Back to Blog
7-Lens AnalysisVSTUtilitiesNuclear Power

Vistra Corp: 3.8 GW Nuclear Fleet Contracted to Amazon and Meta for 20 Years — Peak Narrative or Structural Moat?

Record $5.9B EBITDA, $4B Cogentrix acquisition, and landmark 20-year nuclear PPAs with hyperscalers. The AI power demand thesis has re-rated this stock years ahead of revenue realization. Our 7-lens committee examined where the scarcity premium is justified and where the narrative has outrun the timeline.

March 22, 202612 min read
FY2025 Adj. EBITDA
$5.9B

Record result, above guidance midpoint

Nuclear Contracted
3.8 GW

20-year PPAs with Amazon + Meta

Acquisitions (6mo)
$5.3B

Lotus ($1.3B) + Cogentrix (~$4B)

Revenue Gap
~2 Years

Data center power: late 2027-2028

Vistra Corp has quietly assembled the most compelling position in the nuclear data center power market. The company operates ~6.4 GW of nuclear capacity across Texas (Comanche Peak) and PJM (Beaver Valley, Perry, Davis-Besse), serves ~5 million retail customers through TXU Energy, and has signed 20-year power purchase agreements with Amazon Web Services and Meta that represent the largest corporate nuclear offtake deals in U.S. history.

The financial performance has been exceptional. FY2025 delivered record $5.9B adjusted EBITDA and $3.6B adjusted free cash flow, both above guidance. The fleet performed well during January 2026's Winter Storm Fern. A Fitch credit upgrade reflects the improved financial profile. Insiders are net accumulating shares across the entire C-suite.

Yet the investment case hinges on a single, critical variable: timeline. Management explicitly and repeatedly states that data center demand “will not meaningfully begin until late 2027 or early 2028.” The stock has already re-rated for revenue that does not yet exist. Our committee of 7 lenses examined whether the nuclear scarcity premium is a defensible moat or a narrative that has outrun the calendar.

Want the full 7-lens analysis with signal assessments and model debates?

Opus + Sonnet ensemble. 7 lenses. 8 signals. 10 debates. Full evidence citations.

View VST Analysis
Central Question
Vistra has contracted 3.8 GW of nuclear capacity to Amazon and Meta for 20 years and posted record $5.9B EBITDA, yet data center power revenue does not meaningfully arrive until late 2027. Is the nuclear scarcity premium justified, or has the AI power narrative outrun the timeline?

Signal Assessments

Revenue Durability
CONDITIONAL
Gravy Gauge

20-year nuclear PPAs provide transformational visibility, but revenue realization is 2+ years away. Current EBITDA reflects legacy operations.

Competitive Position
DEFENSIBLE
Moat Mapper

~6.4 GW nuclear fleet is irreplaceable within 10-15 years. New nuclear costs $10B+/GW. Structural scarcity gives pricing power.

Funding Fragility
STRETCHED
Stress Scanner

$5.3B in acquisitions within 6 months pushes leverage higher at elevated valuations. Fitch upgrade occurred pre-Cogentrix.

Regulatory Exposure
ELEVATED
Regulatory Reader

Multi-agency oversight (NRC, FERC, EPA) with active litigation. PJM capacity reform and NRC uprate approvals pending.

Narrative-Reality Gap
DIVERGING
Myth Meter

Stock re-rated 2+ years ahead of data center revenue realization. Management honestly tempers timeline expectations.

Expectations Priced
DIVERGING
Myth Meter

At ~8.4x EV/EBITDA, valuation requires $7B+ EBITDA by 2028-2029 — demanding flawless execution on multiple fronts.

Capital Deployment
MIXED
Consolidation Calibrator

Cogentrix at ~$727/kW vs Lotus at ~$500/kW — a 45% premium that may reflect narrative-driven inflation. $1B buyback competes with acquisition financing.

Governance Alignment
ALIGNED
Insider Investigator

All 7 C-suite insiders net accumulating shares. CEO's 10b5-1 plan beats VWAP by $10/share. RSU-heavy compensation.

