CEG

Current Thesis

The nuclear-baseload-for-AI narrative has gone fully mainstream and CEG is now the crowded, mature leg of an otherwise still-accelerating industrial-power-ai theme. The binary that mattered — the Q1 2026 print on 2026-05-11 — came and went: CEG beat and reaffirmed 2026 guidance of $11–$12/sh, and the stock DIPPED ~3% on the print (classic sell-the-news). Since then the tape has only gotten heavier: Third Point (Dan Loeb) fully exited its CEG position (13F disclosed 2026-05-15), and on 2026-06-01 CEG priced an 11M-share secondary at $281 with a concurrent 2M-share repurchase from underwriters — a supply overhang that pins spot near $281 and tells you holders are happy to distribute here. Mizuho only carries Neutral, PT $310 (2026-05-12). This is not an accelerating setup; it's post-catalyst digestion under a fresh equity overhang. We do NOT chase CEG at ~$280. We wait for a pullback that holds or a genuinely new catalyst (second-tier hyperscaler PPA or a favorable FERC co-location ruling).

Bull Case

• Q1 2026 beat + reaffirmed $11–$12/sh 2026 guide (2026-05-11) — first quarter with Calpine consolidated; guidance held, no cut. At ~$281 that's ~24x forward, premium to regulated utes (18–20x) but no longer the 28–32x stretch of April.
• Structural demand cited at the index level: Benzinga's 2026-06-01 "$700 Billion / watching the wrong companies" piece reframes AI as a power-and-grid capex story — power generation is now the consensus second-order AI trade, and CEG is the cleanest pure play.
• Behind-the-meter optionality still live: FERC docket ER24-2172 (Amazon/Susquehanna co-location ISA) pending in Q2 2026; a favorable ruling unlocks a template across Byron, Braidwood, LaSalle, Calvert Cliffs.
• IRA §45U nuclear PTC floor (~$43.75/MWh, inflation-adjusted) through 2032 caps merchant downside even if power prices roll.
• MSFT/TMI Unit-1 restart on track for 2026-09-20 — physical proof that retired nuclear can be re-commissioned for hyperscaler load.
• Capital recycling: Pine Creek RNG minority stake (2026-05-14) shows management leaning into long-term offtake/clean-attribute deals, not just merchant exposure.

Bear Case

• Catalyst already fired, sold off: a Q1 beat that drops 3% is the tell that the easy narrative money is in. The re-rate trigger we were waiting for in the last dossier has passed without a new leg.
• Smart money exiting: Third Point sold its entire CEG stake in Q1 (2026-05-15). (Note: Berkshire's 2026-05-15 trim was Constellation Brands / STZ — a different company; do NOT misread it as a CEG signal.)
• 11M-share secondary at $281 (2026-06-01) = ~$3.1B of supply hitting the float plus a signal that holders see fair-to-full value here. Equity overhang caps near-term upside.
• FERC precedent is bearish: FERC rejected the Amazon/Talen co-location amendment 2024-11-01 (2-1). A second rejection of the CEG structure is a 15–25% down-day.
• Calpine debt + gas drag: ~$12B incremental debt; gas now ~45% of MWh. Henry Hub sub-$2.50 compresses spark spreads. Moody's review still an overhang.
• Mainstream saturation flags everywhere: Goldman/Barclays/CNBC/Benzinga all on the story; Mizuho only Neutral. Best risk/reward was 2024–2025, not at $280 with a secondary printing.

Setup & Price Structure

• Spot ~$281 (anchored by the 2026-06-01 secondary price). No live quote this session — confirm on first fill.
• Consolidating / range-bound post-earnings; not RSI-extended, but not breaking out either. The 3% earnings dip + secondary = capped tape.
• Resistance: prior-high zone ~$300–$310 (Mizuho PT $310). Upside to Barclays $360 exists on paper but momentum has stalled.
• Support: ~$260 (pre-offering shelf), then [entry redacted] (200-DMA + post-Calpine gap fill) — the line that defines the trade.
• Archetype-3 trim discipline: exit on weekly close below 20-EMA, not on RSI>75. As a watchlist name, the equivalent is: no fresh entry while price is below the 20-EMA / inside the secondary overhang.

Catalyst Calendar (next 30 days)

• ~2026-06-03 (est.) — Secondary offering settlement: 11M shares @ [entry redacted] hit the float; near-term supply pressure.
• Q2 2026, undated — FERC docket ER24-2172 (Amazon/Susquehanna co-location ISA amendment) ruling could drop any time; binary 15–25% mover in either direction. No firm date — watch the docket, not a calendar.
• ~2026-08-05 (est., OUTSIDE 30d) — Q2 2026 earnings: first clean full-quarter Calpine run-rate and the next real synergy-raise window.
• 2026-09-20 (OUTSIDE 30d) — MSFT/TMI Unit-1 restart, physical proof-of-concept.
• Net: no hard dated CEG catalyst inside the next 30 days. Tape will be driven by theme flow, AI-capex sentiment, and 10Y yields — not a company event.

What Would Change Our Mind

• Flip to BUY / size up on: a NEW named hyperscaler PPA (GOOG/META/ORCL/AMZN) at a CEG nuclear site; a FAVORABLE FERC co-location ruling on ER24-2172; or a pullback to $240–$250 that holds with the industrial-power-ai theme still accelerating.
• Confirm SKIP / stay out on: weekly close < [entry redacted]; 2026 adj-EPS guide cut below the $11 floor; an AI-capex pause signaled by MSFT/META/GOOG; a Moody's downgrade on Calpine debt; or a second FERC co-location rejection.

Correlation Notes

CEG trades as the large-cap anchor of industrial-power-ai — high co-movement with VST, TLN, NRG, and GEV, plus secondary correlation to uranium (CCJ) and AI-capex bellwethers (NVDA/MSFT capex guidance). Gas leg (~45% of MWh post-Calpine) ties P&L to Henry Hub. As a high-multiple regulated-adjacent name it carries rate sensitivity — rising 10Y yields compress the multiple. The single biggest macro factor: "is hyperscaler AI capex still ripping?" — a capex-pause headline hits CEG harder than its EPS would imply because the premium is narrative, not cash flow.