Constellation Energy's Biggest AI Power Deal Doesn't Start Paying Until June 2027

Constellation raised 2026 adjusted operating EPS guidance to $11.50-$12.50, with the $12.00 midpoint about 28% above 2025's $9.39, after Q2 adjusted EPS rose 34% year over year to $2.55. Meta's 20-year agreement for the clean-energy attributes of Clinton's 1,121 MW nuclear plant begins in June 2027, replacing Illinois zero-emission credits as they expire and providing no contribution to current-year earnings. Additional contracted demand includes Microsoft's planned Crane restart in 2027 and 920 MW of 15- to 20-year nuclear contracts starting in 2029-2032, supporting visible long-term growth despite execution and plant-availability risks.

Analysis

The key underwriting distinction is that this arrangement principally de-risks a policy-support rollover rather than creating a new near-term earnings stream. CEG retains merchant exposure on the physical generation while monetizing the clean attribute separately, preserving upside if Illinois/PJM power prices tighten; however, investors should not capitalize the contract as a full incremental AI-power PPA. Disclosure of contract pricing, escalation and curtailment/availability provisions is the missing variable needed to quantify post-2027 EBITDA uplift.

At roughly 25x current-year operating earnings, CEG already embeds a substantial portion of the multi-year contracted-power narrative. The more important 1-3 month catalyst is whether management raises its long-term earnings framework and demonstrates that Calpine integration, fleet availability and hedging can sustain growth before the contracted nuclear pipeline begins. A nuclear outage, weaker PJM capacity/power pricing, or a delay in the Crane restart would compress the premium multiple faster than it would impair current cash flow.

The contrarian view is that hyperscaler clean-energy procurement does not necessarily signal incremental regional electricity demand: attribute purchases can satisfy corporate accounting targets without removing physical supply from the grid. That makes CEG's real scarcity value dependent on dispatchable capacity and transmission constraints, not merely on additional clean-attribute contracts. META and MSFT face immaterial direct P&L exposure; the second-order beneficiaries are instead merchant generators with PJM/Illinois capacity exposure, while regulated utilities may face political pressure if data-center demand raises retail bills.

Key Decisions for Investors

• Maintain a tactical long CEG / short XLU pair over the next 1-3 months, sized 1:1 beta-adjusted. The trade isolates CEG's contracted-growth and merchant-capacity optionality from broad rate sensitivity; exit if CEG fails to reaffirm its long-term earnings trajectory at the next results or if PJM forward capacity/power curves fall materially.
• Do not chase CEG solely on the 2027 contract. Add only on a pullback toward the low-$260s to mid-$270s, where the entry better accommodates execution risk; upside case is a return toward the prior $400-plus high if management monetizes further contracted capacity, while a restart delay or fleet-availability miss can justify a sub-$250 valuation.
• Monitor PJM capacity-auction outcomes, Illinois clean-energy policy developments, and Crane construction milestones as catalyst alerts. A favorable capacity-price reset or on-time restart would support increasing CEG exposure; a material delay beyond 2027 should trigger a reduction because the premium multiple is most vulnerable to deferred cash-flow recognition.
• Avoid treating META or MSFT as direct beneficiaries of this development. Their relevant trade signal is only indirect: accelerating clean-power procurement could increase data-center build commitments, but capex guidance and utility interconnection disclosures—not clean-attribute purchases—would validate that demand thesis.