Built by Grok 4.6 High

VST | Earnings Review

Verdict: earnings-power accelerating; GAAP revenue decelerating. Q2’26 is the textbook merchant-power print: consolidated operating revenues of $4,017M fell −5.5% YoY and missed FMP consensus by −26.5%, while Ongoing Ops Adj. EBITDA rose +31.0% YoY to $1,767M (vs $1,349M), margin expanded ~1,225 bps to ~44%, and generation volumes re-accelerated +10.1% to 50.2 TWh. Management framed the quarter exclusively as an operating success and did not bridge the GAAP miss.

Key metrics. Generation was the engine — East Adj. EBITDA $642M (+53.6% YoY), Texas $311M (+119%), Retail $773M (+2.2%). Call color: Gen Adj. EBITDA ~$994M vs ~$593M in Q2’25 on ~5% higher realized $/MWh, higher PJM capacity, Lotus (3Q25 close), and Martin Lake Unit 1 restart. Commercial availability >97% through July heat (PJM peak >168 GW; ERCOT >91 GW).

Guidance. Full reaffirm of FY2026 Ongoing Adj. EBITDA $6.8–7.6B (mid $7.2B, +21.8% vs FY25 actual $5,912M) and Adj. FCF before Growth $3.925–4.725B, with explicit “at or above midpoint” bias after H1 at ~$3.26B (45.3% of mid). 2027 midpoint opportunity $7.4–7.8B maintained but Q&A biases lower end of the ex-deal band; Cogentrix + Meta still excluded and still worth ~+$700M to midpoint once layered in.

Tone. 2026 confidence up (execution + ≥-mid language); 2027 bifurcated (printed range held, internal bias lower end until deals). Policy/market structure expansive (ERCOT Batch Zero / Abbott audit, FERC co-location, IRAS carrot-vs-stick). Deflections: 2028 regional hedges, Helix MW pipeline, Batch Zero project list beyond Comanche Peak.

Contradictions (2). (1) High: ERCOT load CAGR floor quietly 5–6% → 4–6% while claiming forecast unchanged. (2) Medium: prepared remarks “maintain 2027 range due to offsets” vs Q&A “offsets don’t fully offset → lower end.”

Upcoming catalysts. Next leg is not another GAAP beat — it is Cogentrix close (mid–late 2026), Meta nuclear PPA deliveries (late 2026 → full YE2027), Helix first projects, ERCOT audit / Batch Zero resolution, and PJM co-location tariff filing. Formal 2026/2027 guide update expected on Q3 call (or first call after Cogentrix close).

Primary KPI for Vistra is Ongoing Ops Adj. EBITDA, not GAAP revenue/EPS (hedge MTM and retail pass-throughs dominate the IS). YoY always same quarter prior year.

Consolidated P&L trajectory

| Metric | 2023Q3 | 2023Q4 | 2024Q1 | 2024Q2 | 2024Q3 | 2024Q4 | 2025Q1 | 2025Q2 | 2025Q3 | 2025Q4 | 2026Q1 | 2026Q2 | |---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:| | Operating revenues ($M) | $4,086 | $3,079 | $3,054 | $3,845 | $6,288 | $4,037 | $3,933 | $4,250 | $4,971 | $4,584 | $5,640 | $4,017 | | Revenue YoY % | — | — | −31.0% | +20.6% | +53.9% | +31.1% | +28.8% | +10.5% | −20.9% | +13.5% | +43.4% | −5.5% | | Ongoing Ops Adj. EBITDA ($M) | $1,613 | $965 | $813 | $1,414 | $1,444 | $1,985 | $1,240 | $1,349 | $1,581 | $1,742 | $1,494 | $1,767 | | Adj. EBITDA YoY % | — | — | +46.8% | +40.3% | −10.5% | +105.7% | +52.5% | −4.6% | +9.5% | −12.2% | +20.5% | +31.0% | | Adj. EBITDA margin % | 39.5% | 31.3% | 26.6% | 36.8% | 23.0% | 49.2% | 31.5% | 31.7% | 31.8% | 38.0% | 26.5% | 44.0% | | Gen volume Ong. ops (TWh) | 51.3 | 41.6 | 40.0 | 46.4 | 59.4 | 50.5 | 48.0 | 45.6 | 58.1 | 56.5 | 50.5 | 50.2 | | Gen volume YoY % | — | — | +10.5% | +19.3% | +15.8% | +21.4% | +20.0% | −1.7% | −2.2% | +11.9% | +5.2% | +10.1% | | GAAP diluted EPS ($) | $1.25 | −$0.55 | −$0.24 | $0.90 | $5.25 | $1.09 | −$0.93 | $0.81 | $1.75 | $0.55 | $2.87 | $0.76§ |

