[CEG] Constellation: Q3 2026 Earnings Test Raised Guidance After Calpine

Constellation is the largest nuclear operator in the United States, and after completing its merger with Calpine in January 2026 it owns about 55 GW of nuclear, natural gas, geothermal, hydro, wind and solar capacity[1]. Third quarter 2026 (three months ended September 30, 2026) results are scheduled for 2026-11-09, making Constellation's Q3 2026 earnings the first full-summer test of the combined company[2]. In the latest disclosed quarter, the second quarter of 2026, adjusted operating earnings rose to $2.55 per share from $1.91 a year earlier, while GAAP net income fell to $1.42 per share from $2.67; nuclear output in the same quarter was 44,160 GWh, down from 45,170 GWh, and the capacity factor slipped to 93.0% from 94.8%[3]. With second-quarter results, management raised full-year 2026 adjusted operating earnings guidance to $11.50-$12.50 per share[3]. The company does not give quarterly guidance. Drillr's compilation shows a third-quarter EPS consensus of $3.876 from 9 analysts (range $3.836-$3.926), a third-quarter revenue consensus of $9.233 billion from 5 analysts with a range of $5.962 billion to $11.976 billion, and a full-year EPS consensus of $12.167 from 13 analysts[4]. The unusually wide revenue range shows that analysts do not agree on how Calpine's consolidated revenue should be counted[4].

Three things matter most in these results. The first is whether the commercial margins that drove the guidance increase can last through the summer peak: management attributed the raise to higher realized customer margins and successful portfolio optimization[5], adjusted operating EPS for the first and second quarters totals $5.29[6][3], and the $12.00 guidance midpoint therefore requires about $6.71 in the second half, while management said it would revisit full-year guidance on the third-quarter call after peak summer demand ends[5]. The second is whether the nuclear fleet can hold summer output in a heavy refueling year: the company planned 15 refueling outages for 2026[5], first-half refueling outage days reached 185 versus 129 a year earlier, and the Illinois zero-emission credit (ZEC) reference price fell from $6.64 to $1.12[3], while the PJM nuclear capacity price rose to about $329 per MW-day for the 2026/27 planning year[5]; the third quarter is the first summer in which these forces offset each other. The third is how much Calpine's gas fleet earns in its first summer inside Constellation: Calpine segment RNF was about $1.125 billion in the first quarter and $1.028 billion in the second[7], and the equivalent forced outage factor of the gas, oil and pumped-storage fleet rose from 4.5% in the first quarter to 6.2% in the second[6][3]; if summer spreads recover only modestly, the year's largest source of profit growth will slow.

Company Background and Business Structure

Constellation owns the largest nuclear fleet in the United States, and that is the most basic difference between it and other independent power producers. It holds interests in 14 operating nuclear stations with 25 units, about 22 GW of nuclear capacity that produced 183 TWh in 2025, and nuclear supplied 68% of its total electric supply[1]. After merging with Calpine in January 2026, the company describes itself as the world's largest private-sector power producer[1], and the cash it paid for the acquisition in the first half, net of cash acquired, was $2.537 billion[3]. It is also restarting the shut Three Mile Island Unit 1, renamed the Crane Clean Energy Center, under a 20-year power purchase agreement with Microsoft[1]; the NRC has approved the licensing amendment, and the project remains on track to restart in the second half of 2027[5].

Constellation manages its business by power market region and from 2026 reports six segments: Mid-Atlantic, Midwest, New York, ERCOT, Other Power Regions and Calpine, each measured by RNF, which is operating revenue minus purchased power and fuel[7]. Beyond generation, the company packages owned and purchased power for customers: based on EEI data it is the nation's largest energy supplier by commercial, industrial and residential power volumes, serving about 2 million customer accounts, including three-fourths of Fortune 100 companies[1]. After Calpine joined, the customer business supplies about 275 million MWh of electricity and 800 BCF of natural gas a year across 40 states, more than 80% of it to commercial and industrial customers[8]. Portfolio hedging is concentrated in the prompt three years, when customer demand and market liquidity allow effective price risk management[1].

