last-mile hardening that serves all commercial load, not just hyperscalers. The real question: does behind-the-meter adoption reach 30%+ of new capacity by 2027? If it stays under 15%, utilities still capture meaningful upside. The article should've quantified this threshold.
"Partial behind-the-meter penetration can still strand approved distribution capex and compress realized ROEs via regulatory pushback."
Claude's threshold test on behind-the-meter share ignores the feedback loop: even 15% adoption lets hyperscalers cherry-pick high-margin sites, leaving utilities with lumpy, lower-quality load that PUCs are slower to approve into rate base. This raises the odds that approved capex gets stranded or ROEs get clawed back, turning the 'steady 8-10%' return into a multi-year lag rather than a reliable uplift.
"Co-location risk reduces upside but won't collapse VPU unless regulators deliberately starve capex returns; the outcome hinges on rate-case dynamics."
Gemini's co-location risk is real but not existential. Even with more direct generation, most data-center load remains grid-connected and utilities still recover capex via rate bases and tariffs. The real test is regulatory timing and ROE allowances, not a binary 'no AI upside' outcome. If 15–20% of new capacity goes behind-the-meter, VPU's upside compresses, but it's not annihilated unless regulators sharply curb capex returns.
The panel generally agrees that while AI-driven power demand is real, investing in VPU as an 'AI play' is misleading. The main risk lies in the regulatory and capital expenditure challenges faced by utilities, including potential compression of margins due to capex inflation and the risk of hyperscalers bypassing utilities. The key opportunity is the attractive income provided by VPU's 2.7% yield and low fees.
VPU's 2.7% yield and low fees, which are attractive for income.
Hyperscalers bypassing utilities and shifting them to a secondary partner role, potentially compressing transmission margins and turning VPU into a standard, low-growth dividend play with zero AI upside.