VST

VST is a large fish in a structurally improving pond. The competitive merchant generation sector spent a decade as a bad pond (low capacity prices, excess supply). That reversed sharply: PJM 2027/28 capacity cleared at $333.44/MW-day vs. $28.92 three years prior. ERCOT faces 5–6% annual load growth driven by data centers and industrial reshoring. The nuclear fleet is the key moat — high barriers, zero-carbon credentials, and 20-year hyperscaler PPAs that smaller gas-only peers cannot replicate.

Snapshot

What it does

Vistra Corp is the largest competitive (merchant) power generator in the United States by installed capacity — roughly 41 GW spanning gas, nuclear, coal, and solar across ERCOT, PJM, ISO-NE, NYISO, MISO, and CAISO. Unlike regulated utilities, Vistra earns market prices for its electricity rather than a guaranteed rate-of-return, meaning its revenues fluctuate with wholesale power prices and capacity auction outcomes.

The Energy Harbor nuclear acquisition (closed 2024) added ~6 GW of zero-carbon baseload in PJM: Beaver Valley (two units), Davis-Besse, Perry, and Comanche Peak — transforming VST from a largely gas/coal merchant into a diversified nuclear-and-gas platform. Two signed 20-year hyperscaler PPAs (Meta 2,609 MW; AWS ~3,800 MW) lock in nuclear revenue well into the 2040s.

On the retail side, TXU Energy and related brands control ~32% of the ERCOT residential electricity market — a natural (though imperfect) hedge against wholesale price spikes.

What it's planning

1. Cogentrix acquisition (~$4B, H2 2026): 10 modern gas plants adding ~5.5 GW across PJM, ISO-NE, and ERCOT. Pre-funded with a $4.0B April 2026 senior unsecured note issuance. Subject to FERC/HSR/state approvals — management confirmed on-track in the Q1 2026 call.
2. Additional nuclear PPAs: Management identifies ~3.2 GW at Beaver Valley and Comanche Peak as near-term contracting opportunities. Customer engagement described as "historically high." No dates committed — treat as optionality.
3. Capital returns: Targeting ~$3B of the guided $10B+ 2026–2027 cash generation for shareholder returns via buybacks and dividends. Q1 2026 buybacks totalled $379M (2.37M shares).
4. AWS delivery ramp: Comanche Peak ~1,200 MW PPA begins Q4 2027, ramping to full capacity by 2032.

Catalysts & demand drivers

Near-term (next 12 months):

• PJM 2028/29 capacity auction (June 2026): Single most important binary event for the stock in the near term. If it clears at or above $333.44/MW-day (2027/28 level), it extends price visibility by one planning year and defeats the "peak cycle" objection.
• Cogentrix close (H2 2026): Clean close with updated guidance forces upward revisions to 2027 EBITDA estimates most models do not yet reflect.
• Q2–Q4 2026 earnings: Quarterly confirmation of the $6.8B–$7.6B 2026 EBITDA guidance range (Q1 was $1.494B, +20% YoY).
• Buybacks: ~$379M repurchased in Q1 2026 alone; ongoing mechanical per-share support.

Structural (multi-year):

• AWS ~3,800 MW PPA (Q4 2027 ramp, full 2032) + Meta 2,609 MW PPA (underway)
• Section 45U nuclear PTC (up to $15/MWh, survived first House test May 2025; excluded from guidance = undisclosed upside cushion)
• AI/data-center load growth driving structurally higher PJM capacity prices

Track record

Revenue (XBRL-confirmed):

Note: FY2025 GAAP operating income fell sharply from FY2024 ($4.08B → $1.91B) on higher revenue, driven by mark-to-market hedge adjustments and non-cash charges — not a deterioration of underlying cash generation. Operating cash flow held at $4.07B.

Balance sheet highlights:

• Net debt: ~$19.3B (external sources; not in XBRL fact sheet)
• D/E: ~330% (external sources)
• Cash at 2025 year-end: $785M (XBRL) — thinner treasury as largest acquisition approaches
• Investment-grade: S&P BBB- (December 2025), Fitch BBB- (March 2026)

Share count (XBRL-confirmed):

• 2020-end: 489.3M
• 2025-end: 338.1M
• Reduction: 31% over five years — sustained buyback execution

Dilution risk: Low. Share count declining consistently since 2020.

Valuation

At $157.97/share and 338M shares (XBRL), market cap is $54.0B. Using ~$19.3B net debt (externally sourced), EV ≈ $73.3B.