Revenue does not come only from selling power; it also includes capacity revenue, Illinois zero-emission credits and the federal nuclear production tax credit (PTC), and these policy-linked revenues move opposite to power prices[1][7]. Reported nuclear output includes the company's ownership share of the Salem and South Texas Project stations[3]. On long-term contracts, Constellation has signed with Microsoft (Crane) and with Meta for the Clinton station's output for 20 years starting June 2027[1]; in February it signed a 380 MW agreement with CyrusOne to serve a new data center next to the Freestone gas plant[6], and in the second quarter it signed another 920 MW of long-term nuclear power purchase agreements, 15 to 20 years long, with investment-grade customers and deliveries starting from 2029 through 2032[3].

Financial History and Current Position

2025 was the last full year before Calpine was consolidated, with operating revenue of $25.533 billion, purchased power and fuel of $14.681 billion and RNF of $10.852 billion[1]. According to Drillr financial data, 2025 net income was $2.319 billion, down from $3.749 billion in 2024, with GAAP EPS of $7.40 and adjusted operating EPS of $9.39 (versus $8.67 in 2024). One direct reason for the profit decline is that higher power prices cut nuclear PTC revenue from $2.080 billion in 2024 to $320 million in 2025, while nuclear capacity factors were 94.4%, 94.6% and 94.7% in 2023 through 2025, showing stable operations[1].

Operating cash flow swings widely from year to year and cannot be inferred from earnings. Operating cash flow was −$5.301 billion in 2023, −$2.464 billion in 2024 and $4.237 billion in 2025, capital expenditures were $2.422 billion, $2.565 billion and $2.949 billion, and share repurchases were $992 million, $999 million and $400 million[1]. Only after cash flow turned positive in 2025 did the company have room to fund an acquisition, capital spending and large buybacks at the same time in 2026.

Earnings in 2026 already show the Calpine consolidation clearly. First-half operating revenue was $18.626 billion, operating income $2.912 billion, net income attributable to common shareholders $2.103 billion and GAAP diluted EPS $5.88[3]. First-quarter GAAP net income was $4.49 per share (versus $0.38) and adjusted operating EPS was $2.74 (versus $2.14)[6]; second-quarter adjusted operating earnings were $920 million versus $599 million, adjustments included $149 million of amortization of acquired commodity contracts and $84 million of Calpine merger and integration costs, and average diluted shares rose from 314 million to 360 million[3].

The balance sheet has taken on clearly more leverage because of the acquisition and buybacks. First-half operating cash flow was $1.553 billion (versus $1.584 billion), capital expenditures $2.521 billion, the Calpine cash payment net of cash acquired $2.537 billion, buybacks $1.971 billion and dividends $309 million; at June 30 the company held $697 million of cash, $5.226 billion of short-term borrowings, $5.743 billion of assets held for sale, $11.527 billion of goodwill and $98.253 billion of total assets[3]. It also had $7.0 billion of available capacity under its credit facilities at June 30[7], and Moody's and S&P affirmed its Baa1 and BBB+ credit ratings after the Calpine close[9].

Operating Model

Revenue can be broken into generation volume times price, plus contract and policy revenue. Operating revenue is roughly nuclear output multiplied by realized power price, capacity revenue, state credits and PTC, plus gas generation times power price, retail and wholesale load contract revenue and natural gas sales, plus or minus realized hedges; in 2025 segment RNF was $3.702 billion in the Midwest, $3.411 billion in the Mid-Atlantic, $1.600 billion in New York, $1.137 billion in ERCOT and $819 million in Other Power Regions[1]. In the first half of 2026, total reportable-segment RNF was $7.629 billion, including $2.153 billion from Calpine[7]. Higher contracted prices are the main source of growth in the legacy segments: in the second quarter, retail load revenue rose $160 million in the Mid-Atlantic and $110 million in the Midwest on higher contracted energy prices[7], and these contract prices flow into RNF in the same quarter the power is delivered.