§Q2’26 diluted EPS from FMP (Daloopa series not yet loaded for 2026Q2).

Segment Adj. EBITDA (drivers)

| Metric | 2025Q1 | 2025Q2 | 2025Q3 | 2025Q4 | 2026Q1 | 2026Q2 | YoY (Q2) | |---|---:|---:|---:|---:|---:|---:|---:| | Retail Adj. EBITDA ($M) | $184 | $756 | $37 | $645 | $68 | $773 | +2.2% | | Texas Adj. EBITDA ($M) | $490 | $142 | $784 | $418 | $586 | $311 | +119% | | East Adj. EBITDA ($M) | $514 | $418 | $719 | $631 | $801 | $642 | +53.6% | | West Adj. EBITDA ($M) | $62 | $49 | $63 | $70 | $56 | $68 | +38.8% |

Revenue vs Adj. EBITDA — absolute levels ($M)

YoY growth trajectory (divergence is the story)

| Quarter | Revenue YoY | Gen vol YoY | Adj. EBITDA YoY | Adj. EBITDA margin YoY bps | |---|---:|---:|---:|---:| | 2024Q2 | +20.6% | +19.3% | +40.3% | +517 | | 2024Q3 | +53.9% | +15.8% | −10.5% | −1,651 | | 2024Q4 | +31.1% | +21.4% | +105.7% | +1,783 | | 2025Q1 | +28.8% | +20.0% | +52.5% | +491 | | 2025Q2 | +10.5% | −1.7% | −4.6% | −503 | | 2025Q3 | −20.9% | −2.2% | +9.5% | +884 | | 2025Q4 | +13.5% | +11.9% | −12.2% | −1,117 | | 2026Q1 | +43.4% | +5.2% | +20.5% | −504 | | 2026Q2 | −5.5% | +10.1% | +31.0% | +1,225 |

Annual context (FY2021–FY2025)

| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 | |---|---:|---:|---:|---:|---:| | Operating revenues ($M) | $12,077 | $13,728 | $14,779 | $17,224 | $17,738 | | Ongoing Ops Adj. EBITDA ($M) | $1,941 | $3,115 | $4,140 | $5,656 | $5,912 | | Adj. EBITDA YoY % | — | +60.5% | +32.9% | +36.6% | +4.5% | | Adj. FCF before growth ($M) | $179 | $2,399 | $2,491 | $2,888 | $3,592 |

Read: FY25 Adj. EBITDA growth decelerated sharply to +4.5% (Energy Harbor anniversary + hedge normalization). 2026 H1 has re-accelerated — Q1 +20.5%, Q2 +31.0% — underwriting the FY26 mid guide of +21.8% vs FY25 without a formal range raise. Ignore GAAP sales noise; watch Adj. EBITDA, volumes, and East/Texas generation.

Headline: GAAP double miss, operating beat. Street Adj. EBITDA consensus unavailable this run (VA/BBG disconnected). FMP GAAP is a poor map of how the sell-side scores Vistra.

