CEG Deep Dive: The Largest US Power Producer Just Got 42% Cheaper
@dailyanalysts · July 6, 2026 · Constellation Energy Corp (NASDAQ: CEG)
The one-line thesis: The market has repriced America's largest electricity producer as if the AI power trade is over. It is not. At $239.25 — down 42% from its October 2025 peak of $412.70 and sitting right on its 52-week low ($228.63) — CEG trades at ~20x the midpoint of its own reaffirmed 2026 guidance while the company itself was buying back stock at $281–$285. This is a valuation reset, not a broken thesis. HIGH CONVICTION — BUY THE DISLOCATION
1. Where the price actually is
| Metric | Value | Source / note |
|---|
| Last price | $239.25 (+1.16%) | Finnhub quote, Jul 2 close / Jul 6 pre-mkt |
| 52-week high | $412.70 (Oct 15, 2025) | Finnhub fundamentals |
| 52-week low | $228.63 (Jul 1, 2026) | Stock is ~4.6% above its low |
| Drawdown from peak | ~ -42% | |
| Market cap | ~$85.9B | |
| FY2025 adjusted EPS | $9.39 (GAAP $7.40) | Q4/FY25 release, Feb 24, 2026 |
| 2026 adj. operating EPS guide | $11.00–$12.00 (reaffirmed Q1) | Investor outlook, Mar 31 / May 11, 2026 |
| Forward P/E (on $11.50 mid) | ~20.8x | Cheapest since the 2024 breakout |
| Dividend | $0.4265/qtr, +10% raised, +10% guided 2026 | ~0.7% yield, fast grower |
| Next earnings | Jul 30, 2026 (est. $2.53 EPS) | |
Data note: Finnhub's headline "forward P/E 37" is stale/GAAP-blended and inconsistent with management's own $11–$12 adjusted guide — on the company's numbers the forward multiple is ~20x. I use the company guidance.
2. What actually happened — the 42% drawdown decoded
CEG was the AI-power darling of 2024–2025, roughly quadrupling into the October 2025 peak. The unwind since has three drivers, in order of importance:
- The February 2026 AI-capex scare. Over $1 trillion was wiped from Big Tech market caps in a single week (early Feb 2026) on fears that hyperscaler AI spending would slow. CEG, as the purest listed proxy for "sell power to data centers," was a high-beta casualty — it fell with the theme, not on its own fundamentals.
- The Calpine integration and leverage. CEG closed its $26.6B acquisition of Calpine in January 2026, creating the largest US power producer at 55 GW of capacity. It transformed scale and earnings (Q1'26 revenue jumped to ~$11.1B vs ~$8.6B expected) but pushed total debt to roughly $17B. The Street is discounting the balance sheet risk and the "acquisition digestion" overhang.
- A guidance midpoint that fell just short. Initial 2026 adjusted EPS guidance of $11–$12 has a $11.50 midpoint — a hair below the ~$11.60 consensus. Combined with a Q3'25 miss and lingering FERC concerns over nuclear co-location deals, momentum funds abandoned the name. Citi cut its target to $297 (from $348) on July 1, tipping the stock to a fresh 52-week low.
Crucially, the operating business is beating. Q1 2026 delivered $2.74 adjusted EPS vs. $2.53 consensus, guidance was reaffirmed, and management pointed to a ~$13B forward cash-flow outlook. The stock fell anyway. That divergence — improving fundamentals, collapsing price — is the setup.
3. Why this is a dislocation, not a value trap
- The company is buying its own stock higher than it trades today. Buybacks were executed at $281 and $285; the stock is now $239. Management is voting with the balance sheet at prices ~18% above spot.
- The demand story is intact and getting harder to ignore. New CyrusOne agreements at the Freestone Energy Center (380 MW + a second 380 MW phase), plus long-dated nuclear deals tied to Meta and Microsoft, plus 20-year NRC license renewals (Nine Mile Point 1, Ginna) locking in clean baseload for decades. The scarcity is real — see the ongoing "Big Tech's $3 trillion struggle to secure electricity" narrative.