On the profit side, the key is fixed costs and the automatic adjustment of policy revenue. Adjusted operating earnings equal RNF minus O&M, depreciation, other taxes, interest and income tax; nuclear costs are mostly fixed, so higher output lowers unit cost. The PTC is worth up to $15 per MWh and phases out as annual gross receipts move between $26.00 and $44.75 per MWh[1], and nuclear capacity revenue is included in that gross receipts calculation[10], so higher power and capacity prices automatically shrink the PTC. Management splits earnings into two types: base earnings from contracts, the PTC floor and 10-year-average commercial margins make up 60%-70% of the total, while enhanced earnings from power prices above base assumptions, commercial margins above the 10-year average and volatility make up 30%-40%[9]; management said base EPS will grow from $6.65 in 2026 to at least $11.40-$11.90 in 2029, a compound annual growth rate of at least 20%[11].

Cash flow moves more slowly than earnings and depends more on capital allocation. Operating cash flow was −$2.464 billion in 2024 and $4.237 billion in 2025[1], and uses of cash include maintenance and nuclear fuel spending, about $3.9 billion of growth capital in 2026-2027 and a $5.0 billion buyback authorization[9]. Management projects $8.4 billion of free cash flow before growth for 2026-2027 combined and $11.5-$13.0 billion for 2028-2029[5], and aims to return the balance sheet to target credit metrics by the end of 2027[11]. Regulator-required divestitures are another cash source: five mostly gas-fired PJM plants with about 4.4 GW are being sold to LS Power for $5.0 billion[7], and the 606 MW Brazos Valley plant in Texas is being sold for $860 million, a deal the company said would close by year-end[3].

Industry and Competitive Position

Constellation's advantages are nuclear scale and customer reach, and both are hard for peers to replicate. Based on EEI data, it is the nation's largest nuclear operator and largest supplier by commercial, industrial and residential power volumes[1]; Vistra, NRG and Talen are independent power producers that also run nuclear or gas plants in PJM and ERCOT and serve data centers, but the available material does not provide directly comparable financial figures for them. In PJM's capacity auction for the 2028/29 planning year, every Constellation plant in PJM cleared, including 15,700 MW of nuclear and 3,175 MW of fossil and other capacity, for a total of 18,875 MW at $325 per MW-day[10]; the ComEd capacity reference price in the second quarter was $289.67 versus $109.25 a year earlier[3].

Constellation's weakness is that prices and contracting pace in its main markets depend on market rules and on how fast load connects, which the company does not fully control. Management acknowledged that ERCOT prices are under near-term pressure because battery storage and new generation came online before data center load[5], and that some customers paused contracting to wait for regulatory clarity while others kept negotiating[8]. PJM has proposed a Reliability Backstop Procurement (RBP) with an auction planned this fall for a 6.8 GW target and results expected by year-end[5]. Calpine has no prior-year comparison, and the company cannot calculate Calpine's net income separately, which limits any cross-company judgment of the acquired assets' profitability.

Core Debates

Having raised guidance on customer margins and portfolio optimization, can Constellation carry that momentum through the peak-summer third quarter?

This question determines how credible full-year guidance is. On March 31 the company set initial 2026 guidance at $11.00-$12.00[9], and in August it raised both ends by $0.50 to $11.50-$12.50, citing first higher realized customer margins and successful portfolio optimization, partly offset by higher O&M from performance-related compensation[5]. First-half adjusted operating EPS totals $5.29[6][3], the guidance midpoint requires about $6.71 in the second half, and the third quarter is both the year's peak demand period and the point at which management said it would revisit guidance[5].

The evidence shows real growth in the legacy business, but not all of it comes from customer margins. Second-quarter adjusted EPS growth reflected the addition of Calpine and favorable market and portfolio conditions, partly offset by nuclear outages[3]; RNF across the five legacy regional segments totaled $2.928 billion, up 9.8% from $2.667 billion, with Mid-Atlantic retail load revenue up $160 million and realized hedges down $90 million, Midwest capacity revenue up $65 million and Illinois ZEC revenue down $135 million, and Other Power Regions losing about $205 million of wholesale and retail revenue on lower New England load[7]. An alternative reading is that EPS growth mainly came from consolidating Calpine, whose second-quarter RNF was $1.028 billion[7], and that part of the legacy gain came from PJM capacity prices the company does not control; average diluted shares were also about 15% higher than a year earlier in the second quarter[3]. The transmission runs from higher-priced load contracts and portfolio optimization to regional segment RNF, plus Calpine RNF, less O&M, depreciation and interest, to adjusted operating earnings ($920 million in the second quarter), divided by 360 million diluted shares to reach $2.55 per share[3].

