Vistra: Turning Power Demand into Durable Returns?
Download case PDF (5 pages)Request teacher’s pack
Teaching case | Information cutoff: October 3, 2026
This case is a fictional outside-investor exercise based on public information available by the cutoff. It does not describe an undisclosed Vistra management meeting or decision.
The investor's question
Vistra is positioned across competitive electricity retail and power generation. In 2026 it reported stronger adjusted operating results, signed long-term nuclear power purchase agreements, advanced a gas-plant acquisition, agreed to sell other gas plants, and committed up to $1 billion to an investment fund focused partly on data-center infrastructure. Meanwhile, large electricity users have become central to forecasts for several U.S. power markets.
You are preparing an investor assessment: How much confidence should be placed in Vistra's ability to convert expected large-load demand into durable cash generation? Separate results already reported from contracts awaiting delivery, planned capacity, and broader demand announcements. What evidence would raise or lower your confidence over the next several years?
An integrated platform
Vistra sells electricity to retail customers and owns generation assets in Texas and other U.S. markets. In its FY2025 filing, the company described retail sales and hedging as important ways to manage exposure to natural-gas prices and wholesale market heat rates; retail can also act as a sales channel for generation. This integration may coordinate supply and load, but it does not eliminate exposure to customer demand, plant availability, hedge prices, market rules, or outages. Vistra FY2025 Form 10-K
The Q2 2026 results present more than one signal. Vistra reported $305 million of GAAP net income and $1.767 billion of Ongoing Operations Adjusted EBITDA for the quarter. The net income included a $472 million unrealized hedge loss expected to settle in future years. Year over year, Adjusted EBITDA rose $418 million; management attributed the increase primarily to higher realized energy and capacity prices and three months' contribution from Lotus, a generation acquisition closed in October 2025. Net income fell $22 million, with management attributing the decrease primarily to increased unrealized mark-to-market derivative losses, mostly offset by higher realized prices/capacity and Lotus. Q2 2026 earnings release
| Adjusted EBITDA by segment | Q2 2026 | Q2 2025 | H1 2026 | H1 2025 |
|---|---|---|---|---|
| Retail | $773m | $756m | $841m | $940m |
| Texas | $311m | $142m | $897m | $632m |
| East | $642m | $418m | $1,443m | $932m |
| West | $68m | $49m | $124m | $111m |
Source: Vistra Q2 2026 earnings release. Figures are company-defined, non-GAAP segment measures; they are not GAAP net income or cash flow. Segment movements can reflect wholesale-market conditions, hedges and changes in asset perimeter, as well as underlying operations.
As of August 3, Vistra said expected generation volumes were hedged approximately 100% for 2026, 94% for 2027 and 72% for 2028. The company reaffirmed 2026 Ongoing Operations Adjusted EBITDA guidance of $6.8–$7.6 billion and Ongoing Operations Adjusted free cash flow before growth (FCFbG) guidance of $3.925–$4.725 billion. These are company-defined non-GAAP measures and forward guidance; they are not realized results. The Q2 filing also describes exposure to wholesale prices, market heat rates, retail load, fuel prices and regulatory/market conditions. Q2 2026 Form 10-Q
In July 2026, a PJM capacity auction cleared at $325 per MW-day for 2028–29 across listed zones; Vistra reported clearing 10,924 MW. This is a capacity-market result, not an energy price or proof that a particular data center will operate. Future auction prices and rules may differ. (Q2 2026 Form 10-Q, MD&A p. 52.)
Several paths, several timelines
Vistra's large-load strategy includes contracts, investment and development, but they do not all represent current operating sales.