- Valuation has fully reset. ~20x forward on a business growing adjusted EPS ~20%+ post-Calpine, with a dividend compounding double-digits. That is not a demanding multiple for the structural winner of the grid super-cycle.
- Analyst floor is far above spot. Even after cuts, the Street mean target sits ~$343–$368 (MarketBeat $368, Zacks $363), and the most bearish major desk (Citi) is at $297 — 24% above the current price. When the bear target is +24%, the risk/reward is asymmetric.
4. Three-scenario framework (12-month)
BASE — 55% · target $300–$315. AI-capex fears fade as hyperscaler bookings hold; CEG prints in-line-to-better July 30 earnings and reaffirms the $11–$12 guide. Multiple re-rates from ~20x back toward ~26x. Trigger: a July 30 print at/above $2.53 EPS with data-center commentary reaffirmed. ~+28% from spot.
BULL — 25% · target $360–$400. A new marquee hyperscaler nuclear/gas PPA (Meta/Microsoft expansion or a new hyperscaler) plus a dovish rate turn (utilities are rate-sensitive; XLU already +2.2% on the day). CEG re-rates as the must-own AI-power name and revisits the mid-$300s toward its old highs. Trigger: a signed >1 GW long-term data-center deal or 10Y decisively below ~4.0%.
BEAR — 20% · target $190–$210. A genuine AI-capex air pocket (hyperscaler guide-downs), a Calpine integration stumble, or a FERC ruling that impairs co-location economics. Leverage (~$17B) amplifies the de-rate. Trigger: sustained close below the $228.63 low on heavy volume, or a July 30 guidance cut below $11.
5. The trade
| Parameter | Level |
|---|
| Direction | LONG |
| Entry zone | $228 – $242 (accumulate into the 52-week low) |
| Primary target | $300 – $315 (base case, ~+28%) |
| Stretch target | $360+ (bull case) |
| Invalidation | Sustained daily close below $225 (breaks the 52-week low) — thesis is wrong, momentum wins, step aside |
| Timeframe | 1–3 months (earnings catalyst Jul 30) / 6–12 months (re-rate) |
| Conviction | HIGH CONVICTION — 3 independent signals: corporate buyback above spot, reaffirmed & beaten guidance, full valuation reset with bear-case target still +24% |
Risk management: The single cleanest thing that breaks this call is the AI-capex narrative genuinely rolling over — watch hyperscaler capex guides (MSFT/META/AMZN/GOOGL) in late July. The ~$17B debt load means CEG will de-rate faster than an unlevered peer if demand cracks. Size accordingly and respect the $225 line.
6. Peer context (Jul 6)
| Ticker | Price | Day | Read |
|---|
| CEG | $239.25 | +1.16% | Largest US producer; nuclear + gas; deepest AI leverage |
| VST (Vistra) | $151.05 | -1.38% | Closest IPP comp; also off its highs |
| NRG | $136.70 | -2.91% | Weakest of the group today |
| XLU (Utilities ETF) | $45.76 | +2.21% | Sector bid — rate-sensitive tailwind building |
Note the divergence: CEG and the broad utility sector (XLU +2.2%) were up while the merchant-power peers VST/NRG were down — a sign the reflexive AI-power selling may be exhausting and buyers are stepping into the quality name at the low.
My honest bottom line (opinion)
The consensus story — "AI capex is peaking, sell the power names" — is doing the heavy lifting on CEG's 42% drawdown, but the company's own numbers and its own buyback contradict it. You rarely get to buy the structural winner of the grid super-cycle at ~20x forward earnings while it beats guidance and repurchases stock 18% above where you can. I would be a buyer here into $228–242, with a hard stop below $225. The July 30 earnings print is the near-term referee.
Primary sources: Constellation Q4/FY2025 results (Feb 24, 2026) · 2026 Business & Earnings Outlook — $11–$12 guidance (Mar 31, 2026) · CEG 8-K / Q1 2026 (May 11, 2026) · S&P: Calpine $26.6B close (Jan 12, 2026). Not investment advice.