David Hay — Friday POW!: Vistra Corp (VST)

The frame — AI is an energy story, and the rest is handwringing

• "What we believe is (or will become) clear is that, at its core, the AI story is fundamentally an energy story. Who will win? How will it be (effectively) regulated? Will it massively displace labor? All these questions and handwringing are secondary to the fact that AI is going to require an enormous amount of reliable power to sustain its exponential growth trajectory."
• The hedge is built into the same sentence — that trajectory is "likely to plateau in years to come." The call is on the power requirement, not on perpetual exponential growth.

The asset — 40GW, two grids, the second-largest competitive nuclear fleet

• "VST is one of the largest competitive power generators in the United States, operating approximately 40,000 megawatts (40 gigawatts) across nuclear, natural gas, solar, and battery storage in the ERCOT and PJM grids, with a retail business serving approximately five million customers."
• The scale translation: "equivalent to around 40 large-scale nuclear power plants, a truly immense power-generation base," and it runs "the nation's second-largest competitive nuclear fleet."
• The definitions are spelled out for the reader — "PJM = the Pennsylvania, New Jersey, Maryland interconnection, the largest power grid in the U.S. ERCOT = the Electricity Reliability Council of Texas, the grid operator for 90% of the state."

The price — 36% off the high on a $58B cap

• "Its stock is presently trading at approximately $145, down 36% from its 52-week high of $219.82"; the 52-week range is $132.66-$219.82 and the market cap ~$58B.
• Earnings context, from the technical section: "largely an earnings-driven moonshot — earnings have rocketed from just over $1.00/share in 2022 to an estimate of $9.13 this year."

The quarter — EBITDA +30%, generation EBITDA +68%

• Q2 2026 adjusted EBITDA $1.767B, +30% YoY, above the $1.64B consensus; generation EBITDA $994M, +68% YoY; YTD adjusted EBITDA $3.261B vs $2.589B in 2025.
• The operating tell: the fleet "maintained 97% commercial availability during record summer heat in both ERCOT and PJM" — availability is what a power producer is paid for when the grid is stressed.

The revenue "miss" — real in GAAP, irrelevant in economics

• "With a Q2 revenue miss of $4.02 billion versus $5.46 billion consensus, VST stock is predictably under pressure at the moment. It is real in the GAAP sense and irrelevant in the economic sense, but Vistra's revenue figure includes unrealized mark-to-market losses on forward hedges that will settle favorably in future periods."
• The elegant part of the argument: "the hedge program creating the apparent miss is the same driver that gives the company near-complete 2026 earnings certainty" — ~100% of 2026 generation price hedged. The cost of the optics is the source of the certainty.
• "The generation EBITDA of $994 million growing 68% year-over-year, on a fleet that maintained 97% commercial availability during record summer heat… is the operational reality."

The demand proof — a meter reading, not a projection

• "PJM and ERCOT set all-time peak loads of, respectively, 168 and 91 gigawatts in July."
• Why that specific evidence matters: "That is the structural AI demand proof the thesis requires, visible in real-time load data rather than forward projections." The whole data-center-demand debate is settled here with observed load rather than a consultant's forecast.

Helix — "more than just another power purchase agreement"

• "The recently announced Helix Platform is, in our view, the stock's long-term game-changer. Vistra is committing up to $1 billion to a strategic infrastructure partnership with KKR, NVIDIA, and the Kuwait Investment Authority (KIA) to serve as the preferred power partner for at-scale data-center development."
• The division of labour: "NVIDIA provides compute expertise, KKR provides capital and deal sourcing, and KIA (not to be confused with the Korean manufacturer of famously reliable automobiles) provides a long-duration investment mandate. Vistra supplies the one thing none of them can build quickly: 24/7 dispatchable, carbon-friendly nuclear and gas generation in the markets where AI data centers are being built."
• And the valuation consequence: "None of the revenue from Helix is in any current guidance, and the partnership is structured to generate additional contracted capacity commitments as it deploys capital."

The contract book — Meta, AWS, and a licence to 2046

• AWS PPA ~3,800MW at Comanche Peak; Meta ~2,600MW of PJM nuclear; Perry Nuclear licence extended to 2046 — the hyperscaler relationships Haymaker later cites as part of Vistra's incumbency advantage.
• The Meta uplift is timing-gated rather than speculative: "the Meta PPA uplift begins flowing through as contracts reach effective dates."

What is / is not in the numbers — the guidance gap

• Guidance reaffirmed: FY2026 adjusted EBITDA $6.8-7.6B and adjusted FCF before growth $3.925-4.725B, with "management expects at or above midpoint."
• What it leaves out, stated explicitly: "Cogentrix acquisition ($4.7B, ~5,500MW, pending H2 2026 close); Meta nuclear PPAs (~2,600MW PJM); Helix Platform; nuclear production tax credits."
• The arithmetic that follows: "the 2026 guidance midpoint of $7.2 billion… is the starting point, not the destination. Cogentrix at $4.7 billion likely adds $470 to $590 million in annual EBITDA at close… The 2028 EBITDA trajectory incorporating all three almost certainly exceeds $9 billion annualized. At a $58 billion cap, the market is pricing none of it."

2027 — the soft spot, and the hedge book that covers it

• "The 2027 EBITDA opportunity of $7.4 to $7.8 billion, trending toward the lower end due to ERCOT forward curve softness, also excludes Cogentrix and Meta."
• The offset: "hedge coverage of 94% in 2027 and 72% in 2028 provides meaningful protection against that ERCOT headwind, while unhedged upside remains if power prices recover."