This quarter vs consensus

| Metric | Consensus | Actual | Variance | Beat/Miss | |---|---|---|---|---| | Revenue (GAAP) | $5.464B (FMP) | $4.017B | −$1.447B / −26.5% | Miss | | Diluted EPS (GAAP) | $1.61 (FMP) | $0.76 (FMP) | −$0.85 / −52.8% | Miss | | Ongoing Ops Adj. EBITDA | n/a (no street print) | $1,767M | +$418M / +31.0% YoY vs Q2’25 | Operating beat |

Historical beat/miss heatmap (FMP GAAP consensus; Daloopa actuals)

| Window | Revenue beat rate | EPS beat rate | Both-metric beat rate | |---|---|---|---| | L7 (proxy for L12) | 2/7 = 28.6% | 2/7 = 28.6% | 2/7 = 28.6% | | L4 | 1/4 = 25.0% | 1/4 = 25.0% | 1/4 = 25.0% |

Pattern: Mixed / frequent GAAP misser; operating path improving. Absolute revenue variance is deteriorating (L4 avg ~18.5%; Q2’26 largest abs miss at 26.5%). Management variance narrative was entirely Adj. EBITDA + hedges/PJM/Lotus/Martin Lake — no GAAP bridge. Guidance hit-rate remains the better quality signal than GAAP EPS beats. Without VA/BBG Adj. consensus, GAAP beat rates understate how the street scores Vistra.

Management explanation (call)

1. Adj. EBITDA: “nearly $1.8 billion… over 30% increase YoY” (Burke); Moldovan quantified $1.767B.
2. Generation drivers: higher realized prices (~+5% $/MWh), PJM capacity, flexible gas optimization, Martin Lake Unit 1, Lotus assets.
3. Retail: seasonal strength; ~$773M vs ~$756M.
4. Guide: reaffirm + “at or above midpoint” given H1.

Official ranges (reaffirmed)

| Metric | Low | High | Mid | Action vs Q1’26 | |---|---:|---:|---:|---| | FY2026 Ongoing Adj. EBITDA | $6,800M | $7,600M | $7,200M | Reaffirmed + ≥ mid bias | | FY2026 Adj. FCF before Growth | $3,925M | $4,725M | $4,325M | Reaffirmed | | FY2027 Adj. EBITDA midpoint opportunity | $7,400M | $7,800M | $7,600M | Maintained; lower-end bias ex deals | | Revenue / EPS / margin % | — | — | — | Not guided |

Exclusions (explicit): Cogentrix contribution; Meta nuclear PPA premium above market. Stated deal add to 2027: ~+$700M to midpoint opportunity.

Waterfall — hold-and-bias, not raise/cut

Implied growth path

| Bridge | Level | Growth | |---|---:|---:| | FY25 actual → FY26 mid guide | $5,912M → $7,200M | +21.8% YoY | | FY26 mid → FY27 mid opp (ex deals) | $7,200M → $7,600M | +5.6% | | FY26 mid → FY27 mid + ~$700M deals | $7,200M → ~$8,300M | ~+15.3% | | H1’26 vs FY mid | $3,261M / $7,200M | 45.3% of mid |

Guidance history / credibility

| As-of | Guide low | Guide high | Notes | |---|---:|---:|---| | 2025Q1 | $5,500 | $6,100 | FY2025 range | | 2025Q3 | $5,700 | $5,900 | Narrowed; introduced FY26 $6.8–7.6B | | 2026Q1 | $6,800 | $7,600 | Reaffirm | | 2026Q2 | $6,800 | $7,600 | Reaffirm + ≥ mid bias |

FY2025 actual $5,912M beat high end of the final narrowed $5.7–5.9B guide. Pattern: conservative annual framing, mid-year narrow rather than large raises. Correction: any prior internal note citing FY26 “$6.0–6.6B” is obsolete — authoritative is $6.8–7.6B since Q3’25.

Tone scorecard (Q2 vs Q1)

| Dimension | Q1’26 | Q2’26 | Δ | |---|---|---|---| | 2026 financial confidence | Reaffirm ranges | Reaffirm + ≥ midpoint | More constructive | | 2027 financial confidence | Maintain $7.4–7.8B | Maintain band; lower-end bias ex deals | Slightly more cautious ex-deal | | ERCOT load CAGR | 5–6% | 4–6% | Floor softened (see Contradictions) | | Capital allocation | ~$3B equity / ~$4B growth / ~$3B residual | ~$3B equity / $4.5–5B growth (Helix) / $2–2.5B residual | More growth earmarked |

Net: qualitative raise on 2026 via ≥-mid language; 2027 is a bridge problem, not a thesis break.

