Vistra: scarce assets at a fair price

Vistra's second-quarter adjusted EBITDA rose 31% to $1.767 billion, guidance was reaffirmed, and the company says it expects to land at or above the midpoint of both its EBITDA and cash flow ranges. The shares still trade 36% below their 52-week high of $217.10. The question this analysis answers is whether that gap is an opportunity or just the market repricing the same assets more soberly.

Vistra is one of the largest competitive power generators in the country and also runs a large retail electricity business. That combination matters: retail earnings partly offset swings in wholesale power prices, and the generation fleet, which includes nuclear, gas, coal, solar and batteries, is where the scarcity value sits. Data-center demand has turned nuclear output into something buyers will now sign 20-year contracts for, and Vistra has signed two of them.

At $138.02 on September 28, 2026, the shares trade 4.1% below my triangulated Base value of $143.95, an upside of 4.3%. The Bear value of $96.77 is 29.9% below the price and the Bull value of $193.53 is 40.2% above it. The framework's required margin of safety for a Confidence B score with this much leverage and merchant exposure is 30%, which puts the maximum buy price at $100.76. The shares would need to fall about 27% to get there. Until then this is a hold: a good business at roughly a fair price.

Route: SF-01 Energy producer, stable free cash flow (integrated competitive power generation and retail). Economic classification: Quality Growth (9/12). Primary method: normalized adjusted free cash flow before growth (FCFbG) per share. Secondary method: EV/EBITDA. Values are present fair-value estimates, not 12-month price targets. All figures exclude the pending Cogentrix acquisition, in line with company guidance.

What Vistra is now

Second-quarter adjusted EBITDA of $1.767 billion came from four places: Retail $773 million, Texas generation $311 million, East generation $642 million and West generation $68 million, less $27 million of corporate costs. The East segment, which holds the PJM nuclear and gas plants, grew 54% from a year earlier and Texas more than doubled, helped by higher realized prices, higher PJM capacity revenue, the restart of Martin Lake Unit 1 and three months of contribution from the seven gas plants bought from Lotus in October 2025.

The contract story is the reason the stock commands a premium to a typical merchant generator. Meta signed 20-year agreements for 2,609 MW from Vistra's PJM nuclear plants: 2,176 MW of existing output, with deliveries starting in part in late 2026 and reaching full volume by the end of 2027, and 433 MW of uprates that will be built and delivered between 2031 and 2034. Amazon Web Services signed a 20-year agreement for 1,200 MW from Comanche Peak, with delivery beginning in the fourth quarter of 2027 and ramping to full capacity by 2032.

Two more pieces are not in 2026 guidance. The Cogentrix acquisition adds about 5,500 MW of gas plants across PJM, New England and ERCOT for a net price of about $4.0 billion, roughly 7.25x its expected 2027 EBITDA. FERC approved it in August and closing is expected in late 2026. And in the second quarter Vistra committed up to $1.0 billion to Helix Digital Infrastructure, a data-center venture with KKR, the Kuwait Investment Authority and NVIDIA, in which Vistra is the preferred power provider.

The latest quarter

The gap between GAAP and adjusted results is the standing feature of this business. Vistra hedges its power sales years ahead, and the accounting marks those hedges to market every quarter, so a rise in forward power prices books a paper loss now for cash that arrives later. The $472 million loss this quarter is that effect, not weaker operations.

Which numbers matter for Vistra

Because of those hedge marks, GAAP earnings are the wrong yardstick. The most reliable recurring measure is adjusted free cash flow before growth (FCFbG): operating cash flow, excluding working capital and margin deposits, less the capital needed to keep the plants running. The 2026 midpoint is $4.325 billion, which is 60% of the $7.2 billion EBITDA midpoint and matches management's stated medium-term goal of converting more than 60%. For comparison, 2025 delivered $3.592 billion of FCFbG and $5.912 billion of adjusted EBITDA, so the guidance midpoint implies growth of about 20% and 22%.

Two things need to come off that headline number before it is a per-share figure for common shareholders. First, FCFbG starts from operating cash flow, which already adds back non-cash stock compensation, and adjusted EBITDA adds it back explicitly. The 2026 guidance reconciliation puts that add-back at $137 million. It is a real cost of paying employees, so I deduct it. Second, FCFbG is calculated before the dividends on Vistra's preferred stock, which run at $47 million a quarter, or about $190 million a year, and belong to preferred holders, not common shareholders. After both, the $4.325 billion midpoint becomes about $4.0 billion, or $11.90 per share on 336 million shares, against the $12.87 you get from the raw figure.

Shares and capital structure

Vistra had about 336 million shares outstanding on August 3, 2026, down about 30% from November 2021 after roughly $6.5 billion of buybacks. About $1.2 billion of authorization remains, which management expects to use by the end of 2027. The Cogentrix deal will issue 5 million new shares to the seller at an agreed value of $185 each, well above today's price, so it dilutes the count by about 1.5%.