Three sets of figures matter in the third quarter. First is adjusted operating EPS against the $3.88 consensus[4] and the company's $3.04 a year earlier; second is whether the $11.50-$12.50 full-year guidance is kept, narrowed or raised; third is whether RNF for the five regional segments excluding Calpine exceeds the prior-year $2.992 billion, with retail load revenue still a positive driver. A third quarter below $3.60 or a cut to the low end of guidance would show that first-half commercial outperformance is not sustainable, and a year-over-year decline in legacy segment RNF would show that growth relies only on consolidation.

In a heavy refueling year, can summer nuclear output and higher capacity prices offset shrinking Illinois subsidies?

Nuclear is the foundation of Constellation's profit, and 2026 output is weighed down by refueling outages. The fleet has about 22 GW of capacity and produced 183 TWh in 2025[1]; the company planned 15 refueling outages in 2026[5]. First-quarter nuclear output was 44,666 GWh (versus 45,582 GWh) at a 92.3% capacity factor (versus 94.1%)[6]; second-quarter output was 44,160 GWh (versus 45,170 GWh) at 93.0% (versus 94.8%), with 86 refueling outage days versus 41, bringing first-half refueling outage days to 185, 56 more than a year earlier[3]. High power prices automatically shrink Illinois ZECs, carbon mitigation credits (CMC) and the federal PTC, while PJM capacity prices entered a new planning year in June 2026, so the third quarter tests all three forces at once.

The Midwest segment shows that shrinking policy revenue has already begun to outweigh higher capacity revenue. Second-quarter Midwest RNF was $979 million, down 5.5% from $1.036 billion, mainly because revenue recognized for Illinois ZECs delivered in prior planning years fell by $135 million, which a $65 million increase in capacity revenue only partly offset; the average CMC price was −$5.91 per MWh versus −$0.42 a year earlier[7], and the Illinois ZEC reference price fell from $6.64 to $1.12[3]. An alternative reading is that the second-quarter decline mainly reflected refueling timing and a one-time ZEC recognition in 2025, that the third quarter has fewer refuelings, and that the PJM nuclear capacity price rose from $270 in the 2025/26 planning year to $329 per MW-day in 2026/27 (covering 9,350 MW)[5], so the Midwest could return to growth; but the higher summer power prices go, the larger the negative CMC payments become. The transmission runs from nuclear output to saleable volume and replacement power costs, while rising capacity prices and falling ZEC, CMC and PTC revenue offset each other, ending in Midwest and Mid-Atlantic segment RNF and then adjusted operating earnings.

The third-quarter tests are specific. Can nuclear output approach the 46,477 GWh the company reported a year earlier, and can the capacity factor stay above 95% (96.8% a year earlier)? Can Midwest RNF hold at or above the prior-year $943 million? And in which direction do the 10-Q driver explanations move capacity revenue, ZEC/CMC and the PTC? A capacity factor below 94% or more than 20 non-refueling outage days would signal unplanned losses, and a year-over-year Midwest RNF decline of more than 5% would show that shrinking policy revenue exceeded the capacity gain.

In its first summer inside Constellation, can Calpine's gas fleet earn meaningfully higher spreads than in spring?

Calpine is the largest source of 2026 profit growth, and gas plants earn most of their profit at peak demand. First-half Calpine segment RNF was $2.153 billion, while RNF across the five legacy regional segments was $5.476 billion, only about $341 million more than the $5.135 billion a year earlier[7]. The third quarter is the first test of how much the acquired units earn in summer; meanwhile about 5 GW of assets that regulators required the company to divest will close in stages, including about 4.4 GW of five PJM plants sold for $5.0 billion[7] and the 606 MW Brazos Valley plant sold for $860 million with closing expected by year-end[3], after which their earnings leave the consolidated results.