GAAP revenue/EPS are reported-accounting trajectory, not earnings-power trajectory.

| Quarter | Revenue YoY | Rev Accel (bps) | Signal | |---|---:|---:|---| | 2024Q3 | +53.9% | +3,332 | Peak — EH full quarter + elevated MTM base $6,288M | | 2025Q2 | +10.5% | −1,825 | Growth fading as EH anniversarying | | 2025Q3 | −20.9% | −3,148 | Trough — tough prior-year base | | 2025Q4–2026Q1 | +13.6% → +43.4% | +3,449 / +2,985 | Reacceleration — Lotus, PJM capacity, wholesale strength | | 2026Q2 | −5.5% | −4,888 | GAAP re-deceleration — largest rate drop in window; not operating story |

EPS: half the window is NM (sign flips / MTM outliers). Q2’26 GAAP $0.76 is −6.2% YoY vs $0.81 — does not match Adj. EBITDA +30%. Prefer Adj. EBITDA / Adj. FCF.

Why GAAP can fall while Adj. EBITDA rises: IS operating revenues include contracts, wholesale ISO/RTO, and hedging/other that mark with power curves; Adj. EBITDA strips unrealized hedge MTM. Q2’26 is the textbook case.

Operating path that matters

| Period | Adj. EBITDA | YoY | |---|---:|---:| | Q1’25 | $1,240M | +52.5% | | Q2’25 | $1,349M | −4.6% | | Q3’25 | $1,581M | +9.5% | | Q4’25 | $1,742M | −12.2% | | Q1’26 | $1,494M | +20.5% | | Q2’26 | $1,767M | +31.0% |

Net verdict: GAAP growth rate mean-reverting and comp-driven; operating earnings on a compounding path with FY26 guide reaffirmed ≥ mid.

Ranked by expected impact Q3’26 → early 2027. Next leg is deal conversion, not another GAAP beat.

| # | Catalyst | Status after print | Timing | Watch | |---|---|---|---|---| | 1 | Cogentrix close (~5.5 GW gas) | FERC approved; other approvals pending; excluded from guide | Mid–late 2026 | Close 8-K; ~$700M w/ Meta to 2027 mid | | 2 | Meta PJM nuclear PPAs (~2.6 GW) | Premium excluded; deliveries late 2026 → full YE2027 | Late 2026–YE2027 | First delivery disclosure | | 3 | Helix Digital Infrastructure | Founding investor up to $1B (>$500M milestone-gated); preferred power partner | Multi-year | First named Helix + Vistra power project | | 4 | ERCOT large-load audit / Batch Zero | Abbott audit; ~couple months pause (not moratorium); CP end-2027 intact | Weeks–months | PUCT / ERCOT notices | | 5 | Additional large-load PPAs | Active negotiations both markets | Continuous | Named MW / counterparty | | 6 | PJM co-location / IRAS / RBP | FERC June order constructive; TO filing delayed ~30–60 days; oppose BYOC “stick” | Next 1–3 months | Compliance tariff filing | | 7 | FY26 Adj. EBITDA / FCF delivery | Reaffirmed; H1 ~$3.26B → H2 need ~$3.9B to mid | Q3 call ~2026-11-05 | Summer ops + hedge realization | | 8 | 2027 formal guide update | Midpoint opp held; lower-end bias ex deals | Q3 call or post-Cogentrix | Curve path + deal inclusion | | 9 | Capital return | ~$1.2B buyback auth left; YTD repurchases $709M | Through YE2027 | New authorization if lean-in | | 10 | Organic MW | Permian Peakers, Oak Hill 2, nuclear uprates inside $4.5–5B growth stack | 2026–28 | COD / construction milestones |

Hard catalyst hierarchy: (1) already delivered — Q2 Adj. EBITDA +31%, guide reaffirm ≥ mid; (2) highest-signal unresolved positives — Cogentrix, Meta deliveries, named PPAs, co-lo tariff, Helix #1; (3) risks — ERCOT curve softness into open years, Batch Zero misread as multi-year freeze, IRAS stick rules, Cogentrix timing/hedge book, street fixation on GAAP misses.

