Built by Grok 4.6 High

CEG | Earnings Review

Verdict: earnings power accelerating; high-quality mid-year guide raise. Q2’26 adj operating EPS of $2.55 beat Street by +11.4% and rose +33.5% YoY vs $1.91. Revenue of $7.504B missed Street by −2.9% but still +23.0% YoY on the second full Calpine quarter. For merchant generators, adj op EPS is the scorecard — the beat is durable (Calpine accretion, higher PJM capacity prices, commercial margins), not cosmetic.

Key metrics. Nuclear capacity factor $93.0% (−180 bp YoY on planned refueling) with nuclear generation $44,160 GWh (−2.2% YoY). Total supply/sales $89.8 TWh (+36.8% YoY). Calpine regional revenue $2.15B (~28.6% of rev). Gross margin proxy still mid-40s at 46.4% (−228 bp YoY on gas mix).

New guidance. FY26 adj op EPS raised +$0.50 to $11.50–$12.50 (mid $12.00, +2.4% above Street $11.72 / +27.8% YoY vs FY25 $9.39). Drivers: commercial strength + ~$2.2B YTD buybacks. No formal rev/margin guide. LT base EPS >20% through 2029 reaffirmed.

Tone. Upgraded from Q1’s “wait another quarter” reaffirmation to a confident mid-year raise. Expansive on stranded-capacity thesis and FERC/PJM speed; closed on deal pricing ($20–$50/MWh band kept) and counterparties.

Contradictions. Material messaging conflicts on large-load timing: March “not waiting on regulatory clarity” vs multi-quarter deal delays/customer pauses; Q2’s “we were behind” recall is not in the March transcript text. ~920 MW PPAs partially repair execution optics.

Upcoming catalysts. PJM RBP/IRAS clarity into YE26; co-lo FERC order 1H’27; Crane COD 2H’27; Brazos close; Q3 guide revisit post-summer.

Primary KPI is Adjusted Operating EPS (management and Street valuation anchor). YoY = same quarter prior year only. Revenue is noisy (pass-through fuel, MTM, mix); watch EPS trajectory and nuclear ops quality.

Revenue drivers (above consolidated)

| Metric | 2023Q3 | 2023Q4 | 2024Q1 | 2024Q2 | 2024Q3 | 2024Q4 | 2025Q1 | 2025Q2 | 2025Q3 | 2025Q4 | 2026Q1 | 2026Q2 | |---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:| | Nuclear gen (GWh) | 44,125 | 45,563 | 45,391 | 45,314 | 45,510 | 45,494 | 45,582 | 45,170 | 46,477 | 45,459 | 44,666 | 44,160 | | Nuclear GWh YoY % | — | — | +6.9% | +8.2% | +3.1% | −0.2% | +0.4% | −0.3% | +2.1% | −0.1% | −2.0% | −2.2% | | Nuclear CF % | 97.2 | 95.1 | 93.3 | 95.4 | 95.0 | 94.8 | 94.1 | 94.8 | 96.8 | 93.1 | 92.3 | 93.0 | | Supply/sales (GWh) | 72,819 | 67,309 | 68,407 | 65,663 | 70,336 | 65,015 | 68,422 | 65,641 | 70,635 | 64,242 | 93,330 | 89,815 | | Supply YoY % | — | — | +4.8% | +3.9% | −3.4% | −3.4% | +0.0% | −0.0% | +0.4% | −1.2% | +36.4% | +36.8% | | Calpine rev ($M) | — | — | — | — | — | — | — | — | — | — | 2,395 | 2,147 |

