Report contents
At a glance
- Business model: Constellation makes money selling output from the largest US nuclear fleet into wholesale power markets at the clearing price and to commercial and industrial customers under contract, plus capacity payments for availability. The strength is scarce carbon-free baseload; the obstacle is that without clean attributes, constrained-node deliverability, or a long-term guarantee, the wholesale business is a price taker with high fixed costs.
- Financial quality: Operating cash flow was negative in each of fiscal 2021 through 2024, about $11.5bn cumulatively, and free cash flow was negative in four of five years before fiscal 2025's $4.24bn and $1.29bn. Liabilities of $63.09bn against $33.48bn of equity at March 2026 leave a thin cushion, though the acquisition roughly doubled the book.
Business model
Constellation sells the power it makes, then sells it again to businesses
Constellation Energy sells electricity. The parent company sits above Constellation Energy Generation, LLC, which files as a co-registrant and holds the generating assets. Output leaves through two doors: into wholesale power markets at whatever price clears, and to commercial and industrial customers, utilities, and power marketers under contract, through a retail supply arm. That combination is the whole business — a competitive generator with a marketing operation attached, not a regulated utility earning a set return on rate base. Type: cyclical resource producer. Stage: expansion, with a large installed fleet and a contracting strategy that data-center demand is rewriting.
One evidence caveat. Constellation's FY2025 10-K was filed February 24, 2026, and it is the right anchor for this section, but the excerpt available here stops at the cover page. That cover does confirm the December 31 fiscal year end, the dual-registrant structure, the Nasdaq listing, and two facts worth keeping: 361,990,335 shares outstanding as of February 1, 2026, and about $100.6 billion of equity market value held by nonaffiliates as of June 30, 2025. Capacity, segment revenue split, contracted volumes, and hedge ratios live in Items 1, 7, and 8, which are not in front of me. Read what follows as the shape of the model; the numbers require the full filing.
Who pays: businesses buying certainty, wholesalers paying the clearing price
Retail supply customers are businesses — a hospital system, a distribution center, a refinery — that want a defined block of megawatt-hours at a fixed or formula price. What they buy is price certainty and a supplier's credit, not electrons in the abstract. Because those contracts are usually fixed-price for a term, the supplier's margin is the spread between the price collected and the cost of the power and hedges it lined up. Gross margin here is a trading spread, not a manufacturing margin. It widens and narrows with forward power prices and with how well the book was hedged when it was written.
Wholesale buyers pay the clearing price and nothing else. No relationship, no premium for reliability, no ordinary pricing power. Volume is committed; price is taken.
That split tells you where pricing power can exist. A generator charges above market only when it sells something the market cannot otherwise buy: carbon-free attributes, deliverability into a constrained node, or a long-term supply guarantee that lets a data-center developer finance a building. Without one of those, this is a price-taking business with fixed costs and a fuel bill that never goes away.
The tradeoff: every megawatt-hour sold forward is optionality spent
The central operating decision is how much output to sell forward, and for how long. A 20-year contract with a large buyer converts volatile merchant revenue into something close to an annuity — bankable, financeable, and worth more to a lender than to an equity holder who wanted the upside. Staying open preserves the option value of high spot prices and keeps the downside with it. There is a capital dimension too: a long contract can justify an uprate or a life extension that an open merchant position cannot.
Both sides show up in results. Contract more and reported earnings get steadier while the ceiling falls. Contract less and the upside stays live while cash flow becomes a bet on weather, gas, and load. This is the volume-versus-capital-tied-up tradeoff in its purest form: every megawatt-hour sold forward is optionality spent, and every life extension is capital committed on the assumption that someone pays for the output years from now.
Where the model breaks: one commodity, one rulebook, one maintenance calendar
Three fragilities. First, the commodity is identical for everyone; in most hours a gas-fired plant sets the marginal price, so Constellation's realizations track gas and load more than its own execution. Second, a meaningful slice of the fleet's value is set administratively. Capacity markets pay plants for being available rather than for generating, and auction design, transmission rules, and the tax treatment of existing nuclear output are policy outcomes that can move revenue without a single operational change. Third, physical execution: refueling outages and unplanned trips cut volume exactly when replacement power is most expensive.
The deeper structural problem is that all three land on the same line. A rulemaking, a mild winter, and a long outage are unrelated events, yet each shows up as lower realized price or lower volume against a fixed cost base.
What would change this read: a segment table showing the retail supply book is small or structurally low-margin; disclosure that long-dated contracts are signed at prices too low to justify the capital they underwrite; or evidence that the fleet is contracting heavily while the company also spends on growth. Any of those shifts the judgment from optionality preserved to upside already sold.
Questions worth following
- Can Constellation's large-load pipeline still convert to signed contracts if hyperscaler capital spending slows?
The growth story leans on a small number of very large, long-dated supply agreements with technology buyers. If those buyers defer projects, the premium they pay for around-the-clock carbon-free power could compress through each renegotiation rather than disappear at once, shifting the growth narrative.
- Can Constellation's retail supply arm hold its commercial customer base as wholesale power prices rise?
