← All sectors / The AI transformation
002 · Energy & utilities
Powering the buildout
Curve position
Binding constraint
Interconnection queues and turbine delivery slots.
Why this sector sits at Orbit
Structural pressures have moved AI from utility innovation labs into core operations, and most US investor-owned utilities now run production machine learning for demand forecasting and asset health. Adoption is established and broadening at an ordinary pace; the sector AI AI adoption rate remains below the cross-industry average at only 33%, signaling established but incomplete penetration — Growth, not Takeoff.
Reviewed on a two month cycle. The position moves only when a dated, verifiable change in the binding constraint justifies it.
Heatmap / How this sector's stocks are moving
| Stock | Price | 1 Day | 1 Week | 1 Month | Year to date |
|---|---|---|---|---|---|
| $VST | $120.29 | −13.83% | −15.72% | −25.05% | −25.44% |
| $CEG | $245.72 | −6.71% | −4.79% | −18.20% | −30.44% |
| $NRG | $83.25 | −13.52% | −14.15% | −19.64% | −47.72% |
| $TLN | $326.78 | −8.54% | −6.03% | −12.42% | −12.82% |
| $NEE | $88.45 | +4.89% | +9.17% | +13.57% | +10.18% |
| Sector average | −7.54% | −6.30% | −12.35% | −21.25% |
Prices as of November 5, 2026. The sector's move is the equal-weighted average of these listed companies, a market proxy for context, not recommendations. See it on the full heatmap →
Sentiment / The mood in this sector's stocks
Since 2018, Energy & utilities has been in Capitulation and falling 21 separate times. After those, its stocks beat their usual three months 12 times (57%), by a median of +0.5 points.
Confidence: low, a split like this is common by chance, so no edge is called.
AI's appetite for electricity is colliding with a grid built for a different century. Data centers are on track to consume a rapidly growing share of total power demand, and interconnection queues in key markets run years long. For the first time in decades, load growth, not efficiency, is the story in the utility sector, and everything in the energy complex is being repriced around it.
Context matters: u.S. Electricity demand was essentially flat for two decades: utilities planned, staffed, and invested for zero growth. The AI load shock hit an industry with no spare organizational muscle for expansion, which is why responses lag and why the scramble is so investable.
The structural driver is simple arithmetic: a large AI campus can draw as much power as a mid sized city, developers are proposing them by the dozen, and the grid's spare capacity was already committed to electrification of transport and industry. Something has to give, and what gives is price, speed, and the willingness to build.
Every generation source is back on the table. Gas turbine order books have exploded, with delivery slots sold out years ahead. Nuclear is being restarted, uprated, and re licensed while hyperscalers sign long term power purchase agreements and back small modular reactor developers directly. Utility scale solar plus storage remains the fastest new capacity to energize.
Behind the meter generation is the wildcard: data centers co locating at power plants or building dedicated generation to bypass the interconnection queue entirely. Regulators are still deciding who pays for grid upgrades and how co location is treated, decisions that will shift billions between utilities: generators, and tech companies.
Transmission is the quiet bottleneck beneath it all. High voltage transformers, switchgear, cable, and the engineering firms that build lines are capacity constrained worldwide, and every regulated utility's capital plan now leans harder into grid modernization, which, under rate base economics, converts directly into earnings growth.
The value chain runs from fuel (gas, uranium, sunlight) through generation equipment (turbines, reactors, panels, batteries) to delivery (transformers, cable, substations) and finally to markets and retail. Bottlenecks migrate along this chain, and pricing power sits wherever the queue is longest at the moment.
The overlooked layer includes independent power producers with uncontracted capacity in the right markets, grid equipment suppliers several tiers down the chain, fuel and uranium suppliers, and the specialty contractors doing the physical work. Many trade at valuations set before load growth returned.
Competitive dynamics vary by regulatory model: regulated utilities earn on approved capital, so growth is a filing away; merchant generators live on power prices and capacity auctions; equipment makers ride global demand. Each model rewards a different thesis, and mixing them up is how energy investors get hurt.
Risks: regulatory backlash if residential ratepayers bear data center costs, permitting and supply chain delays, commodity price swings, and the possibility that efficiency gains or demand response blunt peak growth. Energy is cyclical and political; position sizing should respect both.
What to watch: interconnection queue reforms, utility capex revisions, turbine and transformer lead times, hyperscaler power purchase announcements, and capacity market auction prices. The research treats watts as the scarcest input in the AI stack and follows the money spent to create them.