Better Energy Play: NextEra vs. Constellation Energy

NextEra is pursuing a $94B capital program through 2030 and an all-stock acquisition of Dominion Energy, though the deal faces intense regulatory scrutiny. Constellation shares are down over 20% in 2026 amid softer profit guidance and concerns about Calpine-related costs and added debt, despite higher AI-driven power demand and generally stronger earnings projections. Overall, the article frames both utilities as benefiting from AI data-center electricity growth, but with materially different risk profiles—regulated stability at NextEra versus higher-volatility upside at Constellation.

Analysis

This is less an AI demand trade than a pricing-power and capital-allocation trade. The market is likely over-indexing on the secular growth story and underweighting the fact that utilities re-rate on regulatory certainty, financing costs, and balance-sheet flexibility as much as on load growth. NEE’s biggest near-term upside is not the acquisition itself but the removal of uncertainty; if the deal is delayed or forced into concessions, the stock can lag even if the long-run earnings base improves.

CEG is the more levered expression of the power scarcity theme, but that cuts both ways. The stock’s drawdown likely reflects investors re-pricing the fact that merchant power exposure plus acquisition debt turns good demand into a higher-beta equity rather than a stable compounder. If hyperscaler demand pauses, power prices soften, or integration costs force another guide-down, CEG can de-rate quickly; if demand stays tight, the upside is real but probably arrives through earnings revisions, not multiple expansion alone.

The cleaner second-order winners are not the obvious utility names but the grid and balancing-capacity ecosystem: transmission, interconnection, peakers, and gas-linked reliability assets. The consensus may be missing that AI power demand initially helps the owners of constrained capacity, but over 6-18 months it also invites new supply, which caps the durability of scarcity rents. In other words, the trade is about whether this is a 12-month shortage or a multi-year structural shortage; that distinction determines whether CEG is cheap or just volatile.

Key Decisions for Investors

• NEE: buy only on regulatory-driven weakness over the next 1-3 months; treat it as a lower-risk utility compounder if the Dominion process clears, but cut if approval risk expands or financing costs move higher.
• CEG: avoid chasing into strength until the next earnings print confirms no further guide-down and no balance-sheet surprise from Calpine; the stock is a higher-beta call on sustained AI load growth, not a defensive utility.
• Relative-value idea: long NEE / short CEG for 1-3 months if you want to express 'certainty over leverage' while the market sorts out regulation versus balance-sheet risk; thesis breaks if CEG posts upward revisions or NEE gets a clean approval path.
• If already long CEG, consider financing via call overwrites into strength rather than adding outright exposure; this monetizes volatility while waiting for a clearer earnings inflection.
• Watch the next 1-2 quarterly updates from META/MSFT hyperscaler capex and power contracting: if data-center buildouts slow, the whole AI-power complex should be de-rated, especially CEG.