Vistra Corp — 8.0/10 — $146.87

Vistra Corp is the largest competitive power producer in the United States and the dominant integrated power-and-retail platform in Texas. The business spans ~41 GW of generation across ERCOT, PJM, NYISO, ISO-NE, MISO, and CAISO -- including ~6.4 GW of nuclear (the #2 US merchant nuclear fleet) plus the only contractable nuclear plant in Texas (Comanche Peak, 2.4 GW). Through TXU Energy, VST serves ~5M+ retail customers, holding ~25-30% of ERCOT residential retail share in an effective duopoly with NRG.

The bull case rests on four monetized monopolies plus an embedded options stack. (1) Comanche Peak -- single-asset ERCOT nuclear monopoly -- already monetized via the Amazon-Comanche Peak 1,200 MW 20-year PPA signed September 2025. (2) ERCOT thermal -- VST #1 in a 4-player oligopoly with extreme demand growth (LTRA3 peak load 85 GW to 139 GW by 2030, +63 GW with ~85% from data centers/crypto/electrification). (3) ERCOT retail -- TXU duopoly with NRG. (4) US merchant nuclear -- 6-player oligopoly with VST at #2 -- already monetized via the Meta-PJM 2,609 MW PPA signed January 2026 across Perry, Davis-Besse, and Beaver Valley (the largest corporate nuclear PPA in US history). Embedded options include 3.2 GW of un-contracted nuclear (Beaver Valley plus Comanche Peak uprate), a 4,500 MW organic gas / bridge / Permian pipeline, and Cogentrix accretion explicitly excluded from 2026/27 guidance.

The quality profile is best-in-class. 3 of 3 quality gates pass -- no scoring cap. 8-year public track record under the same operational leadership (Burke as CEO since August 2022 and COO/President since 2008; Moldovan as CFO since 2022). Promise hit rate ~100% across 12 tracked promises -- 2024 EBITDA beat top of range by $600M+, 2026 guidance raised three times, Energy Harbor delivered $900M vs $700M promised. Capital allocation has been exceptional: $6B cumulative buybacks at a ~$36 average price created ~$20B+ of value at today's $147; share count -28%; investment-grade ratings achieved early (S&P Q4 2025, Fitch Q1 2026). The only checks against management are the Moss Landing battery fire ($400M write-off, January 2025) and the pending Cogentrix integration.

Bottom line versus CEG and TLN. CEG (composite ~13x 2026E EBITDA) is the consensus AI-power name; VST trades at ~9-10x despite arguably better-positioned commercial wins. TLN (composite 6.80) is single-asset concentrated (Susquehanna ~50% of equity value) and fails the 3-year track record gate (less than 2 years public). VST is the cleanest expression of the AI-power thesis -- highest-conviction name in the entire screen.

Vistra receives a composite score of 8.00/10, the highest in the IPP cohort and the highest-conviction AI-power name in the entire screen. All 3 quality gates pass -- no scoring cap applied. The score reflects best-in-class scoring across Financial Trends (8), Thematic Exposure (9), and Management Quality (9), offset only modestly by Investor Sentiment (6, inverted -- still consensus) and Concerns/Risks (7). Probability-weighted fair value ~$160-180 vs spot $146.87 -- asymmetric to the upside.

Bull case (~$200-240, +36% to +63%): Next hyperscaler PPA signed on un-contracted 3.2 GW or Comanche Peak uprate -- Q1 2026 mgmt commentary describes interest 'as high as ever.' Cogentrix closes Q3 2026 and accretion explicitly excluded from current guide proves out at $700-750M to 2027 EBITDA per Moldovan. ERCOT capacity auction prints validate the 5-6% load CAGR mgmt has been signaling but forward curves don't reflect. Multiple re-rates toward CEG ~13x. Buybacks continue at $525M/4mo pace.