Consolidated P&L trajectory

| Metric | 2023Q3 | 2023Q4 | 2024Q1 | 2024Q2 | 2024Q3 | 2024Q4 | 2025Q1 | 2025Q2 | 2025Q3 | 2025Q4 | 2026Q1 | 2026Q2 | |---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:| | Revenue ($M) | 6,111 | 5,796 | 6,161 | 5,475 | 6,550 | 5,382 | 6,788 | 6,101 | 6,570 | 6,074 | 11,122 | 7,504 | | Revenue YoY % | — | — | −18.6% | +0.5% | +7.2% | −7.1% | +10.2% | +11.4% | +0.3% | +12.9% | +63.8% | +23.0% | | Gross margin % | 44.9% | 30.7% | 44.5% | 58.1% | 52.4% | 51.9% | 35.4% | 48.7% | 45.7% | 40.8% | 42.9% | 46.4% | | Adj Op EPS ($) | 2.13 | 1.74 | 1.82 | 1.68 | 2.74 | 2.44 | 2.14 | 1.91 | 3.04 | 2.30 | 2.74 | 2.55 | | Adj Op EPS YoY % | — | — | +133% | +2.4% | +28.6% | +40.2% | +17.6% | +13.7% | +10.9% | −5.7% | +28.0% | +33.5% | | GAAP EPS ($) | 2.26 | −0.11 | 2.78 | 2.58 | 3.82 | 2.71 | 0.38 | 2.67 | 2.97 | 1.38 | 4.49 | 1.42 | | NI to common ($M) | 731 | −36 | 883 | 814 | 1,200 | 852 | 118 | 839 | 930 | 432 | 1,590 | 513 |

Absolute levels — revenue ($B) vs adj op EPS ($)

YoY growth stack (same-quarter trajectory)

| Quarter pair | Rev YoY | Adj Op EPS YoY | Supply YoY | Nuclear GWh YoY | |---|---:|---:|---:|---:| | Q2'24 / Q2'23 | +0.5% | +2.4% | +3.9% | +8.2% | | Q2'25 / Q2'24 | +11.4% | +13.7% | −0.0% | −0.3% | | Q2'26 / Q2'25 | +23.0% | +33.5% | +36.8% | −2.2% | | Q1'26 / Q1'25 | +63.8% | +28.0% | +36.4% | −2.0% | | Q4'25 / Q4'24 | +12.9% | −5.7% | −1.2% | −0.1% | | Q3'25 / Q3'24 | +0.3% | +10.9% | +0.4% | +2.1% |

Q2 earnings-power chain (Adj Op EPS YoY): +2.4% → +13.7% → +33.5% — clear acceleration.
Q2 revenue chain: +0.5% → +11.4% → +23.0% — accelerating, dominated by Calpine.
Why: Calpine accretion (~$2/sh full-year framework), higher PJM capacity prices, commercial margin/optimization; offsets = planned nuclear outages and IL ZEC timing ($85M vs $200M prior-year Q2 — FY-neutral per mgmt).

Annual context (FY2021–FY2025)

| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 | |---|---:|---:|---:|---:|---:| | Revenue ($M) | 19,649 | 24,440 | 24,918 | 23,568 | 25,533 | | Revenue YoY % | — | +24.4% | +2.0% | −5.4% | +8.3% | | Adj Op EPS ($) | n/a | n/a | 6.28 | 8.67 | 9.39 | | Adj Op EPS YoY % | — | — | — | +38.1% | +8.3% | | Nuclear GWh | 172,990 | 173,350 | 174,047 | 181,711 | 182,690 |

Read: Full-year EPS growth decelerated into FY25 (+8%) ahead of the Calpine step-up. 2026 re-bases higher — company mid path +27.8% YoY at $12.00; H1 already +30.6%.

Headline: clean Adj EPS beat +11.4%; soft revenue miss −2.9%. Adj operating EPS is the scorecard. Beat quality is high — Calpine + PJM capacity + commercial — with ZEC/outage offsets planned. L4Q / L8Q beat rate 75%; L12Q 67%. Beat magnitude improving (last 3: +0.9% → +7.0% → +11.4%).

This quarter vs consensus

| Metric | Consensus | Actual | Variance | Beat/Miss | |---|---|---|---|---| | Adj Operating EPS | $2.29 (FMP) | $2.55 | +$0.26 / +11.4% | Beat | | Operating revenue | $7.730B (FMP) | $7.504B | −$0.226B / −2.9% | Miss | | GAAP EPS | n/a | $1.42 | MTM / non-op gap vs adj | n/a |

Historical adj EPS beat/miss heatmap (this quarter ★)

Beat pattern scorecard

| Window | Beats | Misses | Beat rate | Notes | |---|---:|---:|---:|---| | L4Q | 3 | 1 | 75% | Only miss Q3'25 (−2.3%); last 3 all beats | | L8Q | 6 | 2 | 75% | Misses tiny (−1.8%, −2.3%) | | L12Q | 8 | 4 | 66.7% | Consistent majority beater |

Pattern: Consistent beater with occasional small misses. Magnitude trajectory improving — trailing three beats +0.9% → +7.0% → +11.4%. Cultural posture of conservative full-year guides raised through the year supports structural beat bias (FY26 just raised +$0.50 at mid-year).