Pushing back on the WSJ — conceding the premise, moving the moat

• The concession: the "Heard on the Street" author "is right that the easy version of the power thesis has weakened to a degree. Specifically, that regulators in ERCOT and PJM are trying to accelerate new generation and prevent existing plant owners from capturing unlimited scarcity rents as data-center demand rises."
• The rebuttal is physical, not rhetorical: "the bottleneck is more than just policy, it is the physical ability to deliver reliable megawatts. New gas plants, nuclear capacity, transmission, transformers, and interconnections still take years to build. Vistra already owns scarce generation, existing grid connections, nuclear and gas assets, and relationships with hyperscalers."
• The reformulation to keep: the market is shifting "from 'own existing power and enjoy higher prices' to 'own existing power and be one of the few companies capable of expanding reliable supply'… we think the author underestimates Vistra's long-term position."
• And the regulation cuts both ways: it "may cap some merchant-market upside, especially in ERCOT, but it also creates opportunities for direct contracts, uprates, brownfield expansions, and new capacity at existing sites. Vistra's Meta, AWS, Cogentrix, and Helix initiatives are already moving it in that direction."
• The summary line: "the thesis is evolving, not breaking: the next phase of the power trade is less about passive scarcity rents and more about who can actually deliver incremental dispatchable power."

Valuation — a discount to Constellation on numbers that exclude the catalysts

• "At approximately $141 and 10.6x 2026 adjusted EBITDA guidance midpoint on an enterprise value basis, Vistra trades at a meaningful discount to Constellation Energy at approximately 14.5 times earnings (P/E)" — restated at the close as 13.6× trailing EV/EBITDA for CEG.
• The double-count point: "this is despite Vistra's multiple being calculated on forward guidance that explicitly excludes Cogentrix, the Meta PPAs, and Helix."

The entry metric — FCF yield on a normalised base

• "FCF yield is the most compelling entry-point metric: 2026 adjusted FCF guidance midpoint of $4.325 billion against a market cap of approximately $47.5 billion is approximately 9.1%."
• The forward version: "management's medium-term target of converting over 60% of adjusted EBITDA to FCF, applied to a 2028 EBITDA base approaching $9 billion as the excluded catalysts begin contributing, implies normalized annual FCF of approximately $5.4 billion. That would represent an 11.4% yield on current market cap before any re-rating."
• Sell-side: "consensus across 20 analysts is Buy with an average price target of approximately $217, implying roughly 54% upside."

Arguing the other side — three legs, and they all have to break

• "The bear case has three components, all credible to some degree." (1) "ERCOT forward curve softness is compressing unhedged 2027 and 2028 generation margins as renewable additions outpace Texas' near-term demand growth." (2) "Helix's $1 billion commitment is capital subject to milestones and execution risk in an unfamiliar infrastructure-development business." (3) "PJM regulatory uncertainty around the IRAS (Interim Resource Adequacy Service) capacity market framework could reduce capacity revenue that has been a significant earnings tailwind."
• The test: "we believe the bear case is justifiable only if all three occur simultaneously. At 10.6x EV/EBITDA on guidance that excludes Cogentrix, Meta, Helix, and nuclear-production tax credits, with a fully hedged 2026 book and the most constrained power-generation asset class in the U.S. at the center of the AI infrastructure buildout (the PJM region), we think that threshold is too high."

The build-out risk — answered with a slowdown, not a denial

• "To many, the big risk with this type of name is that the great data center build-out hits the wall. It might be prudent to say that we think it could slow meaningfully, but it is almost certain to continue at a rapid clip, just not as fast as is now projected."
• The mechanism keeping it going: "the hyperscalers view not keeping pace with their peers as an existential risk" — a competitive-dynamics argument, not a return-on-capital one.

The technical set-up — a stop instead of a confirmation

• "Clearly, VST has been range-trading for the last two years. Its big breakout was in 2023, prior to an extremely explosive run… largely an earnings-driven moonshot."
• The caveat, stated against his own pick: "it is somewhat concerning that VST has been trading below the 200-moving average for over a year. This isn't ideal, but it has been stabilizing in the 150 range for most of 2026."
• The discipline offered in place of confirmation: "to be on the safe side, you could put in a stop at 130, right below short-term support" — note the contrast with Aug-7's GILD, bought resting on its 200-day.

The bottom line — and the verdict

• "Investors in VST today get the second-largest competitive nuclear fleet in the United States, a company that delivered 30% adjusted EBITDA growth in Q2 on a fully hedged book, maintained 97% fleet availability during record summer heat, and generated $3.261 billion in adjusted EBITDA in the first half of 2026 alone."
• The mispricing restated: "a stock trading at a 36% discount to its 52-week high, where the primary narrative overhang is a GAAP revenue line distorted by hedge accounting rather than any deterioration in the operational business."
• "The case rests on a pretty straightforward gap: Vistra at 10.6x forward EV/EBITDA on conservative guidance versus Constellation Energy at 13.6x trailing EV/EBITDA, with VST's multiple calculated before Cogentrix, the Meta PPAs, and Helix contribute a single dollar."
• Verdict: "We are buyers of VST at today's price (notwithstanding the aforementioned technical caveats)."

Portfolio tables — images, with two text notes

• "(Highlighted cells indicate today's updates.)" The Buys, Holds / Trims and Sells tables were published as images and are not transcribed.
• Buys note: "DVN added, along with earlier WDOFF recommendation, for visibility." Sells note: "we've revised the overall % Change formula to more directly indicate simple price movement" — a scorecard methodology change, not a stance.