Current data show Calpine weakening sequentially in spring. Calpine segment RNF was about $1.125 billion in the first quarter and $1.028 billion in the second, when operating revenue was $2.147 billion and purchased power and fuel $1.119 billion[7]; Calpine generated 24,914 GWh in the second quarter[3], and the gas fleet's equivalent forced outage factor rose from 4.5% in the first quarter to 6.2% in the second[6][3]. On the first-quarter call management said the company was well hedged against near-term price weakness and saw long-term Texas power prices as too low[8], and on the second-quarter call it acknowledged that ERCOT prices were under pressure because storage and new generation arrived before data center load[5]. An alternative reading is that the second-quarter decline reflected seasonal spring maintenance and low demand and that summer will rebound; but if spreads in ERCOT and California stay weak, the summer rebound may be limited. The transmission runs from gas generation times spark spread, plus capacity and resource adequacy revenue and Calpine retail load margin, to Calpine RNF, then through O&M, depreciation and acquisition-related interest to adjusted operating earnings; once divestitures close, the related earnings leave and the proceeds go to debt reduction or buybacks.

In the third quarter, the questions are whether Calpine segment RNF is clearly above the second quarter's $1.028 billion, whether the gas fleet's forced outage factor falls back below 6%, and how approvals and closings progress for the five PJM plants and Brazos Valley, which were carried as $5.743 billion of assets held for sale at June 30[3]. If Calpine RNF shows no sequential increase in summer, the acquired gas assets are earning less than expected; if divestiture closings slip, funding for buybacks and debt reduction will be delayed.

After the October Amazon and Google agreements, can contracting, uprate investment and buybacks all be funded from cash flow and divestiture proceeds?

Long-term contracts are central to management's base earnings growth target and are the way to turn volatile enhanced earnings into visible earnings. Management targets base EPS growth of at least 20% a year compounded over 2026-2029[11]. According to trade press reports, in October the company signed a 20-year agreement with Amazon covering 690 MW of the 1,790 MW Calvert Cliffs station and two agreements with Google: a 20-year power purchase agreement for 890 MW of uprates at 11 existing units and a 15-year supply agreement covering 2,700 MW of existing units that adds no capacity[12]; the Google agreement involves more than $4.3 billion of investment by Constellation, with the first uprate due by 2028[13]. These investments, buybacks and repayment of Calpine acquisition borrowings all have to be paid for from cash flow and divestiture proceeds.

Current funding shows that capital allocation is already stretched. In the second quarter the company signed 920 MW of new long-term nuclear contracts with an average term of 18.5 years, all with investment-grade counterparties, and long-term contracts now cover about 30% of its clean baseload output; it has repurchased about $2.2 billion of stock this year, leaving $2.8 billion of authorization[5], and in July it bought back about one million more shares for about $250 million[7]. First-half operating cash flow was $1.553 billion, capital expenditures $2.521 billion, the net Calpine cash payment $2.537 billion and buybacks $1.971 billion, leaving $5.226 billion of short-term borrowings and $697 million of cash at June 30[3]. An alternative reading is that contract prices have not been disclosed and the new megawatts arrive only from 2028 on[13], while near-term cash flow must first carry capital spending and buybacks; PJM told FERC that allowing existing capacity resources to contract with new large loads would not solve the problem its bring-your-own-new-capacity (BYONC) scheme is meant to address, so the regulatory treatment of the 2,700 MW existing-unit contract remains uncertain[12]. The transmission runs from long-term contracts priced above the PTC floor, phasing in from 2027, to contracted clean margin and base earnings; about $3.9 billion of growth capital in 2026-2027[9] and divestiture proceeds together shape debt and interest; buybacks lower diluted shares, with full-year guidance assuming 357 million average diluted shares[5] versus 361 million in the initial guidance[9], and that flows through to EPS.

The Amazon and Google agreements are so far known only from press reports, and the third-quarter report will be the company's first formal disclosure of them. Investors need to see how the report confirms the capacity, terms and capital spending of the two agreements and updates long-term contract coverage; whether September-end short-term borrowings fall below the June-end $5.226 billion and whether divestitures have closed; and whether average diluted shares fall below 357 million as buybacks continue. If short-term borrowings keep rising while divestitures slip, capital allocation has outrun cash flow; if PJM rules turn against contracts for existing units, the economics of the 2,700 MW contract will suffer.