Retail supply earns the spread between the price a business agrees to pay and the cost of procuring and hedging that power. If wholesale prices rise faster than the book reprices, margins compress or customers leave, weakening the marketing arm that gives the generation fleet a demand outlet.
- Can Constellation grow output from its existing plants through uprates and life extensions?
Adding megawatt-hours at plants already built is the cheapest growth available, because capital per incremental megawatt is far smaller than new construction. If uprates and life extensions stall, growth must come from acquisitions or newbuild at much higher cost, changing returns on reinvested capital.
Financial quality
Constellation is a competitive power producer, not a regulated utility. Earnings follow power prices, hedges, weather, and outage timing, and the balance sheet carries a thin equity layer against a large asset base. Two facts dominate the supplied record: four consecutive years of negative operating cash flow from fiscal 2021 through fiscal 2024, followed by a positive fiscal 2025; and a balance sheet that changed size in the March 2026 quarter. Total assets rose to $96.91bn from $57.25bn, goodwill to $11.53bn from $0.42bn, and equity to $33.48bn from $14.52bn. That is the signature of a large acquisition, and the latest 10-Q lists Calpine entities in its glossary of terms.
One data caveat matters. The supplied gross-profit lines mix definitions — quarterly figures near 15–25% of revenue against 76% for the fiscal 2025 annual — and investing cash flow is positive in every year, unusual for a generator. I rely on revenue, operating income, net income, operating and free cash flow, and balance-sheet totals, which reconcile internally: quarterly net income, operating cash flow, and free cash flow sum to their fiscal 2025 totals, while quarterly revenue does not ($26.14bn versus $25.53bn).
Five years: strong returns only in the best commodity year
| USD bn unless noted | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 | FY2026 Q1 (interim) | Assessment |
|---|---|---|---|---|---|---|---|
| Revenue | 19.65 | 24.44 | 24.92 | 23.57 | 25.53 | 11.12 | Adequate — flat, then a step-up with the deal |
| Operating income | -0.35 | 0.49 | 1.61 | 4.35 | 3.09 | 2.32 | Adequate — volatile |
| Net income | -0.20 | -0.16 | 1.62 | 3.75 | 2.32 | 1.60 | Weak — two loss years |
| Net margin | -1.0% | -0.7% | 6.5% | 15.9% | 9.1% | 14.4% | Adequate — cyclical |
| Gross margin | 14.0% | 8.7% | 13.3% | 25.4% | 75.8% | 42.9% | Not comparable across periods |
| ROE | -1.7% | -1.4% | 14.8% | 31.1% | 16.8% | 6.7% | Adequate — peaked, then fell |
| Asset turnover (x) | 0.41 | 0.51 | 0.51 | 0.45 | 0.46 | 0.14 | Adequate for asset-heavy generation |
| Liabilities / assets | 75.8% | 75.8% | 77.8% | 74.4% | 74.1% | 65.1% | Weak — high, eased by issuance |
| Current ratio (x) | 1.00 | 1.19 | 1.31 | 1.57 | Not disclosed | Not disclosed | Weak — thin coverage |
| Operating cash flow | -1.34 | -2.35 | -5.30 | -2.46 | 4.24 | 0.42 | Very weak until FY2025 |
| Free cash flow | -2.67 | -4.04 | -7.72 | -5.03 | 1.29 | -0.85 | Very weak — negative four of five years |
| Operating cash flow / net income (x) | n.m. | n.m. | -3.3 | -0.7 | 1.8 | 0.3 | Weak — erratic |
* Computed: FY2025 ROE uses net income over average equity of $13.85bn; FY2026 Q1 ratios are single-quarter figures and are not annualized. n.m. = not meaningful on a negative base.
Fiscal 2024's 31.1% ROE was not asset efficiency — turnover was 0.45, ordinary for the industry — but $3.75bn of profit spread over average equity of $12.05bn, with liabilities at 74% of assets. In fiscal 2022, the same structure turned a small loss into a negative return.
Cash never covered those earnings. Operating cash flow totaled roughly -$11.5bn across fiscal 2021–2024, and free cash flow was negative in four of five years. Fiscal 2025 broke the pattern with $4.24bn of operating cash flow and a ratio of 1.8 times net income. Whether that reflects the business or the reversal of earlier working-capital and collateral swings is the central question here, and the supplied statements do not answer it.
Five quarters, then a different company
Only five recent quarters are in the dataset. The 10-Q for the period ended June 30, 2026 was filed August 6, 2026, but its figures were not supplied, and net PP&E appears only for fiscal 2022 ($19.82bn), so I cannot test whether capital spending is adding productive assets.
