Vistra Q2 2026 Earnings: EBITDA Up 31%, Revenue Missed, and the Meta Deal Is Still Not in the Guide

---

The EBITDA Beat, the Top Line Fell, and Meta Is Still Not in the Numbers

UPDATE (August 7, 2026): Vistra has reported, and the scoring note at the bottom of this page was the whole value of it. Ongoing Operations Adjusted EBITDA came in at $1,767 million, up 31% from $1,350 million in Q2 2025 and above the roughly $1.64 billion expected. On the lines this page told readers to ignore, it was ugly: revenue of $4,017 million, down from $4,250 million a year ago and nowhere near the $5.8 billion consensus quoted below, with net income of $305 million against $327 million and GAAP EPS of $1.70 against the $2.43 and $2.06 consensus quotes flagged below as unreliable.

Revenue falling while adjusted EBITDA rises 31% is what hedge mark-to-market does to a merchant generator's reported top line, and it is the reason the grading instruction below was to measure the quarter against the EBITDA track rather than the per-share line. Read the other way round, this print looks like a collapse. It was not one.

Guidance was reaffirmed, not raised: full-year 2026 Ongoing Operations Adjusted EBITDA of $6.8 to $7.6 billion and Ongoing Operations Adjusted Free Cash Flow before Growth of $3.925 to $4.725 billion. Management also put a 2027 Ongoing Operations Adjusted EBITDA midpoint opportunity range of $7.4 to $7.8 billion on the table.

The catalyst this page was waiting for did not fire. Both ranges still explicitly exclude the Meta power purchase agreements and the pending Cogentrix acquisition. That is the second quarter running in which the largest contract in company history has been kept out of the numbers. On Cogentrix there was real movement: FERC approval is in, the deal covers about 5,500 MW across ten gas plants, and it is expected to close in the second half of 2026. President and CEO Jim Burke said the team "delivered another strong quarter, building on our momentum from the start of the year and continuing to execute at a high level."

The stock was down about 1.2% in pre-market trading. That is a pre-market figure, not a close, and a verified regular-session quote was not sourceable at the time of writing. The 10:00am ET call is under way as this update publishes, so a dollarised Meta number could still come from the Q&A rather than the release: the conditional graded below stays open until the transcript is out. The other two results from the same morning: Take-Two and Oklo.

---

TL;DR

• Vistra reports Q2 2026 results Friday, August 7, before the open, with the call at 10:00am ET, ninety minutes after the July jobs report hits the same tape.
• The headline expectation is violent growth: consensus EPS quotes run up to $2.43, up about 140% from $1.01 a year ago (a second provider sits at $2.06; treat the precise number loosely), on revenue near $5.8 billion, up roughly 36% from $4.25 billion.
• The structural story is what is not in the numbers: the 20-year Meta agreements for 2,609 MW of nuclear capacity (Perry, Davis-Besse, plus uprates), signed in January, are explicitly excluded from the reaffirmed $6.8-7.6 billion adjusted EBITDA guidance, with contributions starting in 2027.
• The stock has de-rated with the whole AI-power complex on PJM price-cap proposals: it closed Monday near $148.6, about 32% below its 52-week high, slightly down for the year.
• One scoring note: Vistra's GAAP EPS is noisy (hedge mark-to-market), so grade Friday on adjusted EBITDA against the $6.8-7.6 billion full-year track, not the EPS line.

When Does Vistra Report Earnings?

Friday August 7, before the market opens, with the call at 10:00am ET. It reports into the most macro-exposed slot of the week, ninety minutes after nonfarm payrolls, alongside Oklo, the other electricity name on the day.

The Board

The guidance excludes the biggest contract the company has ever signed. That gap is the bull case.

The Guidance That Ignores Its Own Best News

In January, Vistra signed 20-year power purchase agreements with Meta covering 2,609 MW of nuclear capacity: 1,268 MW at Perry, 908 MW at Davis-Besse, and 433 MW of uprates across three plants, with deliveries beginning late 2026 and full delivery by end-2027. In May, management reaffirmed full-year adjusted EBITDA guidance of $6.8-7.6 billion and explicitly excluded both the Meta PPAs and the pending Cogentrix acquisition from it.