Near-term calendar

| Window | Event | |---|---| | ~2026-08-20 | PUCT open meeting — ERCOT audit / Batch Zero process | | Next 30–60 days | PJM/TO co-location compliance filing | | Mid–late 2026 | Cogentrix closing | | Late 2026 | Meta first deliveries (portion) | | 2026-11-05 | Q3 2026 earnings (FMP) — guide update, summer ops, deal status |

Call: 2026-08-07 · Burke / Moldovan / Doré / Stuckey · 7 analysts · ~15 substantive threads.

Headline: Expansive on policy/market structure; unusually clear on 2027 lower-end bias + ~$700M Cogentrix/Meta still out. Soft spots: 2028 regional hedges, Helix MW pipeline, Batch Zero inventory beyond Comanche Peak.

| # | Analyst | Topic | Badge | |---|---|---|---| | 1–2 | Constantine (Wells) | ERCOT audits / Batch Zero; contracting IRRs / hybrid | Well Answered | | 3–5 | Tonet (JPM) | PJM vs ERCOT; RBP vs bilaterals; IRAS | Well Answered | | 6–7 | Sullivan (Wolfe) | Helix path; 2026 ≥ mid / 2027 bias | Well Answered (soft on milestones) | | 8–9 | West (Melius) | Helix public path; Abbott “moratorium” | Deflected (IPO) / Well Answered | | 10–11 | Davenport (GS) | Buybacks / residual cash; 2028 hedges by region | Well Answered / Deflected | | 12–13 | Arcaro (MS) | Helix MW pipeline; Batch Zero projects | Deflected / Deflected | | 14–15 | Singh (BofA) | FERC co-lo timeline; energy+capacity appetite | Well Answered |

Tally: 10 well-answered · 1 soft · 4 deflected.

Highest-information answers

• Moldovan on 2027 (Sullivan): ERCOT forwards “meaningfully lower” vs 10/31/2025 basis; PJM + hedges + nuclear PTC “don’t fully offset” → trending lower end of $7.4–7.8B; Cogentrix + Meta ~+$700M still excluded; formal update on Q3 call (or post-close).
• Burke on queues: May-2024 ERCOT 2030 load still 115–120 GW; realistic DC ~12–15 GW by 2030 vs media “400 GW” queues (>20× overstated); supports thinning.
• Burke on prices: YTD ERCOT wholesale ~$30/MWh — “will not get new build done”; equipment/EPC costs 2–3×; existing-gen premium still discount to new-build.
• Doré on FERC co-lo: June order crystal clear; TO filing delayed; expect 30–60 days for rates/services detail.
• Burke on hedges (Davenport): explicit “I don’t think we’re going to talk that much about hedging strategies in detail on the call” — clearest full deflection.

What Street still does not know

1. 2027–28 hedge coverage by region (open MWh / % / price).
2. Helix sizing — MW, CODs, capital-call cadence, second-$500M milestones.
3. Batch Zero sites/partners beyond Comanche Peak (end-2027).
4. Helix permanent capital path (private only vs possible public vehicle).
5. Formal 2027 guide revision — only bias until Q3/post-close.

Tone: confident on ops; measured on ERCOT forwards; classic IPP deflection pattern (policy over-answered; hedge book and deal-level pipeline under-answered).

Two genuine management contradictions from seven transcripts (Q4’24 → Q2’26). Progressive capital allocation and labeled guide continuity excluded.

Checked tensions that are not contradictions

| Topic | Verdict | |---|---| | Peak vs energy load growth (Q4’25 peak 3–5% vs Q1’26 energy 5–6%) | Compatible if metrics kept distinct | | FY26 guide reaffirmed; Meta/Cogentrix still excluded | Labeled continuity | | Residual cash $3B → $2–2.5B | Progressive allocation (Helix + buybacks), not dual claims | | FMP GAAP miss vs company Adj. EBITDA beat | Different metrics — not a management self-contradiction |

Bottom line: strategic bull case is highly consistent. Only decision-relevant credibility items are the ERCOT CAGR floor and the 2027 “maintained” vs lower-end bias framing.