What sits ahead of the common shares is larger than the headline debt number suggests:

With a market value of about $46.4 billion, the enterprise value is about $68.9 billion, or 9.6 times the 2026 EBITDA midpoint. Net debt is 2.8 times the 2026 midpoint, but 3.05 times trailing twelve-month EBITDA of $6.58 billion. Management has said it is aiming for about 2.3 times by the end of 2027. Cogentrix will work against that in the short term: it brings about $2.3 billion of cash consideration and $1.5 billion of assumed debt for a stream of earnings that only starts contributing in 2027. Separately, Vistra priced $1.5 billion of junior subordinated notes on September 10, after the balance sheet date, so the debt figure above does not include them.

Primary valuation: normalized FCFbG per share

Vistra trades today at 11.6 times its guidance-midpoint FCFbG per share on this basis. A Base multiple of 12x is a small premium to that: enough to reflect the nuclear scarcity, the 20-year contracts, hedge visibility and buybacks, but capped by leverage and the fact that power markets are cyclical. The primary Base is $144.54.

Secondary valuation: EV/EBITDA

This method starts from the 2026 EBITDA range and subtracts the full $22.56 billion of claims from the table above. The secondary Base is $143.06. The 10x Base multiple is slightly above the 9.6x the market pays today, on the same logic as the primary method.

Historical multiples as context

Current market readings are 9.6x EV/EBITDA on the 2026 midpoint, 13.4x forward earnings and 11.6x FCFbG per share. A long, comparable history is not available: Energy Harbor, Lotus, the hedge book, the buybacks and soon Cogentrix have changed what the company is, so an old multiple describes a different business. I give this comparison no weight in the triangulation.

Valuation triangulation

The weighted average is $143.95 and I use it unrounded. The two methods differ by only 1.0%, far below the 20% warning threshold. That is partly by construction, since both start from the same guidance, so it shows the arithmetic is consistent rather than that the answer is precise.

What is the market already pricing in?

At $138.02 and the 12x Base multiple, the market is paying for $11.50 of FCFbG per share, or about $3.87 billion after stock compensation and preferred dividends. Add those back and that is about $4.19 billion of headline FCFbG, roughly 3% below the $4.325 billion guidance midpoint and comfortably above the $3.925 billion low end. On EV/EBITDA, the current enterprise value at a 10x multiple implies about $7.03 billion of adjusted EBITDA, about 2% below the midpoint. So the price does not assume anything heroic for 2026. What it cannot see is what happens after the hedges roll: only 72% of 2028 generation is hedged, and the 2027 figure is an opportunity estimate, not guidance.

Bear, base and bull scenarios

The scenarios come from the two methods above, weighted 60% and 40%; I assign no probabilities to them. The downside case takes away about $41 a share and the upside case adds about $56, a reasonable payoff but not one that meets the margin of safety the framework requires.

At today's price the shares are fairly valued at roughly $11.50 per share and 12x, $9.85 and 14x, or $13.80 and 10x. They only look clearly cheap if cash flow per share rises toward $13.75 at 12x (about $165) or the multiple reaches 14x on $12 (about $168).

How the buy price is set

For a Confidence B score, the framework's base margin of safety is 25% to 35%. I use 30%: leverage, merchant cyclicality and Cogentrix integration argue for more, while 94% hedging of 2027 and the 20-year contracts argue for less. Applied to the $143.95 Base, the maximum buy price is $100.76. That is only about 4% above the Bear value of $96.77, so buying there would mean paying close to what the pessimistic scenario says the shares are worth. The shares would need to fall about 27% from $138.02.

Dividend

The common dividend is $0.23 a quarter, or $0.92 a year, a yield of 0.67%. It is not the thesis. In the first half of 2026 Vistra paid $154 million in common dividends and $96 million in preferred dividends, against $709 million spent on buybacks. Total return to shareholders depends on per-share value creation from buybacks and growth spending, not income.

Reasons to own VST

• Scarce dispatchable nuclear output. The PJM and Comanche Peak plants are hard to replicate and increasingly valuable to large power buyers.
• Longer contract duration. 3,809 MW of nuclear output is contracted for 20 years to Meta and Amazon, which replaces some merchant exposure with revenue tied to named customers.
• Demand runway. Data centers and electrification support a multi-year power runway, and Vistra has said it sees room to contract more nuclear capacity.
• Buybacks amplify per-share results. The share count is down about 30% since late 2021 and the program continues.
• Integration reduces volatility. Retail earnings, generation and hedging together make results more resilient than a pure merchant generator. Hedge coverage is about 100% for 2026 and 94% for 2027.

What could go wrong?