Revenue beat/miss (secondary; do not over-weight)

| Quarter | Rev actual | Rev est | Variance | Result | |---|---|---|---|---| | Q2'26 ★ | $7.50B | $7.73B | −2.9% | Miss | | Q1'26 | $11.12B | $8.46B | +31.5% | Beat (Calpine inclusion) | | Q4'25 | $6.07B | $5.60B | +8.4% | Beat | | Q2'25 | $6.10B | $4.91B | +24.4% | Beat |

Primary signal: clean mid-year raise of +$0.50 on adj operating EPS. New mid $12.00 sits +$0.28 / +2.4% above Street $11.72 and implies +27.8% YoY vs FY25 $9.39. Q1 explicitly deferred this decision (“at least another quarter”) — Q2 delivered.

FY26 adj op EPS guide waterfall

Guidance inventory

| Metric | Prior (Q1 / Mar 31) | New (Q2 / Aug 6) | Consensus | Signal | |---|---|---|---|---| | FY26 Adj Op EPS | $11.00–$12.00 (mid $11.50) | $11.50–$12.50 (mid $12.00) | $11.72 | +$0.50 raise; mid above Street | | Base EPS FY26 | $6.65–$6.75 | $6.65–$6.75 | n/a | Unchanged modeling tool | | Base EPS FY27 | $7.60–$7.70 | $7.60–$7.70 | n/a | Unchanged | | Base EPS 2029 | $11.40–$11.90 | $11.40–$11.90 | n/a | ≥20% CAGR from 2026 base mid | | FCF before growth | 2026–27 $8.4B; 2028–29 $11.5–$13B | Reaffirmed | n/a | Multi-year scaffold intact | | 2029 cap-alloc sensitivity | ~$0.50+ upside | Floor $0.20 (done buys) to >$0.75 | n/a | $2.2B YTD locks floor | | LT base EPS growth | >20% through 2029; 10%+ rolling | Reaffirmed | Street FY27 $13.26 | Architecture intact | | Revenue / margins | Not guided | Not guided | FY26 rev $33.11B | Adj EPS is the rail |

Implied growth vs FY25 actual $9.39

| Reference | Level | YoY | |---|---:|---:| | Prior guide mid | $11.50 | +22.5% | | Street FY26 | $11.72 | +24.8% | | New guide mid | $12.00 | +27.8% | | New guide high | $12.50 | +33.1% | | H1'26 reported | $5.29 | +30.6% (run-rate ahead of FY path) |

H2 bridge: H1 $5.29 → new mid $12.00 implies H2 $6.71 vs H2'25 $5.34 (~+25.7% YoY). Raise is largely H2-facing (commercial + buyback share count); mgmt will revisit after summer on the Q3 call.

Street setup: FMP high/low $11.31–$11.99 — entire street range is at or below company mid. Constructive if CEG continues to hit mid-or-better (FY25 actual above mid).

Tone: Q1 → Q2

| Dimension | Q1'26 | Q2'26 | Delta | |---|---|---|---| | Overall | Confident, restrained reaffirm | Bullish raise + “terrific update” | Upgrade | | Guide | Reaffirm $11–$12; need another quarter | Raised +$0.50 | Action matches patience | | Contracting | Some customers paused for PJM clarity | ~920 MW signed; ~30% baseload LT | De-risking | | Regulatory | Hopeful whitepaper path | CIR waiver; FERC speed validated | More constructive | | Capital return | $335M YTD | ~$2.2B YTD; $2.8B left | Aggressive opportunistic |

What is not guided: quarterly EPS/revenue; GAAP EPS; gross/operating/EBITDA margins; formal FY27 adj op EPS range.