Base case (~$160-180, +9% to +22%): Cogentrix closes on time and is modestly accretive. Existing PPAs (Amazon Comanche Peak, Meta PJM) execute on schedule with no surprises. ERCOT and PJM demand growth continues but multiple expansion is muted by AI capex narrative concerns. Buybacks continue, share count compresses further. Mgmt delivers another ~100% hit rate year. Stock grinds to fair value range as commercial wins flow through P&L.

Bear case (~$100-130, -32% to -11%): AI capex narrative cracks -- multiple compresses from ~9-10x to ~7-8x across the IPP cohort. FERC PJM behind-the-meter overhang escalates (though less acute than TLN since VST's PPAs are FOM structure). Moss Landing residual liabilities expand. Coal retirement political pressure forces accelerated stranded-cost writedowns. Cogentrix integration disappoints. Even in this scenario, IG balance sheet and ongoing buybacks provide downside support.

Bottom line: VST is the highest-quality, best-priced way to own the AI-power thesis. Four distinct local oligopolies, two largest corporate nuclear PPAs in US history already signed, 8-year public track record with ~100% hit rate, $20B+ of value created via buybacks at ~$36 avg, IG balance sheet, ~9-10x 2026E EBITDA -- a clear discount to CEG at ~13x. 33% pullback from $220 has compressed the multiple and de-positioned the trade. BUY.

Key catalysts and monitoring points:

• Next hyperscaler PPA announcement: 3.2 GW un-contracted nuclear (Beaver Valley + Comanche Peak uprate) plus Vistra Zero brownfield optionality. Q1 2026 mgmt described hyperscaler interest as 'as high as ever.' Single biggest near-term catalyst -- another Meta-scale deal would re-rate the multiple.
• Cogentrix close (Q3 2026) and integration commentary: Acquired at 7.25x 2027 EBITDA. Accretion explicitly excluded from 2026/27 guide -- Moldovan implied +$700-750M to 2027 EBITDA. First quarter of consolidated reporting will be a key proof point.
• ERCOT and PJM capacity auction prints: Forward curves don't currently reflect even mgmt's own 5-6% ERCOT load CAGR. Each auction print is a referendum on the data-center demand thesis -- and feeds directly into uncovered position economics.
• Buyback pace ($525M in first 4 months of 2026): Annualized run-rate ~$1.6B is aggressive relative to free cash. Continuation signals mgmt views the stock as undervalued; pace moderation would be a yellow flag.
• Moss Landing residual liabilities: $400M write-off booked. Watch for any additional claims, reputational impact on battery storage strategy, or regulatory follow-on.
• Coal retirement pathway: Political pressure plus potential stranded costs. Texas coal economics depend on ERCOT scarcity pricing -- any policy shift on coal retirement timing materially impacts the bridge to the brownfield gas pipeline.
• FERC PJM behind-the-meter ruling: Sector-wide overhang. Less acute for VST than TLN (VST's nuclear PPAs are FOM structure from inception) but still relevant for sentiment.
• Insider activity: Heavy Burke/Moldovan 10b5-1 selling through 2025 at $160-200+ has tapered. Zero open-market insider buys despite the 33% drawdown -- any open-market buy would be a meaningful positive signal.

For the full analysis, see the Business Model, Financials, Thematics, Management, Valuation, and Sentiment pages.

VST and TLN both express the merchant-nuclear / AI-power thesis. The screen scored VST at 8.00 and TLN at 6.80. The table below shows the 10 axes that drive the 1.2-point composite gap -- VST wins 9 of 10.

Reading the table: TLN's only win is on the single-asset concentration axis -- and even there it's a double-edged sword. Concentrating ~50% of equity value in Susquehanna is what made TLN the highest-beta AI-power name in 2024, but the FERC behind-the-meter ruling shows the downside of single-asset / single-counterparty exposure. VST's PPAs are FOM structure from inception, deliberately designed to sidestep the BTM overhang. Across every quality, valuation, and execution axis, VST is the structurally superior way to own the same thesis.