That is the whole setup in two sentences: the market is valuing Vistra on a guidance track that leaves out its largest-ever contracted revenue stream. Friday's question is whether management starts quantifying the 2027-plus uplift, raises the range to absorb early Meta economics, or keeps the powder dry another quarter. Any dollarised Meta disclosure is the catalyst; the Q2 print itself is secondary.

Why the Stock Is Down Anyway

Because regulation moved faster than contracts. Proposals to cap electricity prices in PJM, the grid where Vistra's nuclear fleet and the Meta deal live, have pressured every merchant generator, and the AI-power trade broadly has deflated on "when does the capex become profit" doubts. Vistra trades about 32% below its 52-week high with the group. The bear case is simple: price caps compress exactly the scarcity premium that 20-year hyperscaler PPAs monetise. Management's regulatory commentary on Friday carries real information for the whole complex (Constellation, Talen, NRG), not just VST.

How to Grade Friday

• Adjusted EBITDA versus the $6.8-7.6 billion track. The cleaner scoreboard; GAAP EPS swings on hedge mark-to-market and the +140% consensus figure inherits that noise.
• Any Meta quantification. A 2027 EBITDA contribution number, even a range, converts the deal from narrative to model input.
• PJM commentary. The regulatory line every power investor will trade on.
• Cogentrix close timing, the other excluded item.

The Options Angle

A single provider quotes the implied move near 6.4%, which we flag as thinly sourced. The macro overlay is the real reason to stay small: this print reacts to payrolls first and fundamentals second for the opening hour, and the week-ahead playbook's warning about selling premium into a payrolls Friday applies squarely.

• No pre-print options position. A thin implied quote plus a macro gap risk that has nothing to do with the company is two reasons, either sufficient.
• The conditional: any dollarised Meta/2027 disclosure with guidance intact is the entry, via shares or calls once the payrolls dust settles.

Trade log

#	Stance	Structure	Strikes and expiry	Cost or credit	Spot at writing	Implied move	Breakeven
1	Pass	Any pre-print options position	Aug expiries	Not sourced	~$148.6, Aug 3 close	~6.4%, single source	n/a; pass scored against the realised move
2	Conditional	Post-print long (shares or 1-2 month calls) if management quantifies Meta PPA economics with guidance intact	Struck off the Aug 7 post-open	Struck off the Aug 7 post-open	To be struck Aug 7	n/a	Scored against the post-print entry if triggered

The Options Angle, Now Graded

Scored against a pre-market move of about 1.2% on a thinly sourced implied of roughly 6.4%. Both grades are provisional: the regular-session close is the number that settles row 1, and the 10:00am call settles row 2.

#	Play as logged	Condition	Result
1	Pass on any pre-print options position, either direction	none	Win on the long-vol side, forfeited credit on the short side. A 6.4% implied against a roughly 1.2% early move is a premium seller's print, and this page took neither leg.
2	Conditional post-print long, shares or 1-2 month calls	Meta PPA economics quantified and guidance intact	Not triggered on the release. Guidance held; Meta stayed excluded, from the 2026 range and the new 2027 range alike. Open until the call transcript.

The stated reason for row 1 was that this print reacts to payrolls before fundamentals for the opening hour, and that a single-provider implied quote is not something to trade against. That reasoning survives the outcome even though the payrolls gap did not dominate the tape: a pass justified by not being able to price the thing is not retroactively wrong because the thing turned out cheap.

Row 2 is the one that matters, and it is now a pattern rather than a one-off. This page argued that the market is valuing Vistra on a guidance track that leaves out its largest-ever contracted revenue stream, and that the moment management dollarises it, the deal converts from narrative to model input. Management has now had two consecutive quarters to do that and has instead published a fresh 2027 range that also excludes it. That is a deliberate choice, and the most defensible read is that the Meta economics are not yet certain enough in timing to guide on, with deliveries only beginning late 2026. The bull case below is unchanged in substance and has simply lost a quarter of time value.

What did change is Cogentrix. FERC approval plus a second-half 2026 close turns the other excluded item into something with a date on it, and a closed acquisition is far harder to leave out of a guidance range than a contract that has not started delivering.

The One-Line Read

Vistra reports a quarter expected to more than double its earnings while its own guidance still pretends the largest contract in company history does not exist, and the stock, parked a third below its high on regulatory fear, moves on whichever of those two silences breaks first.