| Initiative | Disclosed position as of cutoff | What remains uncertain |
|---|---|---|
| Meta nuclear agreement | A 20-year agreement covers up to 2,609 MW associated with existing output and planned uprates at Perry, Davis-Besse and Beaver Valley. Existing-output delivery was expected to begin in late 2026 and reach full contracted portions by year-end 2027; uprate energy was expected to begin in 2031 and reach full delivery by 2034. | Delivery timing for uprates depends on approvals, engineering and capital decisions. Contract pricing and several risk-allocation terms are not fully disclosed in the reviewed filing. |
| AWS / Comanche Peak | A 20-year agreement for 1,200 MW; delivery was expected to start in Q4 2027 and ramp to full capacity by 2032. | It is a future delivery schedule, not current production. Full public contract economics are unavailable. |
| Helix fund | Vistra committed up to $1 billion to a KKR-managed fund targeting data centers and enabling infrastructure. The initial $500 million commitment was callable; another $500 million is tied to power-supply milestones for existing Vistra assets or may be made at Vistra's election. | Vistra does not consolidate the fund and accounts for its interest under the equity method. Its commitment is not equivalent to ownership of all fund assets or consolidated data-center revenue. |
| Cogentrix | Agreement to acquire 10 gas facilities totaling about 5,500 MW across PJM, ISO-NE and ERCOT. FERC approval was received in August; closing remained subject to other conditions and was expected late 2026. | Pending acquisition at cutoff; closing, financing, integration and subsequent performance remained ahead. |
| Casco Bay, Beaver Falls, Syracuse | June 2026 agreement to sell a 754 MW gas-plant group in ISO-NE and NYISO. The assets were held for sale at June 30; closing was expected in H2 2026 subject to approvals and other conditions. | Pending disposition at cutoff. If completed, the East-segment generation perimeter changes. This is a sale, not an addition like Cogentrix. |
| ERCOT development | Plans included up to 2,000 MW of gas generation and upgrades, including 860 MW of Permian peakers expected in 2028 and a possible coal-to-gas repower at Coleto Creek. | Planned capacity is not completed capacity. Timing, approvals, construction, equipment availability and costs matter. |
The source for the transaction and PPA facts is Vistra's Q2 2026 Form 10-Q, Notes 2 and 7 and MD&A pp. 53–54. Lotus is already acquired: its seven gas facilities added approximately 2,600 MW across four markets. Company management identified three months of Lotus contribution in its Q2 year-over-year explanation. Do not treat acquired capacity or its contribution as organic load growth.
Load growth: forecasts are not deliveries
Vistra's Q2 filing cited data centers, oil-field electrification and electric vehicles as contributors to projected fast load growth in regions it serves. The company also reported that labor and supply-chain shortages raised equipment costs and lead times, leading it to defer or abandon some solar and battery capital projects. ERCOT publishes system load forecasts and scenarios, but these are not Vistra customer contracts. Interconnection interest, forecasts, signed power agreements and realized consumption are different measures.
Academic work provides mechanisms and boundary conditions rather than a Vistra return estimate. Brown and Sappington's theoretical electricity-market model finds that vertical integration often reduces retail prices and increases capacity investment and welfare, while universal integration can lower aggregate industry profit. The results depend on stylized market assumptions. Brown & Sappington (2022)
Hortaçsu and Puller studied ERCOT balancing-energy bids from weekdays between September 2001 and January 2003, focusing on the 6:00–6:15 p.m. interval. They found significant differences in bidder performance; market stakes were an important determinant. The study covers a narrow, historical balancing market and does not estimate current Vistra economics. Hortaçsu & Puller (2008)
Potts and Castellanos model alternative ERCOT large-load scenarios through 2050. Under their assumptions, modeled portfolios rely on combinations of wind, solar and natural gas, and peak-load flexibility reduces modeled capacity requirements in some scenarios. The paper uses a simplified capacity-expansion model with perfect foresight; its results are not a forecast of Vistra's load or project returns. Potts & Castellanos (2025)
Prepare your assessment
- Assign a confidence level—low, moderate or high—to the claim that Vistra can translate expected large-load growth into durable cash generation. Defend the level with evidence available by October 3, 2026.
- Sort the major growth claims into: realized operating evidence; signed but future delivery; pending transaction or conditional commitment; and planned capacity or system-level forecast.
- Explain how the integrated retail/generation model and hedging could help, and what they cannot protect against.
- Identify two indicators you would monitor through 2028 to update your assessment. State what result would change your view.
Sources
- Vistra FY2025 Form 10-K
- Vistra Q2 2026 Form 10-Q
- Vistra Q2 2026 earnings release
- ERCOT 2025 long-term load forecast materials
- Academic citations are linked above. Method, source-access and transfer limits are described in the research dossier.