This Magnificent Energy Stock Is Down 20%. Buy It Before It Sets a New All-Time High.

Constellation Energy expects 2026 adjusted EPS of $11 to $12, with the midpoint up 55% from 2025 and only slightly below the $11.60 consensus. Management also reaffirmed 20%+ EPS CAGR through 2029, highlighted Calpine acquisition synergies, and raised the dividend 10% to $0.4265 per quarter with a plan for 10%+ annual growth. Near-term pressure remains from the delayed Three Mile Island restart and a weak stock performance, but long-term demand from AI, data centers, and long-dated power contracts supports the outlook.

Analysis

CEG is increasingly becoming a quasi-infrastructure asset rather than a pure utility: the market is still pricing it like an earnings-sensitive power producer, while hyperscaler demand makes a meaningful slice of future output look more like contracted datacenter infrastructure. That matters because long-duration power contracts can re-rate the multiple if investors become convinced earnings are not just growing, but de-risking through 2027-2029. The Calpine asset base also adds optionality in gas-linked regions where data-center load growth is strongest, which should support higher utilization and better pricing leverage than the market is currently modeling.

The bigger second-order effect is that CEG’s competitive edge pressures smaller merchant generators and independent power producers that lack nuclear baseload and balance-sheet scale. If CEG keeps locking in multi-year load with hyperscalers, it can cherry-pick the best contracts while peers face rising interconnection costs and shorter-duration agreements, compressing spreads across the merchant power complex. Microsoft is a validation point, but the more important signal is that the next wave of AI demand can be served by a limited club of dispatchable, carbon-light assets; that scarcity should eventually pull capital toward CEG rather than away from it.

Near term, the main risk is execution slippage on restart timing and transmission, which matters because the stock is likely to trade on visible 2026-27 milestones rather than the 2029 growth path. Another risk is that the market has already partially discounted the AI power narrative; if power procurement cycles slow or contract pricing proves less favorable than expected, the de-rating could continue for another 1-2 quarters. Conversely, if management converts even a modest amount of the remaining generation base into additional hyperscaler deals, earnings revisions can re-accelerate quickly and overwhelm near-term EPS noise.

The dividend guidance is more important as a signal than as income: a steadily rising payout with a low payout ratio tells the market the company sees durable free cash flow, not just cyclical upside. That can attract long-duration, quality-oriented capital and reduce downside volatility, especially if rate-cut expectations keep improving the relative appeal of an asset with contractual growth. The contrarian view is that the selloff may be too shallow relative to the business quality upgrade from Calpine plus AI load, creating a setup where the stock can re-rate before the reported numbers fully catch up.

Key Decisions for Investors

• Go long CEG over a 3-6 month horizon on weakness below recent support; risk/reward favors a re-rating if the market starts valuing contracted AI-linked megawatts rather than just 2026 EPS.
• Pair trade: long CEG / short a merchant power peer with less nuclear or contract visibility over 3-9 months; thesis is that scarce dispatchable supply with hyperscaler demand should outperform exposed generators as contract economics diverge.
• Use call spreads on CEG for a 6-12 month window rather than outright equity if concerned about restart timing volatility; upside is tied to catalyst realization while downside is cushioned by limited premium outlay.
• Trim or hedge if CEG fails to show additional contract announcements within the next 1-2 quarters; the stock is vulnerable to narrative fatigue if AI power demand does not convert into backlog growth.
• Add MSFT as an indirect beneficiary only on pullbacks if you want exposure to the same power-secured AI capex cycle without single-asset execution risk; the hedge is that hyperscalers with balance-sheet strength will secure scarce power first.