US power infrastructure is entering its largest capital investment cycle in generations. Two decades of flat grid spending, combined with surging AI data center demand and industrial reshoring, have created structural electricity scarcity across parts of the country. Eli Horton of TCW and Jim Worden of The Wealth Consulting Group examine what that shift means for advisors and their clients.
A capital cycle forms when prolonged underinvestment in physical assets creates scarcity, which drives pricing power and eventually a surge of new supply. Eli Horton, managing director and senior portfolio manager at TCW, describes the current grid situation as a classic capital cycle. US electricity output grew roughly five percent annually from 1950 to 2000, then went flat for two decades. That stagnation starved the grid of new capacity at exactly the moment AI data centers, reshoring factories, and broad electrification began drawing simultaneously on the same constrained infrastructure. The convergence of those forces, hitting a grid that received almost no incremental investment for 20 years, is what Horton regards as the defining feature of the current opportunity. "As an investor, we love the bottlenecks," Horton says. "That's where economic value accrues."
Two decades of flat demand left US grid infrastructure chronically underfunded, and the consequences are now measurable. Equipment lead times have stretched sharply: a high-voltage transformer ordered today may not arrive for three years. Utility contractors cannot find enough crews to erect power poles, and engineering and construction firms have themselves become labor bottlenecks. Horton notes that large technology companies are already paying extraordinary premiums to secure power supply in constrained regions, a sign that scarcity is real and not just forecast. He projects close to three percent annual electricity demand growth for the next several decades, enough to effectively double the grid from its current size. The federal Energy Information Administration's near-term forecast is closer to two percent annual growth, which still represents a meaningful acceleration from the flat years that preceded it.
PWRD is an actively managed ETF run by Horton that invests across the full electrical grid value chain, from generation and transmission through distribution, grid modernization, and the software systems that manage complex electrical networks. Unlike an index fund, PWRD does not mechanically track a benchmark. Instead, Horton's research process synthesizes what competitors, customers, suppliers, and management teams reveal about individual businesses, then acts when a stock appears mispriced against that mosaic of evidence. Core investment themes include electrification, energy security, AI infrastructure, and industrial reshoring. The fund is designed to capture companies that supply the physical build-out of the grid rather than the technology companies that consume its output. Investors should read the fund's prospectus at tcw.com before investing.
The 10 largest stocks in the S&P 500 now account for nearly 40 percent of the index, one of the highest concentration levels in decades, per data from S&P, FactSet, and TCW. Jim Worden, chief investment officer of The Wealth Consulting Group, points out that a capitalization-weighted index automatically allocates the largest share of capital to companies that have already performed best, which then pulls in more capital, making those winners larger still. He uses NVIDIA as a concrete example: the chipmaker was known mainly for gaming hardware before that same technology proved suited to large language models. The issue is not whether NVIDIA is a good business. The issue is sizing. "We're not going to love NVIDIA at 6 to 8 percent," Worden says. "We're going to love it at 2, 2.5 percent."
Horton argues the opportunity extends well past the largest technology companies. Rising power demand pulls through utilities, transmission networks, infrastructure providers, equipment manufacturers, and industrial companies that supply the grid's physical components. Worden calls this the second-order effect: as electricity consumption grows, the infrastructure required to produce and move it becomes increasingly valuable regardless of which AI model or application wins market share. Horton estimates that up to 12 percent of US electricity could be consumed by data centers by 2028, compared with about 4.4 percent in 2023. That fivefold shift in share, occurring across a grid that is simultaneously growing in absolute size, creates durable demand for grid hardware and construction capacity. Horton believes many portfolios hold far less exposure to this build-out than advisors realize, because broad benchmarks remain weighted toward asset-light winners from the low-rate era.
Worden uses the phrase to push back on headline-chasing in a market prone to speculative behavior. His test for any investment theme is simple: will the underlying force still matter in a few years, or is it the latest wave of enthusiasm? AI will generate its share of speculative meme-stock behavior, which is precisely why he keeps the focus on underlying fundamentals rather than narrative momentum. Horton reinforces that discipline by invoking investor Stan Druckenmiller, whose approach is to position for what comes next rather than what is already in the news. A headline, by definition, reflects the present. "When in doubt, zoom out," Worden says, reflecting a shared conviction that durable fundamentals outlast any story cycle. For both Worden and Horton, the electricity demand trend behind AI is the kind of multi-decade force that passes that test with room to spare.
Horton's core argument is that most client portfolios already carry heavy exposure to the last cycle's winners through broad-market benchmarks weighted toward mega-cap technology, while owning comparatively little of the companies likely to benefit from electrification, energy security, and the physical infrastructure build-out supporting a more digital economy. Worden agrees that position sizes in mega-cap names should reflect deliberate allocation decisions made by an advisor, not index drift made by passive vehicles. Advisors who have never actively chosen certain outsized weights may find that a capitalization-weighted index made that choice for them over several years of strong technology returns. Revisiting those exposures, and considering whether clients hold any dedicated allocation to grid infrastructure, energy transition, or industrial reshoring, is the starting point for a more intentional approach to the capital cycle ahead.
Eli Horton: managing director and senior portfolio manager, TCW; focuses on electrification, energy security, AI infrastructure, and industrial reshoring; portfolio manager of the TCW Transform Systems ETF (PWRD); TCW is a privately held global asset manager with more than 50 years of experience across fixed income, public equity, and alternative credit.
Jim Worden: chief investment officer, The Wealth Consulting Group.