# Equity Research Update: Constellation Energy Corporation (CEG)
**Analyst Note:** This is an UPDATE to my prior thesis dated 2026-04-26 (abandoned, conviction 6/10). Since that abandonment, the stock has appreciated ~6% over the past month, three new 8-Ks have been filed, and news flow continues to emphasize CEG as a primary beneficiary of data center / AI power demand. I will be explicit below about what has changed in my view — and what has NOT.
---
1. THESIS SUMMARY
Constellation Energy is the largest producer of carbon-free electricity in the United States, operating ~31,676 MW of generating capacity heavily weighted toward nuclear (the bulk of the fleet), supplemented by wind, solar, hydro, and natural gas (source: company business description, yfinance). The company sells power across five regional segments — Mid-Atlantic, Midwest, New York, ERCOT, and Other — through a combination of wholesale markets and direct contracts with commercial/industrial customers. Recent strategic emphasis has been on monetizing baseload nuclear capacity through long-duration power purchase agreements (PPAs) with hyperscale data center operators.
The core investment thesis is straightforward: CEG owns an essentially irreplaceable fleet of dispatchable, zero-carbon, 24/7 baseload generation at the precise moment that AI/data center load growth is creating an unprecedented squeeze on firm clean power. New nuclear capacity in the U.S. takes 10+ years to permit and build (source: well-established industry data; NRC licensing timelines). This creates a structural scarcity premium for existing operating reactors that CEG can extract via long-term contracts and capacity payments.
**The moat:** Existing nuclear assets with operating licenses are effectively non-replicable on any relevant investment horizon. CEG has scale advantages in nuclear operations (lower per-MWh O&M), regulatory expertise, and increasingly, customer relationships with hyperscalers willing to pay above-market rates for clean firm power. This is a real moat — but it is concentrated in a finite set of assets whose economics depend heavily on power-price and regulatory dynamics.
---
2. BULL CASE
**AI/Data center demand is structural, not cyclical.** Hyperscaler capex commitments and load forecasts from grid operators (PJM, ERCOT) have repeatedly been revised upward. CEG's nuclear fleet sits in the most demand-constrained zones (PJM Mid-Atlantic). If realized hyperscaler PPAs continue at premium rates, forward EPS estimates ($13.59 vs. TTM $7.41 — source: yfinance) imply meaningful earnings step-up by 2027-2028.
**Forward P/E of 22.3x vs. TTM P/E of 41x** (source: yfinance) suggests the market is already pricing in significant earnings growth — but if CEG can deliver above the consensus EPS path via incremental PPAs and/or uprates, multiple compression to a more utility-like 18–20x forward becomes a tailwind, not headwind.
**Policy tailwinds remain intact.** The IRA's nuclear production tax credit ($15/MWh floor, source: IRA legislative text) provides downside protection on power prices. Bipartisan support for nuclear (a rare zone of agreement) reduces regulatory tail risk.
**Revenue growth of 12.9% (source: yfinance)** in a sector that typically grows 2–4% confirms CEG is monetizing the demand surge in real-time, not as a future promise.
---
3. BEAR CASE
**Valuation has run hard and offers limited margin of safety.** P/B of 6.5x and EV/EBITDA of 20.6x (source: yfinance) are extreme for a utility. Even on forward P/E of 22.3x, this is priced as a growth stock, not a regulated power producer. ROE of 16.4% on D/E of 63.9 (source: yfinance) is solid but not exceptional. **Free cash flow of just $1.26B against a $110B market cap is a 1.1% FCF yield** — a key concern unchanged from my prior abandoned thesis.
**Hyperscaler PPA economics are unproven at the regulatory level.** The FERC rejection of the Talen-Amazon "behind-the-meter" arrangement set a precedent that complicates CEG's ability to extract premium pricing without grid-cost-sharing concessions. This risk has not been fully resolved.
**Single-event tail risk on nuclear assets.** An unplanned outage at a major reactor, a refueling overrun, or any safety incident has outsized financial and reputational impact. Concentrated asset base = concentrated operational risk.
**Power price mean reversion.** If natural gas prices stay low or new gas combined-cycle / SMR / battery capacity comes online faster than expected, the scarcity premium for nuclear could compress. Wholesale power markets are notoriously cyclical.
---
4. EXIT CONDITIONS
I would re-engage (upgrade) or formally re-abandon this thesis upon:
**FERC rulings** that substantively limit behind-the-meter hyperscaler PPAs across the industry (bear trigger).
**Forward EPS revisions downward** below $12 for FY27 (would invalidate the growth narrative).
**Stock breaching $250** without an accompanying fundamental deterioration — would re-evaluate as a value entry.
**A new hyperscaler PPA disclosure** with disclosed pricing and term — would materially de-risk the thesis (bull trigger). The 8-Ks filed 2026-03-26, 2026-03-31, and 2026-05-01 may contain such information; I have not yet reviewed primary filing text and flag this as an information gap.
**Material change in IRA nuclear PTC** under any future legislation.
---
5. 5-YEAR EXPECTED OUTCOME RANGE
**Bear (~25% probability):** Hyperscaler PPA economics get capped by FERC; power prices mean-revert; multiple compresses to 15x forward EPS. Price target: **$200–230** (-25% to -35% TR including dividends).
**Base (~50% probability):** CEG executes incremental PPAs at decent-but-not-spectacular margins; EPS grows to $16–18 by 2030; multiple settles at 18–20x. Price target: **$310–360** (~5–8% annualized TR).
**Bull (~25% probability):** AI demand outstrips even bullish forecasts; CEG signs multiple multi-decade PPAs at premium pricing; EPS reaches $22+ by 2030; multiple holds at 22x+. Price target: **$480–550** (~10–13% annualized TR).
**Probability-weighted return: ~5–7% annualized** — adequate but not compelling for a 3–5 year mandate seeking idiosyncratic alpha.
---
WHAT HAS CHANGED FROM PRIOR ABANDONED THESIS
**What's different:** Continued positive price/news momentum, three 8-Ks suggesting active corporate development, and persistence of the AI-power narrative.
**What's NOT different — and why I am NOT upgrading:** The fundamental valuation concerns that drove abandonment at conviction 6/10 remain intact. P/B of 6.5x, EV/EBITDA of 20.6x, and a 1.1% FCF yield are not rectified by a +6% monthly price move — they are *worsened* by it. Stock is up, fundamentals haven't materially changed, and probability-weighted returns are now thinner. **I am moving from "abandoned" back to active "monitoring" given the news flow, but not recommending.** I want to read the three 8-Ks before adjusting conviction further.
---
```json