Net read: Structural inflection is Q1'26 (Calpine close), with Q2'26 confirmation on the EPS line. Revenue YoY cools to +23% off the Q1 +64% step-up (deal math, not demand deterioration). EPS still accelerating: +33.5% YoY, +547 bps vs prior quarter’s YoY rate.

8-quarter YoY / acceleration (Q3'24 → Q2'26)

| Metric | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26 | |---|---:|---:|---:|---:|---:|---:|---:|---:| | Revenue YoY % | +7.2% | −7.1% | +10.2% | +11.4% | +0.3% | +12.9% | +63.8% | +23.0% | | Rev Accel (bps) | +665 | −1,433 | +1,732 | +126 | −1,113 | +1,255 | +5,099 | −4,085 | | EPS YoY % | +28.6% | +40.2% | +17.6% | +13.7% | +10.9% | −5.7% | +28.0% | +33.5% | | EPS Accel (bps) | +2,620 | +1,159 | −2,265 | −389 | −274 | −1,669 | +3,378 | +547 |

YoY trajectory chart

Inflection annotations

| # | Point | What happened | Magnitude | Read | |---|---|---|---|---| | 1 | Q4'24 rev trough | Rev YoY −7.1% while EPS +40.2% | Rev −1,433 bps | Earnings quality decoupled from headline rev | | 2 | Q4'25 EPS trough | EPS YoY −5.7% (only negative in 8Q window) | EPS −1,669 bps | Sets up 2026 re-acceleration | | 3 | Q1'26 Calpine step-function | Rev +63.8%; EPS +28.0% | Rev +5,099 bps; EPS +3,378 bps | Primary structural inflection | | 4 | Q2'26 confirmation | Rev +23.0% (cool off spike); EPS +33.5% | Rev −4,085 bps; EPS +547 bps | EPS keeps accelerating — cleaner signal |

Drivers of the +$0.64 YoY EPS lift (mgmt): (1) Calpine accretion, (2) higher PJM capacity prices, (3) commercial margins/portfolio optimization; offsets = planned refueling and IL ZEC timing. Capital return (~$2.2B YTD buybacks) also widens the EPS vs revenue gap.

Q2 unlocked the catalyst stack: 920 MW nuclear PPAs, Crane NRC/FERC milestones for 2H 2027, final DOJ plant sale, and +$0.50 guide raise. Next watch is PJM RBP/IRAS clarity into YE26.

| Catalyst | Timing | Street watch | Implication | |---|---|---|---| | LT nuclear PPAs / contracting | Continuous; starts 2029–2032 | Can CEG re-accelerate after Q1 “behind schedule”? | ~920 MW signed; avg 18.5 yrs; ~30% clean baseload LT; Walmart 176 MW disclosed | | Crane Clean Energy Center | COD target 2H 2027 | CIR/transmission risk into 2030–31 | NRC fuel LAR + FERC Eddystone CIR waiver — binary path largely derisked on paper | | PJM RBP + IRAS | Bilateral now; auction fall 2026; results YE26 | How much of 6.8 GW matched bilaterally | Macro grid catalyst re-opens deal engine | | FERC co-lo / large-load rules | PJM ~Nov 2026; FERC 1H 2027 | Bankable co-lo structures | Structural enabler; not a 2026 P&L event | | Brazos Valley sale (final DOJ) | Announced 8/6; close 2H26 | Calpine integration checklist | $860M / ~$1,420/kW to LS Power; ~$5.9B total DOJ divestitures | | Share repurchases | YTD ~$2.2B; ~$2.8B left | EPS accretion already in guide | 2029 CA sensitivity floor $0.20 locked | | FY26 guide / summer weather | Q3 call ~2026-11-06 | Hold high end? Further raise? | Street Q3 EPS est ~$3.79; rev ~$9.39B | | Nuclear PTC inflation | IRS 2025 adj locked | Model lag on strike | 2030 strike $50.88/MWh → ~+$0.30/sh 2030 base | | Ginna / NMP1 SLR filings | Filed Q2 | Multi-year NRC process | Extends terminal value (NY ZEC backdrop) | | IL ZEC → Meta Clinton PPA | ZEC ends May 2027; PPA starts Jun 2027 | Handoff risk | Largely derisked commercially |