Risks and Falsifiers

The first risk is a policy change to the federal nuclear PTC and state credits. The PTC is worth up to $15 per MWh and acts as a floor when power prices are weak[1]; the base earnings framework values available nuclear output at the PTC floor price, so a change to the credit or unfavorable Treasury guidance would remove that support. Nuclear PTC revenue fell from $2.080 billion in 2024 to $320 million in 2025[1] and was only $25 million in the first half of 2026[7], showing that it contributes almost nothing in high-price years but acts as a floor when prices fall; if the PTC remains in place through 2032 with continued inflation adjustment, this risk does not materialize.

The second risk is commercial margins reverting to the mean. Management counts commercial margins above the 10-year average and volatility gains as enhanced earnings, which make up 30%-40% of 2026 total earnings[9]; these move with market conditions and are not as visible as contract revenue. The line directly exposed is consolidated adjusted operating EPS; if the third-quarter driver explanations keep listing higher realized customer margins as a positive and full-year guidance is maintained or raised, the concern weakens.

The third risk is unplanned nuclear outages or extended refuelings. When refueling outages run longer than planned or units trip unexpectedly, capacity factors fall, the company bears higher replacement power costs or sells less power, and O&M rises[1]; in the second quarter, 45 more refueling outage days than a year earlier lowered the capacity factor by about 1.8 percentage points[3]. The exposure sits in nuclear regional segment RNF, and losses grow as summer prices rise; a third-quarter capacity factor of at least 95% with no more than 10 non-refueling outage days would show that the risk has not materialized.

The fourth risk is weak ERCOT and western power prices. Storage and new generation are arriving faster than data center load and compressing spark spreads[5], and Calpine's gas assets are concentrated in ERCOT and California, with first-half Calpine segment RNF of $2.153 billion[7]. If third-quarter Calpine RNF is at least $1.25 billion and the driver explanations do not list ERCOT prices as a main negative, the concern weakens.

The fifth risk is that PJM rules for large-load interconnection and co-location are not settled. Results of PJM's Reliability Backstop Procurement are expected by year-end[5], and both customer contracting pace and the treatment of contracts for existing units depend on these rules; Constellation cleared about 15,700 MW of nuclear capacity in PJM for the 2028/29 planning year[10], and how the rules treat existing units directly affects whether that capacity can lock in premiums through long-term contracts. If third-quarter official filings confirm more than 1 GW of new contracts and FERC approves PJM's RBP design, this risk falls markedly.

What to Watch Next

• Commercial margins and raised guidance: compare third-quarter adjusted operating EPS with the $3.88 consensus and $3.04 a year earlier, and check whether $11.50-$12.50 guidance is kept or raised; below $3.60 or a lower guidance floor would falsify the current view.
• Legacy segments: second-quarter RNF for the five regional segments was $2.928 billion (+9.8%); watch whether third-quarter RNF beats the prior-year $2.992 billion, since a decline would mean growth relies only on consolidation.
• Nuclear summer output: second-quarter output was 44,160 GWh at a 93.0% capacity factor; watch for output near 46,477 GWh and a capacity factor above 95%, while below 94% or more than 20 non-refueling outage days signals unplanned losses.
• Midwest segment: second-quarter RNF was $979 million (−5.5%); watch whether it holds at or above the prior-year $943 million, with a decline of more than 5% as the falsifier.
• Calpine's first summer: second-quarter RNF was $1.028 billion with a 6.2% forced outage factor; watch for clearly higher RNF and an outage factor below 6%, with no sequential summer growth as the falsifier.
• Capital allocation: June-end short-term borrowings were $5.226 billion and second-quarter diluted shares 360 million; watch for lower borrowings, fewer than 357 million shares and closed divestitures, while rising borrowings with delayed divestitures would falsify the funding case.
• Amazon and Google agreements: the second quarter added 920 MW and long-term contracts cover about 30% of clean baseload output; watch the formal disclosure of capacity, terms, capital spending and coverage, and any PJM rule that disadvantages existing units.

Conclusion

Constellation's earnings rest on three businesses: low-marginal-cost baseload power from about 22 GW of nuclear capacity, peak gas generation from Calpine, and a commercial business that sells that power to commercial and industrial customers. In the first half of 2026, adjusted operating EPS totaled $5.29 and full-year guidance was raised to $11.50-$12.50, but nuclear output fell year over year, Illinois credits shrank sharply and short-term borrowings rose to $5.226 billion at June 30. The central unresolved relationship is whether the customer margins and portfolio optimization behind the higher guidance are a durable commercial capability, or the combined effect of the Calpine consolidation, higher capacity prices and a favorable market.