Key Findings

Nuclear Scarcity is Factual, Confirmed Across 3 Lenses

Gravy Gauge, Moat Mapper, and Myth Meter all converge: Vistra's ~6.4 GW nuclear fleet is an irreplaceable asset. New nuclear construction takes 10-15+ years and costs $10B+/GW. SMR alternatives are 5-10 years from commercial viability. For data centers requiring 24/7 carbon-free power, existing nuclear plants are the only option at scale. This is not narrative; it is physics and permitting reality.

The 2-Year Revenue Gap is the Core Vulnerability

Three lenses (Gravy Gauge, Myth Meter, Stress Scanner) independently identify the same problem: nuclear PPA revenue does not materially accrete until data centers are built and operational in late 2027-2028. Current $5.9B EBITDA reflects legacy operations, not contracted hyperscaler revenue. The market has priced through this gap, assuming flawless execution. Any delay announcement from Amazon or Meta could trigger outsized stock price reaction.

Cross-Lens Finding: Management Credibility
An unusual pattern emerged across our lenses: management is simultaneously the beneficiary and the temperer of the nuclear renaissance narrative. CEO Burke and the executive team have been “consistently messaging” the late 2027/2028 timeline for data center demand impact. They benefit from the stock re-rating but are honest about the timeline. Insider Investigator confirmed all C-suite insiders are net accumulating shares. This alignment between conservative guidance and insider accumulation is a positive signal.

Cogentrix at $727/kW vs. Lotus at $500/kW: The Price of Narrative

Vistra acquired Lotus Infrastructure (2,600 MW) in October 2025 for ~$500/kW. Just months later, the Cogentrix acquisition (5,500 MW) came at ~$727/kW — a 45% premium. Both are modern gas generation assets in competitive markets. The price difference likely reflects the data center narrative's intensification between deals. While Cogentrix includes exceptionally efficient plants (sub-7,000 BTU/kWh heat rates), the premium raises the question: did Vistra capture value or buy at the peak?

Temporal Limitation
This analysis uses financial data through Q4 FY2025 earnings (February 2026) and SEC filings through March 2026. The Cogentrix acquisition remains pending regulatory approval. PJM capacity market reform proceedings are ongoing. NRC nuclear uprate timelines are subject to change. All assessments reflect conditions as of March 22, 2026.

Multi-Agency Regulatory Compound Risk

Vistra faces simultaneous regulatory proceedings across NRC (nuclear safety and uprate approvals), FERC (PJM capacity market design), EPA (environmental compliance), and Congressional oversight (IRA nuclear tax credits). Each body operates independently, creating compound risk where adverse outcomes in any two proceedings simultaneously could materially impair plant economics. The PJM capacity market redesign alone could shift hundreds of millions in annual nuclear revenue.

Where Models Disagreed

1

Nuclear Moat: Permanent Scarcity or Transitional Advantage?

OPUS: Effectively permanent through 2035

NRC licensing, 10-15+ year construction, and $10B+/GW capital costs make new nuclear economically infeasible at scale. The scarcity is structural, covering the entire term of current PPAs.

SONNET: Time-limited first-mover advantage

SMR technologies, fusion progress, and hyperscaler investments in behind-the-meter generation (Microsoft/nuclear, Google/geothermal) signal that grid PPAs may be a bridge, not a destination.

Resolution: Converged on DEFENSIBLE (not DOMINANT). Real and durable for 5-7+ years, but calling it permanent overstates certainty given technology evolution.

2

Cogentrix Valuation: Fair Price or Cycle-Peak Premium?

OPUS: Rational relative to replacement cost

New-build gas generation exceeds $1,000/kW. Cogentrix plants have sub-7,000 BTU/kWh heat rates. At $727/kW, Vistra is buying best-in-class assets below replacement cost.

SONNET: Narrative-inflated cycle-peak pricing

Gas asset valuations doubled in 18 months on data center hype. The Lotus deal 3 months earlier was 45% cheaper. Historical cycle-peak utility M&A frequently leads to goodwill impairments.

Resolution: Converged on MIXED. Assets are high-quality but timing and pricing create overpayment risk if data center demand disappoints.

3

Valuation: Justified Re-Rating or Narrative Overshoot?