Catalyst scorecard this quarter

| Theme Street measured | Outcome | Score | |---|---|---| | Re-accelerate LT nuclear deals? | 920 MW signed; 30% baseload contracted | Beat | | Crane 2027 still real? | NRC fuel LAR + FERC CIR waiver | Beat | | Final Calpine/DOJ plant? | Brazos $860M announced | Beat | | Guide: hold or raise? | Raised to $11.50–$12.50 | Beat | | Adj EPS vs Street | $2.55 vs $2.29 | Beat | | Revenue vs Street | $7.50B vs $7.73B | Miss (secondary) |

Forward calendar (priority): Aug–Sep RBP workshops → Sep–Nov bilateral/auction window → ~Nov PJM co-lo response → Q4 Brazos close → YE26 RBP results → 1H27 FERC order + IL ZEC/Meta handoff → 2H27 Crane COD.

Call: August 6, 2026 — Joe Dominguez, Shane Smith, Dan Eggers, David Dardis. 7 analysts; ~14 substantive Qs. Street focused on nuclear PPAs, PJM/FERC large-load clarity, and stranded-capacity strategy — not the print. Tally: 10 Well Answered · 4 Deflected/Avoided.

| Analyst (firm) | Question | Management response | Assessment | |---|---|---|---| | Nick Campanella (Barclays) | 920 MW deals: price/term/customer vs Walmart? | Customers announce their own deals; consistent with LT value | Deflected (policy) | | Nick Campanella (Barclays) | PJM bilaterals / hybrids / 5 GW queue | Stranded capacity thesis; peak with batteries/DR/peakers; confidential matchmaking | Well answered | | Steve Fleishman (Wolfe) | Are new nuclear contracts in PJM? | “Most of our stuff is in PJM”; no origin pin | Soft deflection | | Steve Fleishman (Wolfe) | Does IRAS undercut existing-gen contracting? | Still need energy for ~99% of hours from existing resources | Well answered (high signal) | | Steve Fleishman (Wolfe) | Where in $20–$50/MWh range? | Keep disclosed range; transactions fit profile | Deflected | | David Arcaro (MS) | ERCOT Batch Zero / sparks / batteries | Soft expected; batteries ahead of load; will tighten as DCs land | Well answered | | Jeremy Tonet (JPM) | Cap-alloc / growth investment updates | Floor $0.20 from done buys; growth side nothing newly disclosed | Well answered | | Jeremy Tonet (JPM) | Customer tone as rules clarify | Ambiguity kills deals; clarity → velocity “with a bang” | Well answered | | Sophie Karp (KBCM) | Co-lo definitive timeline | PJM ~Nov; FERC order 1Q–2Q 2027 | Well answered (most calendar-specific) | | Sophie Karp (KBCM) | Organic new-build nuclear CapEx? | NY conversations; nothing imminent for investment dollars | Well answered (honest negative) | | James West (Melius) | Customer view of stranded capacity | “Definitely all of the above” — uprates + existing MW | Well answered | | Julien Dumoulin-Smith (Jefferies) | Bilateral timing vs RBP | RBP one avenue; signed contracts not contingent on FERC/PJM | Well answered | | Julien Dumoulin-Smith (Jefferies) | IL IRP / CMC expiration | No separate IL negotiation; NY is template | Deflected (thin on CMC) |

Themes management repeated

• Stranded capacity: peak problem, not energy shortage; use existing gen for 99% of hours.
• Rule clarity → deal velocity: ambiguity froze contracts; FERC/PJM speed unlocking pipeline.
• Customer disclosure discipline: no price/counterparty pin.
• Capital return over new nuclear FID: $2.2B YTD buybacks; new nuclear not imminent CapEx.

Open Street follow-ups

1. Midpoint of $20–$50/MWh band for new nuclear PPAs
2. Counterparty mix of the 920 MW (hyperscaler vs C&I)
3. Illinois CMC bridge post-2027
4. Bilateral MW vs 6.8 GW RBP target
5. EPA backup-gen 50-hour clarification
6. Brazos close timing / use of proceeds

Material contradictions center on large-load timing messaging. Numeric operating claims (Crane, capacity factor, guide path) are internally consistent once scopes are respected.

