Language | 
English
Spanish
Believe it or not, the best-performing S&P 500 sector over the past five years is not technology. It is energy. The sector, which bottomed at roughly 2% of the index in 2020 when oil briefly traded below zero and no investor wanted to touch it, has quietly delivered stronger returns than the technology giants that have dominated every headline since. In 2026, the Iran War has blown the gap wide open: Energy is up over 25% while Technology has managed less than 8%.[1]
The war has accelerated the oil and gas side of this trade. But the deeper, more durable story is in the power sector, where structural electricity demand is surging, driven by AI data centers building out at record pace, the electrification of transport, and industrialization across the developing world. The companies that generate, transmit, and distribute electricity are being rerated from defensive income plays into genuine growth stories.
The Demand Story
The International Energy Agency’s (IEA) latest Global Energy Review claims the “Age of Electricity” has arrived. Global electricity demand grew 3% in 2025, adding approximately 800 terawatt-hours, equivalent to the combined consumption of Germany and France. That growth rate is well above the 2.7% annual average of the prior decade, and follows a 4.3% surge in 2024 that marked the fastest expansion in years. Electricity grew 2.3 times faster than total energy demand, a structural divergence projected to widen further in 2026.
The drivers are structural rather than cyclical, and three stand out. The first is AI infrastructure. Data center electricity consumption grew 17% in 2025 and is projected to reach 1,000 terawatt-hours by 2030, equivalent to Japan’s entire national consumption; in the U.S., data centers accounted for half of all electricity demand growth. The second is transport electrification: electric vehicle (EV) sales surged 20% to over 20 million units globally, with one in four new cars sold electric and EVs capturing more than half of new car sales in China for the first time. The third is industrialization across Asia, the Middle East, and Africa, where manufacturing growth, urbanization, and rising temperatures are lifting power and cooling demand simultaneously.
The scale is reshaping corporate capital allocation. Spending by the five largest hyperscalers surpassed $440 billion in 2025 and is projected to exceed $600 billion in 2026, a 36% increase. To put that in perspective, hyperscaler capex now represents over 2% of U.S. GDP. A growing share is going to long-term power supply, and within that, nuclear is emerging as a clear strategic priority. Microsoft, Amazon, Google, Meta, and Oracle have signed a wave of 20-year nuclear power purchase agreements, and together have now contracted more than 13 gigawatts of capacity, with Meta’s January 2026 agreements alone accounting for 6.6 gigawatts, the largest private nuclear procurement on record. These are multi-decade commitments that reflect a recognition that AI infrastructure requires firm, 24/7, carbon-free baseload power at a scale few other sources can match.
Renewables: A Milestone Year
Renewables hit a genuine milestone in 2025: growth in renewables and nuclear generation exceeded total electricity supply growth for the first time, solar PV posted a record 600 terawatt-hour increase, and battery storage additions rose 40%. Yet coal still provided 34% of global generation, and when drought and poor wind constrained renewable output in Europe and South America, the shortfall was met by fossil fuels. The reality for the next decade is that power demand will require all-of-the-above: gas, renewables, nuclear, and storage expanding simultaneously rather than sequentially.
The Grid Bottleneck
The grid is emerging as the binding constraint on the AI buildout and arguably the most significant differentiator in the global AI arms race. AI-related data center electricity demand is expected to at least triple by 2030. At the chip level, Nvidia’s transition from H100 Hopper to GB200 Blackwell brought a roughly 300% increase in power draw, and the next generation, Rubin, adds a further 48%. Every leap in AI capability translates directly into a step-change in electricity requirements at the rack level.
The U.S. grid is ill-equipped to absorb this. Compute can only scale as fast as the grid and generation allow. Connection timelines now stretch to five years or more. Transformer prices have risen 60 to 80% since 2020, with lead times exceeding two and a half years and a projected 30% supply deficit. Prices for cables, conductors, and switchgear are all surging.
The political backlash is building in parallel. Northern Virginia, home to over 400 data center campuses, has halted new permits through 2028. At least 12 U.S. states have filed data center moratorium bills in 2026, dozens of municipalities have enacted local construction pauses, and a federal moratorium bill has been introduced in Congress. The industry’s demand for power is colliding with grid limitations, rising residential electricity costs, and community opposition.
The Iran war has compounded these pressures. The Strait of Hormuz closure has disrupted oil and gas flows, with the Gulf accounting for 20% of global natural gas supply and damage to Qatar’s Ras Laffan complex expected to take three to five years to repair. Rising energy costs are flowing through to industrial supply chains and consumer prices. But the supply shock is likely temporary while the demand growth is structural.
The Investment Opportunity
Grid and storage. The IEA estimates that annual grid investment needs to rise roughly 50% from today’s $400 billion by 2030, with over 2,500 gigawatts of projects already stalled in connection queues worldwide. Companies manufacturing grid components and deploying utility-scale storage sit at the intersection of every major energy trend.
Renewables and nuclear. 78 gigawatts of nuclear are under construction globally, the highest in 30 years, with Big Tech funding advanced projects. The nuclear value chain is entering a structural growth cycle that could last decades.
Utilities. The shift from defensive to growth is real but narrow, applying to operators with signed long-duration hyperscaler contracts rather than the broader sector.
Threats to Watch
The risks to this thesis are real and operate on different time horizons. Component inflation is the most immediate, compressing margins across grid equipment and renewables supply chains even as revenues grow. Grid connection delays and the moratorium wave are medium-term, capping how fast capacity can actually come online regardless of how much is contracted. The long-term risk is hyperscaler capex discipline. Many of the nuclear and data center projects underpinning current forecasts are at early development stages, and given their scale relative to existing infrastructure, the cancellation or delay of even a handful could materially reshape the demand curve.
Energy has moved from a cyclical trade into a structural theme. The Iran war and the grid bottleneck have accelerated the underlying trends, not created them, and the companies that benefit most are those positioned for a decade of execution rather than a quarter of headlines.
[1] Measured by the S&P Information Technology Sector GICS Level 1 Index and the S&P Energy Sector GICS Level 1 Index total return from 31 December 2025 to 24 April 2026.
The information contained in this article is for information purposes only, and represent the views of the author. It is not intended as specific investment or financial advice, or a recommendation or solicitation to buy or sell any security. Any investments or strategies listed in this article may not be suitable for all investors. Past performance is not indicative of future performance, and as with any investment, prices may fluctuate. It is recommended that advice is sought from a qualified investment professional prior to implementing any financial plan. LOM has made every effort to ensure that the contents herein have been compiled from sources believed reliable, however LOM does not warrant the accuracy, adequacy, timeliness, or completeness of this information expressly disclaims liability for errors or omissions in this information.