Macro (power-market, not classic macro)

One-line takeaway: Structural load growth intact (AI + industrial/reshoring + TX population/electrification), but soft ERCOT (~$30/MWh) + 2–3× equipment costs favor existing baseload/co-lo and Helix capital over speculative queues and battery overbuild; Meta/Cogentrix still off-guide upside.

| Theme | Statement | Beyond-VST read | |---|---|---| | Peak demand strength | July all-time peaks: PJM >168 GW, ERCOT >91 GW | Supportive for merchant/capacity owners with availability | | Load growth held | ERCOT 4–6%, PJM 2–3% through 2030; ERCOT 2030 still 115–120 GW | Not cutting multi-year demand thesis despite soft near-term prices | | Non-DC load material | TX: ~3 ppt non-DC / ~2 ppt DC inside 5–6% style CAGR | Dilutes “AI-only” risk for power complex | | New-build cost inflation | Equipment “doubled, if not tripled” | Scarcity premium for existing gen; pressure on RBP-type constructs | | Battery overbuild | Returns ~1/5 expected; queues slowing; mild days “chase to the bottom” | Bearish near-term for standalone ERCOT storage | | Texas queue audit | Supports thinning (>20× overstated queues); ~months pause not multi-year freeze | Short-term delay risk for speculative interconnects; constructive for real projects | | FERC co-location | June order constructive; TO filing delayed | Positive structural for nuclear/gas co-lo IPP model industry-wide |

Not said: Fed, rates, CPI, consumer confidence — channel is almost entirely electric load + power prices + construction cost inflation + grid policy.

Named counterparties

| Entity | Role | Implication | |---|---|---| | KKR / NVIDIA / KIA / Helix | Founding capital + preferred power partner (VST up to $1B) | Digital infra + chip capital internalizing power; multi-year AI power credibility | | Meta | ~2.6 GW PJM nuclear PPAs; premium excluded from guide | Template hyperscaler nuclear offtake; industry upside if more Meta-like deals print | | Cogentrix | Pending ~5.5 GW gas acquisition | Flexible gas consolidation continues; closing risk open for 2027 | | Lotus | Closed 3Q25 generation assets | Already in Gen EBITDA run-rate | | CEG (analyst-named Q1) | Peer “pause” color | VST still describes active negotiations Q1–Q2 — live peer divergence to track | | Battery developers (generic) | Return collapse / queue slow | Worse for ERCOT storage sponsors; better for thermal/flex if build slows into load ramp |

Cross-name implications

1. AI power stack is institutionalizing (KKR + NVIDIA + KIA + VST on Helix).
2. Meta nuclear PPAs remain the template; still outside VST guide.
3. ERCOT near-term soft, long-term intact — pure ERCOT energy beta may lag PJM scarcity names.
4. New-build cost inflation + ~$90/MWh peer new-build chat raise the bar for greenfield IRRs.
5. Policy is a share-of-wallet driver: FERC co-lo (existing-site owners), TX audit (speculative interconnects), PJM BYOC stick (could steer demand to co-lo).

| Lens | Read | |---|---| | This quarter | GAAP Miss / Operating Beat. Rev −26.5% and EPS −52.8% vs FMP; Adj. EBITDA +31% to $1.767B. | | Trajectory | Earnings-power re-accelerating (Q1 +20.5% → Q2 +31.0%); GAAP revenue noisy/decelerating. | | Guide | FY26 reaffirmed ≥ mid; 2027 band held with lower-end bias ex deals + ~+$700M deal stack still out. | | Quality signal | Guidance credibility > GAAP beat rates (L4 GAAP beat rate 25%; frequent misser on FMP tape). | | Credibility flags | ERCOT CAGR floor 5–6%→4–6% while “unchanged”; 2027 “maintained” vs lower-end Q&A. | | What moves the stock next | Cogentrix close, Meta deliveries, named PPAs, co-lo tariff clarity, Helix project #1 — not another GAAP EPS print. |