← Feed📈 Board🎯 Predictions🚧 Bottlenecks📐 Calibration📊 AnalysisCEG Thesis2026-10-23 12:00 PM · Constellation Energy Corporation

CEG — Constellation Energy Corporation

$252.54-2.67 (-1.05%)
Open 253.12 · High 253.91 · Low 252.32 · Vol 1,655,424
Delayed quote · 2026-10-23 11:00 AM · change vs prev close · source: Stooq
Running conviction (P)
0.75prior 0.68

Investment Read as of 2026-10-23

The Read

CEG remains a SOLID investment based on its unique position providing reliable power exactly when AI demand spikes. The recent flurry of hyperscaler deals and industry filings confirm strong, sustained interest in firm, carbon-free baseload power, supporting the running conviction of P=0.75.

Bull case

Bear case / what breaks it

What the latest signal says

The recent Google Virtual Power Plant deal source reinforces the thesis by showing hyperscalers are not just buying raw megawatts, but complex, reliable grid services—exactly what CEG's baseload power provides. The high volume of regulatory filings also confirms sustained industry focus on nuclear infrastructure.

Posterior history

DatePΔCallDriver
2026-10-230.75-0.07DOWN/MEDGoogle expanded its PJM virtual power plant (pooled flexible demand paid like a power stat
2026-10-160.82-0.13REVIEW/-Strengthening, but the tracker is overconfident. The PPA pillar now has real proof: three
2026-10-130.95+0.10UP/HIGHMultiple signals (S2, S28) highlight advanced nuclear applications beyond the grid, reinfo
2026-10-090.85-0.08REVIEW/-Strengthening. In under a week, three hyperscalers have signed 20-year deals: Microsoft (8
2026-10-060.93+0.05UP/MEDGoogle signed 20-year PPAs totaling 3,590 MW in PJM (890 MW of new nuclear uprates + 2,700
2026-10-020.88+0.03REVIEW/-Strengthening — a second hyperscaler signed a 20-year PPA and the NRC cleared the Crane re
2026-10-020.85+0.05UP/MEDAmazon reported joining the hyperscaler-rescues-existing-nuclear trade at Calvert Cliffs —
2026-10-010.80-0.07DOWN/MEDPJM shelved its data-center power auction because nobody agrees who pays the bill — the co
2026-10-010.87+0.02UP/LOWAmazon signed a 20-year PPA with Constellation for 190 MW at Calvert Cliffs — a second hyp
2026-09-300.85-0.07DOWN/MEDSame FERC/PJM delay names CEG directly. Doesn't touch the fleet or the MSFT PPA, but it sl
2026-09-260.92+0.02UP/LOWNRC (with DOE's Energy Dominance Financing office as cooperating agency) issued an environ
2026-09-250.90+0.05UP/MEDNRC issued an environmental assessment and finding of no significant impact for the Christ

Thesis detail

Core thesis

Largest US nuclear operator (~55 GW post-Calpine, ~95% capacity factor) sitting

on irreplaceable firm baseload exactly when AI is starved for it — converting

uncontracted nuclear into 20-year hyperscaler PPAs at scarcity prices.

Pillars (with priors)

1. Existing nuclear fleet = the scarcest firm, carbon-free power for AI · P = 0.85

2. Hyperscaler PPAs (MSFT 835 MW; ~147M MWh uncontracted) re-rate the fleet · P = 0.70

3. Crane / TMI-1 restart adds ~835 MW by Q3–Q4 2027 · P = 0.60

4. Calpine synergies land as underwritten · P = 0.55

Expected news (the prior)

Surface the residual: anything *not* on this list (a cancellation, a delay,
a competing supply source) is the signal.

Thesis-breaking triggers (→ set P near 0)

Leading vs lagging indicators

Key metrics

Valuation anchor

Premium (~27x fwd P/E); priced for execution. Consensus ~$368 (range $272–441).

The premium is the bear's lever.

Cross-arena sensors

B2 (PPAs/capacity/restart), B1 (AI demand + the efficiency risk), B7 (FERC/policy).

Posterior log