Explicitly not contradictions

| Topic | Why cleared | |---|---| | 13% → >20% base EPS growth | Open recalibration after Calpine | | Crane 2H'27 vs capacity credit into 2030s | Energy restart vs full deliverability — consistently distinguished | | CF >99% heat-wave vs 93% quarter | Event vs full-quarter scope | | Buyback floor $0.20 / upside >$0.75 | Explicit update for completed $2.2B | | FY26 guide $11–12 → $11.50–$12.50 | Guide raise, not conflicting ranges |

CEG’s Q2 call is not a classic macro call — useful content is power-market structure, AI/data-center build-out, nuclear PTC inflation gearing, regulatory speed, and third-party valuation signals.

Macro / sector implications

| Theme | What management said | Read-through | |---|---|---| | AI / data-center demand | Contracting resumed with rule clarity; cites MS report that AI returns are “proving out”; ~$1T data-economy infra narrative | Strong for nuclear-exposed power complex (VST, TLN, NRG) and large-load suppliers | | Inflation / nuclear PTC | 2025 IRS adj 2.8%; still assume 2% thereafter; 2030 strike $49.88 → $50.88/MWh → ~+$0.30/sh | Inflation is a modest positive for PTC-backed nuclear | | ERCOT power prices | Soft near-term as batteries lead load; DCs still under construction; will tighten | Near-term cautious for ERCOT merchant sparks; medium-term constructive if load lands | | PJM capacity / large-load rules | RBP 6.8 GW; co-lo clarity 1H'27 (was ~2029); FERC speed “unprecedented” | Rule clarity unlocks deal flow; favors existing connected generation | | Public / political risk | Data-center moratoriums real; Crane playbook (jobs/tax base); TX Batch Zero = midterm optics | Siting risk political but manageable | | Fed / rates / consumer | Not addressed | N/A |

Third-party signals

| Entity | Role | Implication | |---|---|---| | Walmart | Nuclear PPA (176 MW incl. 30 MW Dresden uprate) | Large retail C&I now contracts multi-year nuclear — not only hyperscalers | | LS Power | Buyer of Brazos at ~$1,420/kW | Competitive PE bid for efficient ERCOT CCGT even in soft market → gas fleet valuation floor | | Calpine (integrated) | DOJ divestitures ~$5.9B gross at ~$1,200/kW vs ~$960/kW purchase | Deal thesis vindicated; battery skillset now inside CEG | | Duke / Brattle studies | Cited for stranded-capacity thesis | Independent support for “use existing capacity first” | | Morgan Stanley research | AI/data-economy returns “proving out” | Supports continued hyperscale power contracting | | DOE Sec. Chris Wright | Credited on capacity-vs-energy framing | Federal messaging aligned with existing-fleet utilization | | NRC / FERC | Crane fuel LAR + CIR waiver | Restart path tangible; positive for nuclear restart peers | | EPA (backup gens) | Seek 50-hr limit carve-out for FERC curtailments | If granted, faster large-load connect; peak-management tool |

Cross-cutting summary

1. AI/data-center power demand remains the industrial macro — deal flow reaccelerating with rule clarity.
2. Inflation is a nuclear PTC tailwind, not a cost-shock narrative.
3. ERCOT soft prices = load lag, not demand death — PE still paid ~$1,420/kW for efficient gas.
4. Policy speed (FERC → PJM, EPA, DOE) is the main non-price catalyst.
5. Customer mix expanding — Walmart signals retail C&I adoption of 24/7 nuclear.

CEG printed a high-quality +11.4% adj EPS beat with accelerating earnings power (+33.5% YoY; Q2 YoY stack +2% → +14% → +34%) and raised FY26 adj op EPS by $0.50 to a mid above Street. The structural story is intact: Calpine platform + PJM capacity + commercial alpha + buybacks, with nuclear as the cash engine. Contracting re-opened (~920 MW), Crane regulatory path derisked for 2H27, and final DOJ plant exit announced. Residual risks are messaging credibility on large-load timing, ERCOT near-term softness, residual PJM design detail, and Street inability to underwrite PPA economics inside the $20–$50/MWh band. Next fundamental catalyst is Q3 print + any further guide move after summer, with RBP/IRAS clarity into YE26 the multi-year multiple driver.