Assessing the Utility AI Boom: A Growth Investor's Look at TAM, Scalability, and Valuation

The investment thesis for utilities is undergoing a fundamental transformation. It is no longer just about steady dividends and regulated returns. The sector is entering a multi-decade growth cycle, powered by a structural surge in electricity demand from artificial intelligence. This shift is quantifiable and massive.

The core driver is the explosive expansion of data centers. In 2023, they consumed about 4.4% of total U.S. electricity. By 2028, that share is projected to nearly triple, rising to an estimated 6.7% to 12%. This isn't a marginal increase; it's a demand acceleration that will require over 50 gigawatts of new power capacity. The Department of Energy's analysis shows data center electricity usage has tripled over the past decade and is expected to double or triple by 2028. This is the new baseline for utility growth.

This structural shift is already translating into exceptional financial performance. The sector's earnings trajectory has flipped from weakness to strength. After a 15% decline in earnings at the end of 2023, utilities delivered a sharp rebound in 2024. The momentum has continued into 2025, with the sector posting year-to-date gains of 20.25%. The most telling metric is earnings growth: in the third quarter of 2025, utilities reported year-over-year earnings growth of 23.1%, the third-highest among all S&P 500 sectors. This acceleration is directly tied to the capital deployment needed to meet AI-driven demand.

The bottom line is a re-rating of the entire sector. Utilities are being transformed from defensive yield plays into high-growth infrastructure plays. The projected demand surge, the massive capital investment cycle, and the resulting earnings acceleration point to a durable growth story. For investors, this is a secular shift where the total addressable market for power is expanding at a rate not seen in generations.

Business Model Scalability and Competitive Positioning

The utilities best positioned to capture AI-driven growth are those that can scale their infrastructure to meet massive, new power demands while locking in long-term, low-risk revenue. Three models stand out, each demonstrating a unique path to dominance.

NextEra Energy exemplifies a dual-model powerhouse. Its regulated utility, Florida Power & Light (FPL), provides a stable, cash-generating base. Simultaneously, its infrastructure arm, NextEra Resources, acts as a pure-play developer for the AI boom. This structure is now being validated by new contracts, with the company striking agreements with Google and Meta Platforms to develop energy infrastructure for data center campuses. This dual approach allows NextEra to leverage its regulated cash flow to fund growth projects while directly monetizing its development expertise. The model's scalability is clear: FPL's approved large load tariff and Florida's data center tax exemption create a favorable environment for rapid deployment, while the infrastructure arm can replicate deals nationwide. This setup provides a direct pipeline to the AI power demand surge, turning a sector tailwind into a multi-year growth story.

Dominion Energy's strategy is one of massive, state-backed capital commitment. The utility is investing $50 billion through 2029 to build the generation and transmission capacity needed for Virginia's data center market. The scale of this plan is its competitive moat. With 47.1 gigawatts of pending power requests in its service area, Dominion is positioning itself as the indispensable utility for a major tech hub. This isn't speculative; it's a regulated, long-term investment that converts future demand into guaranteed returns. The model's scalability comes from its ability to execute a multi-year capital plan while maintaining a strong dividend, allowing it to fund its own growth and capture a disproportionate share of the region's AI power needs.

Entergy's model is the most direct and lowest-risk. Instead of waiting for demand, it is funding its own power plants to serve a single, massive customer: Meta's $10 billion AI data center in Louisiana. The utility is constructing two new gas-fired plants to power the facility, with Meta funding the full cost of the required infrastructure. This creates a direct, off-balance-sheet revenue stream that is essentially guaranteed for the life of the agreement. The model's competitive advantage is its ability to de-risk growth by shifting capital costs to its customer while securing a long-term, high-margin revenue contract. It turns a volatile, speculative build-out into a predictable, cash-positive project, demonstrating a scalable path to earnings growth.

