Vistra owns one of the largest fleets of nuclear, natural gas and coal power plants in the country, and rising demand for electricity from data centers and artificial intelligence computing is pushing up the prices it can charge, especially in the mid Atlantic grid where several of its nuclear plants sit.
The company has signed twenty year contracts to sell power directly to Meta and Amazon from its nuclear plants, agreed to buy a large natural gas power fleet, and become the preferred power supplier for a new multi billion dollar data center venture, all of which should add to earnings later this decade.
At the same time, Vistra has taken on more debt to pay for this growth, and rising household electricity bills tied to data center demand have triggered political pressure that could cap how much further power prices are allowed to climb.
Today's stock price already gives credit for next year's growth staying on track, without yet paying up for every expansion plan working out. Fairly valued.
Most of its profit comes from running power plants, mainly in Texas and in the mid Atlantic and Northeast grid known as PJM, and selling that electricity into wholesale power markets or through long term contracts. It also has retail businesses that sell electricity directly to households and companies, and it locks in fuel and power prices ahead of time to smooth out swings from weather and commodity costs.
We'd revisit this call if the price falls below $95.00 or rises above $250.00.
These are review triggers for us to take another look, not price targets or a promise the stock will reach them.
| Metric | VST (subject) | Peer median |
|---|---|---|
| EV/Sales | 3.96x | N.M. |
| EV/EBIT | 36.55x | N.M. |
| EV/EBITDA | 17.90x | N.M. |
| P/B | 10.48x | N.M. |
| ROIC | 7.5% | N.M. |
| ROE | 18.5% | N.M. |
Vistra is a large US power generator and retail electricity provider with roughly 41 GW of nuclear, natural gas, coal and renewable capacity, concentrated in Texas (ERCOT) and the mid Atlantic/Northeast (PJM). Classification is CYCLICAL per the deterministic screen, reflecting a multi-decade history as a merchant generator, but the business mix has shifted meaningfully since the 2023-2024 Energy Harbor nuclear merger and the 2026 Meta/AWS contracts and pending Cogentrix gas fleet purchase.
FY2025 revenue was $17,738 million (up from $14,761 million in 2024). GAAP net income was $944 million, EPS $2.18, down sharply from FY2024's $7.00 EPS, but the decline is mostly an $808 million unrealized, non-cash hedge loss, not an operating deterioration; Ongoing Operations Adjusted EBITDA actually rose to $5,912 million from $5,643 million. Q1 2026 flipped the pattern: net income of $1,029 million included a $723 million unrealized hedge gain, on revenue of $5.64 billion (versus $3.93 billion in Q1 2025) and Adjusted EBITDA of $1,494 million, a Q1 record.
FY2025 operating cash flow was $4,070 million; capital expenditures were $2,752 million, up every year since 2020 ($1,259M in 2020). Standardized free cash flow (OCF minus capex minus $113 million of stock compensation) was $1,205 million, matching the deterministic input exactly. Vistra's own non-GAAP metric, Adjusted FCF before growth capital, was $3,592 million for 2025, a much larger figure because it excludes growth-related capital spending; 2026 guidance for that metric is $3.925 to $4.725 billion. Q1 2026 alone had $1,199 million of operating cash flow against $883 million of capex.
Net debt is about $19.3 billion, 2.84x trailing EBITDA, within Vistra's own long-term target of under 3x. The company issued $2.25 billion of secured notes in January 2026 and $4.0 billion of unsecured notes in April 2026 (partly refinancing existing debt) to help fund the pending $4.7 billion Cogentrix acquisition. Both S&P and Fitch upgraded Vistra to investment grade during this period.
See checklist.q10_kpi_pack for hedge ratio, leverage, forward multiple and buyback detail. Return on equity (18.5%) and return on invested capital (7.5%) for 2025 are solid but not exceptional for the sector; 2025 net margin was 5.4%, depressed by the unrealized hedge loss.
The deterministic screen found no valid peer set (gate G4) and fell back to a 100% single-method intrinsic value built on a 10-year median historical price to earnings ratio (13.56x). That method predates Vistra's current nuclear fleet and contracted-growth profile and likely understates fair value materially. Using public data instead: Vistra trades around 10x forward EV/EBITDA versus about 13x for Constellation Energy (CEG), the most contracted, least indebted nuclear peer; Vistra carries about $19.3 billion of net debt versus CEG's roughly $5.9 billion. This discount is a reasonable reflection of Vistra's higher leverage and larger merchant (uncontracted) earnings share, not an anomaly to be arbitraged away.
The dominant macro driver is data center and AI-related electricity demand, which pushed PJM's 2026/2027 capacity auction to the FERC-approved cap of $329.17/MW-day, up from $28.92 two delivery years earlier. That price surge is politically contentious: household bills in PJM territory are projected to rise about $70/month by 2028, and a temporary price cap (from a Pennsylvania governor-brokered settlement) is due to expire after the most recent auction. Rates (10-year Treasury near 4.56%) and elevated beta (1.409) also raise the discount rate applied to Vistra's cash flows.
Recent Form 4 filings (May-June 2026) show a senior finance officer and several directors selling shares, mostly under pre-arranged Rule 10b5-1 plans, at prices in the $160 to $170 range. No material open-market insider buying was found over the same window. This is a mild, not decisive, negative signal.
Catalysts: Cogentrix close (H2 2026), further PJM/ERCOT auction results, 2027 guidance firming as Meta/AWS contracts ramp, further Helix-related contracts. Risks: reinstated or tightened capacity price caps, integration and leverage risk from Cogentrix, continued high growth capex suppressing standardized free cash flow, and the non-cash hedge accounting that makes reported EPS volatile and headline P/E unreliable quarter to quarter.
See checklist.q6_disconfirming for the full list. The single clearest falsification trigger: if PJM's capacity price cap is extended or tightened for 2027/2028 or later, or a future auction clears meaningfully below roughly $330/MW-day, the AI-driven capacity re-rating thesis is falsified and the bear case should take precedence.
Revisit this view if: (a) a capacity price cap is reinstated or extended past the current auction cycle; (b) the Cogentrix deal fails to close or closes on materially worse terms; (c) Meta/AWS contract revenue fails to appear in 2027 guidance updates; (d) standardized free cash flow does not recover as growth capex moderates.
Mechanical model output, a screen, not a fair value.
A scenario range of enterprise value to Adjusted EBITDA multiples (7.5x bear, 9.5x base, 12x bull) applied to 2026 to 2027 guided and near-term visible Adjusted EBITDA, less roughly $19 to 20 billion of net debt and divided by about 338 million diluted shares, puts fair value near the base case; the current price sits close to that base case rather than in the bear or bull tail.
The research read is more favorable than the mechanical screen's output.
Confidence = 0.7 × research sub-score + 0.3 × mechanical sub-score. Mechanical sub-score rewards the Layer 1 screen for agreeing with Layer 2's direction (corroboration), not just for being conviction=NORMAL in isolation.