• Thesis breaker: FCFbG settles below about $3.5 billion to $3.8 billion after hedges roll. Today's cash flow was locked in when power prices were high. If sustainable owner earnings are structurally lower, the valuation multiple has nothing to stand on.
• Leverage and Cogentrix. About $3.8 billion of cash and assumed debt goes out for earnings that start in 2027. Net debt is 2.8x the 2026 midpoint but above 3x on trailing EBITDA.
• Growth spending sits outside FCFbG. The $1.0 billion Helix commitment, two new Permian gas units, solar projects and the Meta uprates (with capital spending running through 2034) all come on top of the free cash flow figure.
• Commodity and capacity-market volatility. Wholesale prices and capacity auction results drive a large part of earnings.
• Operational and regulatory risk. Nuclear outages, licensing, environmental rules and market-design changes can each affect earnings, as the Martin Lake and Moss Landing incidents showed.

Management execution

Management beat its original 2025 guidance midpoints by $112 million on adjusted EBITDA and $292 million on FCFbG, reaffirmed 2026 guidance after both the first and second quarters, closed the Lotus acquisition, signed the Meta and Amazon contracts and kept buying back stock. CEO Jim Burke bought 2,000 shares at $135 on August 24, 2026, a $270,000 purchase. The key test ahead is per-share value creation from Lotus, Cogentrix and the new growth investments, not just their headline size. Assessment: 16/20, good. The next test is whether Cogentrix and Helix add value per share rather than just megawatts.

Stock Analyza scorecard

Acquisition-driven growth, capital intensity, leverage and merchant exposure make a classic compounder model too precise for this business, so the engine is a growth-adjusted multiple. Business quality is 78/100: moat and assets 17/20, return versus cost of capital 15/20, balance sheet 10/15, cash flow quality 12/15, growth runway 13/15, management 8/10, governance and dilution 3/5. Nuclear scarcity is the strongest moat-like feature, and the balance sheet and preferred stock are what hold the score back.

Confidence score

Filings are detailed and guidance has been reliable. Accounting transparency is marked down because hedge mark-to-market swings make GAAP earnings hard to read, and a comparable long history of the FCFbG multiple is not available.

What I would watch from here

Net leverage is 2.8x on the 2026 midpoint, which is green, but only just, and Cogentrix will push it up before it comes back down. Review with the third-quarter report, expected around November 5, the Cogentrix close and integration, and any update to the 2027 outlook.

The three most important thesis breakers

1. Adjusted FCFbG settling structurally below about $3.5 billion to $3.8 billion once today's hedges roll off.
2. Leverage above 3.5x without a credible path to deleveraging.
3. A major nuclear contract impairment or cancellation.

Adversarial review

The first draft's arithmetic on FCFbG per share, the guidance figures, the second-quarter results, the contract terms, the buyback and share count data and the CEO purchase all checked out against Vistra's filings and press releases. Several things did not, and correcting them lowered the Base value from $158 to $143.95.

The strongest counter-thesis remains that current FCFbG may overstate long-run owner earnings if favorable hedges and tight power markets normalize while debt and capital intensity stay high. The multiples themselves are judgment calls, the Bear, Base and Bull cash flow and share inputs are analytical estimates rather than company figures, the preferred stock value comes from third-party capital structure data, and I have not verified how the September notes will be used. That is why $138.02 is not treated as a high-margin-of-safety entry.

Checksum: ($4.325B - $0.137B - $0.190B) / 336M = $11.90; $138.02 / $11.90 = 11.6x; $143.95 x 70% = $100.76.

Final verdict: hold at $138.02, watchlist for new money below $100.76

Vistra has a strong operating story: scarce nuclear assets, long-duration contracts with two of the largest technology buyers of power, high hedge visibility and aggressive buybacks. Each of those is real. But the price already gives credit for most of them, the balance sheet carries about $22.6 billion of claims ahead of common shareholders, and the cash flow being valued was locked in at higher power prices than may persist.

Verdict: WATCHLIST at $138.02. Triangulated Base value $143.95, upside 4.3%, Confidence B (20/25). Required margin of safety 30%, maximum buy price $100.76.

The case improves if the third-quarter report confirms guidance and the 2027 outlook, if Cogentrix closes without stretching leverage, or if the price falls into the buy zone with the fundamentals intact.

Sources

Primary sources include Vistra's second-quarter 2026 earnings release (August 7, 2026) and Form 10-Q for the quarter ended June 30, 2026, its fourth-quarter and full-year 2025 earnings release, its first-quarter 2026 Form 10-Q, its Form 10-K for 2025, the January 2026 announcement of the Cogentrix acquisition, its 8-K on the Meta agreements, and the CEO's Form 4 filing of August 2026. Management's leverage target is from the fourth-quarter 2025 earnings call.

The reference market price is $138.02, the September 28, 2026 close from StockAnalysis.com. The preferred stock amount is from S&P Global capital structure data as shown by Seeking Alpha. Balance sheet figures are as of June 30, 2026 and share data as of August 3, 2026. Normalized FCFbG, EBITDA, scenario inputs, multiples and the value estimates are analytical estimates, not company guidance.