Together, these models show that the winning utilities are those that can either scale infrastructure to meet broad regional demand, act as a pure-play developer for tech giants, or directly fund projects for a single, anchor customer. Each approach leverages the utility's core strengths-regulation, capital, and grid access-to capture a disproportionate share of the AI power boom.

Financial Impact and Valuation: Growth vs. Price

The sector's explosive growth story is now reflected in its price. The S&P 500 Utilities Sector trades at a forward P/E of 22.26, a premium to its 5-year average of 20.13 and the broader market. This valuation embeds significant optimism, pricing in the structural demand shift from AI and electrification. The trade-off is clear: investors are paying up for growth, accepting a higher multiple for the potential of earnings acceleration.

That growth is real and accelerating. The sector reported 23.1% year-over-year earnings growth in Q3 2025, the third-highest among all S&P 500 sectors. Consensus expects continued expansion, with forward earnings growth estimated at 9.1% for 2026. This creates a dual-income stream: a dividend yield of 2.68% provides a defensive floor, while the growth trajectory offers the potential for capital appreciation. It's a rare blend in a sector historically seen as a pure yield play.

Yet the valuation premium makes execution critical. The key risk is whether the projected demand, particularly from data centers, materializes as forecast. Some analysts warn utility load forecasts may be overstated, with one report suggesting actual growth closer to 65 GW by 2030 versus utility projections of 90 GW. If the AI-driven power surge proves speculative or slower than expected, the high P/E multiple could face pressure as earnings growth fails to meet lofty expectations.

The bottom line is a valuation that demands flawless execution. The sector's premium multiple reflects a bet on a multi-decade growth cycle, but it leaves little room for error. Investors are trading a stable income stream for a high-growth opportunity, and the price they pay is a valuation that assumes the growth thesis will be validated.

Catalysts and Risks: What to Watch in 2026

The growth thesis for the energy and utilities sector in 2026 hinges on a few clear catalysts and risks. The primary driver is the sustained capital expenditure from hyperscalers, which is forcing utilities to build new power capacity at an unprecedented pace. This is already translating into concrete projects. For example, Dominion Energy has moved from exploring to executing a 300 MW Small Modular Reactor (SMR) project with Amazon, backed by a $500 million joint venture. Similarly, Entergy Louisiana is constructing two new power plants adding approximately 1,500 megawatts of capacity to support Meta's data center. These successful project milestones are critical validation that the sector can meet the AI-driven demand surge.

A secondary catalyst is a potential sector rotation into utilities as a 'quiet value play.' As some investors believe many of the best-performing stocks of 2025, notably artificial intelligence (AI) stocks, are simply overvalued, they may look for stocks trading below fair value. Utilities, with their average P/E ratio of around 18x, could benefit from this shift, especially if they are seen as providing essential, inflation-resistant cash flows.

The primary risks are regulatory and demand-related. First, there is the narrative of an 'AI bubble' if the monetization of data center power demand lags. This could trigger a reassessment of the sector's growth trajectory. Second, regulatory pushback on power costs is a tangible threat. As data centers drive up electricity prices, with residential retail electricity prices in September up 7.4%, politicians are making affordability a key issue. Virginia's governor-elect, for instance, has pledged to "make sure data centers don't drive up energy costs for everyone else in Virginia." This creates a political risk for utilities that may be forced to absorb or pass on higher costs.

Finally, the sector faces the risk of overbuilding. The scale of demand is immense, with data centers projected to consume anywhere from 6.7% to 12% of total U.S. electricity by 2028. If the pace of new power plant construction outstrips actual data center load growth, it could lead to stranded assets and margin pressure.

The key watchpoint is the 2026 earnings trajectory. The sector's growth rate is expected to moderate to ~9%. Investors must scrutinize whether this deceleration is a natural settling of expectations or a sign that the initial AI-driven demand surge is fading. The successful execution of large projects like Dominion's SMR and Entergy's plants will be the first tangible evidence of the sector's ability to sustain its growth story.