• EV/EBITDA (2026 guided midpoint $7.2B): ~10.2x
• FCF yield: ~8% (FCFbG midpoint ~$4.3B / $54B market cap)
• Peer CEG: ~13x EV/EBITDA — reflecting purer nuclear balance sheet and lower leverage
• TIKR quantitative 2030 model: $184 target (only ~16% total return from current levels over four years on a pure fundamental basis)
• Street consensus: $225 mean (18/20 buy-equivalent) — ~42% above current price

Fair value range: $165–$210

• Low end (~$165): Applies ~9x 2026 EBITDA midpoint ($7.2B) to EV, backs out $19.3B net debt, divides by 338M shares. Reflects no near-term re-rating, debt paydown required for meaningful equity appreciation.
• High end (~$210): Assumes Cogentrix adds $600M–$800M incremental EBITDA, 2027 run-rate ~$7.8B–$8.0B, valued at 9x, minus net debt after partial paydown. Requires clean execution.
• Method: EV/EBITDA on 2026–2027 guided figures, backed out with external net debt estimate. Note: the bull's stated "7.5x EV/EBITDA" is a methodological error (market cap / EBITDA, not EV/EBITDA); the correct figure is ~10x.

Ownership & insiders

• Short interest: 3.99% of float — low. No large skeptical camp.
• Buybacks: 31% share count reduction (2020–2025, XBRL); $379M in Q1 2026 alone — demonstrated execution, not stated intention.
• Institutional: Investment-grade upgrade (December 2025 / March 2026) opens the company to institutional mandates that previously excluded it as high-yield.
• Insider selling: No evidence of aggressive insider selling from research or public sources.
• Analyst coverage: 18/20 buy-equivalent ratings — when consensus is this lopsided, incremental information risk is to the downside.

Bull case

The re-rating thesis: VST trades at ~10x 2026 EV/EBITDA — a ~3x discount to CEG (~13x). That discount is partly explained by higher leverage and a messier retail book, but on a free-cash-flow basis it is hard to fully justify: guided FCFbG midpoint ~$4.3B against a $54B market cap implies a ~8% FCF yield.

Why the EBITDA is structural, not cyclical:

• PJM 2027/28 already cleared: 10,566 MW at $333.44/MW-day (vs. $28.92 three years prior). This is locked, not projected.
• Meta (2,609 MW) and AWS (~3,800 MW) PPAs are signed 20-year contracts. Near-term nuclear revenue is largely insulated from demand-cycle risk.
• Nuclear PTC (Section 45U, up to $15/MWh through 2028, then reduced) is excluded from guidance entirely — any legislative survival is upside the base case doesn't reflect.

Per-share compounding: Share count fell from 489M (2020) to 338M (2025-end, XBRL) — a 31% reduction. With buybacks continuing at ~$379M/quarter, per-share FCF grows even if enterprise multiple stays flat.

Investment-grade demand catalyst: The S&P/Fitch upgrades to BBB- open VST to institutional mandates that excluded it as high-yield — an incremental buyer base not yet fully reflected.

Bear case & red flags

1. Leverage spike into a cyclical earnings peak — more debt arriving Net debt ~$19.3B against guided 2026 EBITDA of $6.8–$7.6B = 2.5–2.8x net debt/EBITDA at peak earnings. The April 2026 $4B unsecured note pre-funds Cogentrix before its EBITDA flows. FY2025 cash fell to $785M (XBRL) from $3.49B in FY2023 — the treasury is thinner precisely as the largest-ever acquisition approaches. D/E ~330%. Severity: HIGH.

2. GAAP earnings are volatile — FY2025 proves it FY2025 GAAP net income: $944M (XBRL). FY2025 GAAP operating income: $1.91B — less than half of FY2024's $4.08B — on higher revenue ($17.59B vs. $14.76B). FY2021 and FY2022 operating losses (-$1.52B and -$1.18B, XBRL) show this is not hypothetical. The entire bull case rests on Adjusted EBITDA; the GAAP income statement tells a more volatile story. Severity: MEDIUM-HIGH.

3. ERCOT weather concentration — retail is a documented loss generator Q1 2026 retail segment: $68M Adjusted EBITDA on "extremely mild weather." FY2021 operating income was -$1.52B (XBRL), driven by Winter Storm Uri — a single event wiped out multiple years of profits in one quarter. The ~32% ERCOT residential share creates concentration in a notoriously volatile, weather-dependent market with no rate-case protection. In mild weather both retail margins compress and wholesale prices soften simultaneously — the hedge is imperfect. Severity: HIGH.