OPUS: Fundamental risk profile transformation

20-year contracts with investment-grade hyperscalers transform Vistra from cyclical IPP to quasi-contracted utility. Historical utility multiples support 8-10x EBITDA for contracted flows.

SONNET: Re-rating pace exceeds revenue pace

Contracts are not yet generating revenue. Constellation Energy trades at even higher multiples. The market prices nuclear as a scarce commodity rather than doing DCF valuation.

Resolution: Converged on DIVERGING. Narrative and fundamentals move in the same direction, but re-rating pace has outrun revenue realization by 2+ years.

Cross-Lens Reinforcements

Nuclear asset scarcity confirmed across 3 lenses

Gravy Gauge, Moat Mapper, and Myth Meter all independently validate that the nuclear fleet is genuinely irreplaceable. This is the strongest cross-lens reinforcement in the analysis.

Revenue realization gap identified across 3 lenses

Gravy Gauge, Myth Meter, and Stress Scanner converge on the 2-year bridge risk: current earnings do not reflect the contracted future that the stock price embeds.

Management credibility validated across 3 lenses

Insider Investigator (all insiders net accumulating), Myth Meter (honest timeline tempering), and Consolidation Calibrator (disciplined acquisition framework) all support management quality.

What to Watch

CRITICALData center construction milestones

Amazon's Comanche Peak campus and Meta's PJM data center builds are the key milestones. Any delay announcement or capacity scale-back would challenge the revenue bridge assumption that underpins the current valuation.

CRITICALCogentrix acquisition closing and integration

The ~$4B acquisition is pending regulatory approval. Integration of 5,500 MW alongside the recently integrated 2,600 MW Lotus fleet creates organizational strain. Watch for cost overruns or operational issues in the first 12 months.

HIGHPJM capacity market auction results

FERC market design proceedings and capacity price outcomes directly affect Eastern nuclear plant economics. A capacity price decline of 20%+ could reduce annual earnings by hundreds of millions.

HIGHNRC nuclear uprate approval progress

433 MW of uprates (Meta contract) and 200 MW (Comanche Peak potential) require NRC approval. Delays or conditions could affect contracted capacity delivery and PPA terms.

Bottom Line

PROCEED WITH CAUTION

Vistra's nuclear fleet and hyperscaler contracts represent genuinely scarce, high-quality assets with aligned management. The competitive moat is real for the next 5-7 years. However, the valuation assumes flawless execution on a 2+ year timeline, elevated leverage from aggressive acquisitions, and sustained AI power demand growth that remains forward-looking. The narrative-reality gap is the primary concern: the business must grow into its valuation.

Path to More Favorable Assessment

  • • Amazon/Meta data center construction confirms 2027-2028 timeline
  • • Cogentrix closes and integrates smoothly
  • • EBITDA trajectory toward $7B+ by FY2027
  • • Additional nuclear PPA announcements with other hyperscalers

Path to Less Favorable Assessment

  • • Data center build delays or capacity scale-backs
  • • Cogentrix integration issues or financing difficulty
  • • PJM capacity reform reducing nuclear economics
  • • EBITDA at low end of guidance ($5.5B) indicating stall

This analysis is for educational purposes only — it is not a recommendation to buy or sell any security.

Public Sources Used (14 documents)
  • • Annual Report (10-K) — FY2025
  • • Quarterly Reports (10-Q) — Q1-Q3 2025, Q3 2024
  • • Current Reports (8-K) — 10 filings (Oct 2025 - Feb 2026)
  • • Proxy Statement Supplement (DEFA14A)
  • • Q1-Q4 2025 Earnings Call Transcripts
  • • Form 4 Insider Transactions — 20 filings
  • • Form 144 Proposed Sales — 10 filings
  • • CourtListener Litigation Summary — 10 cases

Full Analysis with Signal Breakdowns

Explore the complete 7-lens assessment including debate transcripts, evidence citations, and monitoring triggers for Vistra Corp.

View VST Analysis

This report was generated by the Runchey Research AI Ensemble using primary SEC data and reviewed by Matthew Runchey for accuracy.

This analysis is for educational purposes only and does not constitute investment advice. See our Editorial Integrity & Disclosure Policy and Terms of Service.