| USD bn unless noted | Q1 FY2025 | Q2 FY2025 | Q3 FY2025 | Q4 FY2025 | Q1 FY2026 | Assessment | Trend |
|---|---|---|---|---|---|---|---|
| Revenue | 6.79 | 6.10 | 7.18 | 6.07 | 11.12 | Adequate | Flat, then a step-up |
| Operating income | 0.45 | 0.95 | 1.54 | 0.14 | 2.32 | Adequate | Lumpy; near zero in Q4 FY2025 |
| Net income | 0.12 | 0.84 | 0.93 | 0.43 | 1.60 | Adequate | Improving |
| Net margin | 1.7% | 13.8% | 13.0% | 7.1% | 14.4% | Adequate | Volatile |
| EBITDA | 1.08 | 1.65 | 1.77 | 1.30 | 3.53 | Adequate | Rising with scale |
| Total assets | 52.25 | 53.04 | 56.16 | 57.25 | 96.91 | Changed scale | +69% in one quarter |
| Total liabilities | 38.92 | 39.23 | 41.47 | 42.40 | 63.09 | Weak | +49%, funding the deal |
| Total equity | 12.96 | 13.45 | 14.35 | 14.52 | 33.48 | Strong | +131%, mostly issuance |
| Cash | 1.85 | 2.06 | 4.09 | 3.75 | 1.17 | Weak | Peaked Q3 FY2025, then fell |
| Short-term debt | 1.04 | 1.02 | 1.77 | 1.74 | 5.47 | Weak | Tripled |
| Long-term debt | 7.32 | 7.29 | Not disclosed | Not disclosed | Not disclosed | — | Coverage now incomplete |
| Goodwill | 0.42 | 0.42 | 0.42 | 0.42 | 11.53 | Watch | Jumped with the deal |
| Operating cash flow | 0.11 | 1.48 | 1.85 | 0.81 | 0.42 | Adequate | Positive but uneven |
| Free cash flow | -0.70 | 0.71 | 1.46 | -0.18 | -0.85 | Weak | Negative in two of five quarters |
| Operating cash flow / net income (x) | 0.9 | 1.8 | 2.0 | 1.9 | 0.3 | Mixed | Weakened after the deal |
Where the numbers connect
Leverage multiplies a cyclical earnings stream. Fiscal 2024's 31% ROE and fiscal 2022's negative ROE came off nearly identical liability ratios, because the equity denominator is small relative to assets. The March 2026 issuance roughly doubled that denominator, which lowers ROE arithmetic but also lowers the chance a bad commodity year wipes out book equity.
Profit stopped matching cash for three years. Net income was positive in fiscal 2023 and 2024 while operating cash flow was negative $5.30bn and $2.46bn. Hedge-collateral postings or working-capital swings could explain it and would be reversible; a structural working-capital drain would mean reported profit overstates cash. This distinction matters more than any single margin.
Capex consumes most of the best year's cash. Implied capital spending (operating cash flow minus free cash flow) was about $2.95bn in fiscal 2025 — near 12% of revenue and about 70% of operating cash flow, normal for a nuclear fleet with refueling and license work, but it leaves little free cash against a $33.5bn equity base. The March 2026 quarter's $11.12bn of revenue, roughly 1.7 times an average fiscal 2025 quarter, suggests the acquired plants are producing, though one quarter is not a run rate.
The funding mix shortened as interest cost rose. Short-term debt tripled to $5.47bn while cash fell to $1.17bn, and interest expense of $0.25bn in the March 2026 quarter alone is half the prior full year's $0.51bn. Coverage still looks comfortable — quarterly EBIT of $2.33bn is about nine times interest — but rollover and rate sensitivity both increased.
What to check
- The cash-flow statement's working-capital and collateral lines for fiscal 2022–2024, to see whether the profit-to-cash gap was reversible postings or a permanent drain.
- The purchase-price allocation and goodwill impairment testing, given that $11.53bn of goodwill is about a third of equity.
- Why net income fell to $2.32bn in fiscal 2025 from $3.75bn on higher revenue, with the tax provision rising to $1.19bn from $0.77bn — one-time items or a higher effective rate?
- Liquidity: with the current-asset split undisclosed for the last two balance-sheet dates, the current ratio cannot be computed; the revolver and maturity schedule need to come from the filing.
Questions worth following
- Can Constellation keep cash earnings ahead of reported profit as its larger asset base is integrated?
Four straight years of negative operating cash flow before fiscal 2025 make cash conversion the clearest test of earnings quality here. If the positive fiscal 2025 reading was working-capital or collateral timing rather than a change in the business, reported profit overstates what the fleet actually produces.
- Can Constellation term out the short-term debt it took on without draining its cash balance?
Short-term debt tripled to $5.47bn while cash fell to $1.17bn in the March 2026 quarter, and interest expense for that quarter alone was half the prior full year. How that funding is refinanced sets the refinancing risk and the interest burden carried against a volatile earnings stream.
- Will the equity Constellation issued for its largest acquisition earn a return above its cost of capital?
The March 2026 balance sheet shows equity of $33.48bn against $14.52bn at year-end, mostly newly issued to fund the deal. If the larger equity base does not produce proportionally higher profit, returns on equity fall and the acquisition destroys value even as revenue grows.
Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.
Continue reading the full report
- Key risksWhat could break the investment case
- Valuation and margin of safetyWhat the current price assumes, and how much room for error it leaves
- SummaryOur overall judgement and what to watch next
Free to sign up · 20 research points · unlocking this report uses 10
20 points cover 2 full reports. No card needed.
Research reflects its generation date. Preview the first two modules.
More on Constellation Energy Corp
Constellation Energy Corp
Business, moat and key questions
Open company page