4. Nuclear PTC legislative risk Section 45U survives the first House reconciliation test (May 2025) but with a 20% reduction starting 2029 and subject to future revision. A rough estimate: ~40 TWh of nuclear output at ~$12/MWh post-phaseout residual = ~$480M/year reduction in credit value. Vistra excludes it from guidance — repeal is not a guidance miss, but investors pricing it in face cliff risk. Severity: MEDIUM.

5. Serial acquisition integration risk Energy Harbor (~2024) and Cogentrix (~$4B, pending H2 2026) represent two major acquisitions across six regulatory jurisdictions (PJM, ISO-NE, ERCOT, MISO, CAISO, NYISO). Management bandwidth, systems integration, compliance surface area, and cultural coherence are all stretched. Historical serial acquirers in the power sector have created and destroyed value in equal measure. Severity: MEDIUM.

6. Hyperscaler capex concentration beyond signed PPAs Meta and AWS represent the overwhelming majority of contracted nuclear capacity. The pipeline beyond those two deals is uncontracted and depends on hyperscalers sustaining AI infrastructure capex. If AI monetisation disappoints or the capex cycle compresses, uncontracted gas capacity faces lower realized prices. Management's ~3.2 GW additional contracting opportunity is optionality, not booked revenue. Severity: MEDIUM.

7. FERC co-location regulatory overhang FERC is actively investigating behind-the-meter data center co-location arrangements. An adverse ruling could restrict structures, impose new interconnection fees, or require contract renegotiation. Binary regulatory risk with no clear timeline. Severity: MEDIUM.

8. Valuation: limited margin of safety at peak earnings EV/EBITDA of ~10.2x prices in peak-cycle power prices, full Cogentrix execution, PTC continuation, and no weather disruption. The TIKR quantitative model's $184 five-year target implies only ~16% total return from current levels. When sell-side targets ($225 mean) are 22% above a rigorous fundamental model, one of them will be wrong. At 3.99% short interest, any negative trigger moves the stock only one direction. Severity: MEDIUM-HIGH.

Interesting findings

• The EV/EBITDA confusion: The bull stated VST trades at "7.5x EV/EBITDA" — that is market cap / EBITDA, not EV/EBITDA. The correct EV/EBITDA figure is ~10.2x (adding ~$19B net debt). The discount to CEG (~13x) is real but a ~3x gap, not ~5.5x.
• The bull's own fair value math undercuts the re-rating story: The bull's realistic upside section calculates that a re-rating to 10x EV/EBITDA implies a per-share value of ~$157 — essentially the current price. Multiple expansion alone doesn't get you there without debt paydown.
• Revenue is not new growth: FY2025 revenue ($17.59B, XBRL) nearly matches FY2021's $17.70B. The EBITDA improvement is real and structural (capacity prices, nuclear PTC, PPAs), but revenue has been above $17B before.
• Cash has drained consistently: From $3.49B (FY2023) to $1.19B (FY2024) to $785M (FY2025, XBRL). The company is profitable on a cash-flow basis but deploying capital aggressively.
• The FY2025 GAAP income collapse is underexplored: Net income fell from $2.66B (FY2024) to $944M (FY2025) on higher revenue. Understanding what drove that divergence (mark-to-market hedge losses, step-up in D&A from Energy Harbor, or other non-cash charges) is essential for modeling whether 2026 GAAP will recover.

The read

Vistra is not broken — operating cash flow of $4.07B in FY2025 (XBRL), two locked-in hyperscaler nuclear PPAs, and a capacity price tailwind already cleared make this a genuine compounder thesis. But it is a leveraged bet on peak-cycle conditions holding, and the market knows it: the stock has been essentially flat (-1.24% 52-week) even as reported earnings grew, suggesting some multiple compression has already occurred.

The bull thesis is strongest on contracted revenues (PJM 2027/28 already locked, Meta/AWS PPAs signed) and the FCF yield (~8% at current price). The bear thesis is strongest on leverage-at-peak-earnings, ERCOT weather asymmetry, and the GAAP volatility that the Adjusted EBITDA narrative papers over.

The honest range: $165–$210 depending on Cogentrix execution, the June 2026 PJM auction result, and whether nuclear PTC survives. Neither a table-pounder nor a short — a "show-me" situation where the next 90 days (PJM auction + Cogentrix close) are the most information-rich period in the near-term thesis.

Research, not investment advice. Figures sourced from SEC filings and public data; verify before acting.