Since the August 6 second-quarter report, full sell-side research reports could not be found through public channels, and the independent commentary available centers on the October 6 Google agreement. Sonal C. Patel, senior editor at POWER magazine, noted that only the 890 MW of uprates in the Google deal is new capacity, that the 2,700 MW existing-unit contract adds none, and that PJM has told FERC that letting existing capacity resources contract with new large loads would not solve the problem, so the regulatory treatment of that contract remains uncertain[12]. Michael Douglass, writing for TIKR, argued that long-term contracts from two technology giants for new capacity make earnings estimates more dependable, but the company still has to deliver the uprates on time and on budget, the new megawatts arrive only from 2028, and neither company disclosed what Google is paying[13]. Both pieces agree that the agreements will not change third-quarter profit; they differ in emphasis, with POWER focused on rule risk and TIKR on execution and timing, and both bear on the long-term contracts and capital allocation debate rather than on summer results themselves.

The current understanding would be clearly strengthened if the third quarter delivered this combination: adjusted operating EPS at or above consensus with full-year guidance kept or raised, legacy regional segment RNF above the prior year, a nuclear capacity factor back above 95%, Calpine RNF clearly above the second quarter's $1.028 billion, and short-term borrowings falling as divestitures close. Conversely, if legacy segment RNF declines, Midwest RNF falls by more than 5%, Calpine shows no sequential summer gain, and short-term borrowings keep rising while divestitures slip, the first-half guidance increase would look more dependent on consolidation and one-off market conditions, with the earnings from long-term contracts unable to make up the difference until after 2028.

Sources

[1] CEG FY2025 10-K filed 2026-02-24 · 2026-02-24 · 10-K · https://www.sec.gov/Archives/edgar/data/1868275/000186827526000032/ceg-20251231.htm

[2] Drillr earnings calendar CEG updated 2026-11-02 · 2026-11-02 · Drillr earnings calendar

[3] CEG Q2 2026 8-K filed 2026-08-06 · 2026-08-06 · 8-K · https://www.sec.gov/Archives/edgar/data/1868275/000186827526000097/ceg-20260806991.htm

[4] Drillr analyst estimates CEG Q3 2026 retrieved 2026-11-02 · 2026-11-02 · Drillr analyst estimates · https://gateway.drillr.ai/mcp/private

[5] CEG Q2 2026 earnings call 2026-08-06 · 2026-08-06 · earnings-call · https://investors.constellationenergy.com

[6] CEG Q1 2026 8-K filed 2026-05-11 · 2026-05-11 · 8-K · https://www.sec.gov/Archives/edgar/data/1868275/000186827526000063/ceg-20260511991.htm

[7] CEG Q2 2026 10-Q filed 2026-08-06 · 2026-08-06 · 10-Q · https://www.sec.gov/Archives/edgar/data/1868275/000186827526000104/ceg-20260630.htm

[8] CEG Q1 2026 earnings call 2026-05-11 · 2026-05-11 · earnings-call · https://investors.constellationenergy.com

[9] CEG 8-K filed 2026-03-31 (2026 Business and Earnings Outlook) · 2026-03-31 · 8-K · https://www.sec.gov/Archives/edgar/data/1868275/000186827526000045/ceg-202603318kexh991.htm

[10] CEG 8-K filed 2026-07-14 · 2026-07-14 · 8-K · https://www.sec.gov/Archives/edgar/data/1868275/000186827526000080/ceg-20260714.htm

[11] CEG Business and Earnings Outlook call 2026-03-31 · 2026-03-31 · earnings-call · https://investors.constellationenergy.com

[12] POWER magazine on Constellation-Google deal 2026-10-07 · 2026-10-07 · POWER magazine · https://www.powermag.com/google-deal-backs-890-mw-of-constellation-nuclear-uprates

[13] TIKR on Constellation-Google deal 2026-10-06 · 2026-10-06 · TIKR · https://www.tikr.com/blog/constellation-energy-soars-14-on-model-4-3-billion-google-nuclear-deal