What does a lender mean by contracted revenue?
A defined payment from an identified counterparty for a defined period. Contracted revenue is not merely revenue the sponsor expects; it is revenue somebody has promised, in a document a lender can read, for a term the lender can match against the debt. Everything short of that is a forecast wearing a contract's clothes.
It is a spectrum rather than a switch, and the rungs matter. At the top sits availability-based revenue, where the project is paid for being ready to perform rather than for what anyone consumes, which removes demand risk from the project entirely and places it with the public counterparty. Below that sits a full-requirements or fixed-volume offtake at a fixed price. Below that, a contract for differences or a hedge that fixes price but leaves volume with the project. Below that, a shaped or partial contract covering some of the output. And at the bottom, merchant exposure, where the project sells into a market at whatever the market pays.
Lenders read that ladder for three separate things, and sponsors usually only notice the first. Price certainty is one. Volume certainty is a different one, and a fixed price on an uncertain volume is not a contracted project. Counterparty credit is the third, and it is decisive: a twenty-year contract with a counterparty nobody will underwrite for twenty years is a ten-year contract at best, and lenders size to the shorter number.
What does the contracting market look like right now?
Large, concentrated, and no longer growing in a straight line. BloombergNEF counted 55.9 GW of corporate clean power purchase agreements announced globally in 2025, down about 10% on the record set the year before and the first decline in nearly a decade. The United States set a record within that total at 29.5 GW; the Europe, Middle East and Africa region fell about 13% to 17 GW; and the Asia-Pacific region dropped to 6.9 GW from 10.7 GW (BloombergNEF, drawn from its 1H 2026 Corporate Energy Market Outlook, published 19 February 2026).
Concentration is the part a sponsor should read twice. On the same BloombergNEF count, four buyers, Amazon, Meta, Google and Microsoft, accounted for 49% of global deals, with Meta and Amazon together at 20.4 GW. A market where half the contracted demand sits with four names is a market where offtake availability is a function of whether your asset fits a very small number of procurement programmes, not a function of general appetite.
Pricing has stopped moving in one direction. LevelTen Energy's North American index for the second quarter of 2026 recorded solar offer prices down 4.8% on the quarter on a market-averaged basis, and down 1.8% excluding the California market, while wind offers rose 5.5% on the quarter and 17.5% on the year (LevelTen Energy, North American PPA Price Index, Q2 2026, published 21 July 2026; LevelTen publishes the percentage moves openly and holds the absolute price levels behind subscription, so no price level is quoted here). Solar and wind moving in opposite directions in the same quarter is a reminder that contracted does not mean uniform, and that the technology decides the negotiation as much as the structure does.
| 2025 volume | Change on 2024 | |
|---|---|---|
| Global | 55.9 GW | Down about 10%, first decline in nearly a decade |
| United States | 29.5 GW | A record |
| Europe, Middle East and Africa | 17 GW | Down about 13% |
| Asia-Pacific | 6.9 GW | Down from 10.7 GW |
| Amazon, Meta, Google and Microsoft combined | 49% of global deals | Meta and Amazon alone at 20.4 GW |
“Sponsors describe an asset as contracted when they have a signature. Lenders call it contracted when the price, the volume, the term and the counterparty's credit all survive the downside case at once. Those are different tests and the second one loses deals.”
What is a contract actually worth in debt terms?
More than a margin, and less than a published number. It has to be said plainly, because the internet is full of confident answers: no rating agency, lender association or market data provider publishes a current, general series showing how much more leverage or how much cheaper debt a contracted project raises than an otherwise identical merchant one. The only quantified statement of that kind we could source was a power-sector rating methodology from 2008 that has since been superseded, and whose contracted and merchant coverage measures use different denominators and therefore cannot be compared like for like. We are not going to invent the missing number.
What is published, and what is worth more than a spread, is performance. The Global Infrastructure Hub's Infrastructure Monitor 2023, republishing Moody's project finance loan data as at 2021, records a ten-year cumulative default rate of 1.7% for green energy projects against 6.0% for conventional energy projects (Global Infrastructure Hub, Infrastructure Monitor 2023, published December 2023). That gap is not a statement about technology. Green energy financings in that period were overwhelmingly built on long-dated contracted offtake, and conventional generation carried far more merchant and dispatch exposure. The default record is the revenue structure showing up in the outcome.
The practical effect on a financing shows up in four places at once, and it compounds. A contracted stream supports a lower coverage target because the lender's downside case cuts less deeply into it. It supports a longer tenor, because the contract gives the lender a term to lend against. It carries a lighter reserve and sweep package, because there is less to reserve against. And it opens a wider lender universe, because institutional infrastructure debt mandates are frequently written to require contracted revenue outright. Any one of those alone would matter. Together they change what the asset can carry.
How do lenders finance merchant exposure at all?
By sizing to a floor and structuring around the rest. A lender confronting merchant revenue does not model the sponsor's price curve; it applies its own, usually a consultant's low case, and sizes the debt so that coverage holds at that level. Everything above the floor accrues to the project, and most of it is trapped rather than distributed.
The structural devices are consistent across markets. A cash sweep applies surplus above a defined coverage level to principal, which deleverages the project faster in strong price years and shortens the effective life of the exposure. A reserve tail requires the debt to be fully repaid some years before the asset's expected end of life, so there is always residual value behind the loan. And amortization is front-loaded against the period the lender can see, rather than spread evenly across a horizon it cannot.
The result is a financing that works and costs the sponsor optionality. Merchant projects can be financed; they are financed constantly. What they cannot do is hold leverage and distribute cash at the same time, which is precisely what a contracted asset can do. A sponsor whose return model assumes early distributions from a merchant asset has usually modelled a contracted project by mistake.
What should a sponsor do with a partially contracted asset?
Size the tranches to the certainty, not the asset to the average. Most real projects are neither fully contracted nor fully merchant, and the common mistake is to blend the two into one revenue line and one facility. The better structure treats them as what they are: a contracted stream that supports long-dated, lower-coverage debt, and a merchant stream that supports a smaller, faster-amortizing, more heavily swept layer, or no debt at all.
Then negotiate the contract for what the financing needs rather than for what the commercial team wants. Term matched to the debt rather than to the sales cycle. Termination and change-in-law provisions that a lender will accept without a sponsor guarantee behind them. Credit support from the offtaker sized to survive its own downgrade. And curtailment, shape and basis risk allocated explicitly, because a fixed price on power the project cannot deliver where the contract requires it is not the protection it appears to be.
Finally, sequence the two processes deliberately. A contract signed before lenders have seen it is a contract that gets re-diligenced with no ability to change it, and the terms a lender would have asked for become the terms that reduce the debt. Running offtake negotiation and lender conversations in parallel costs a little coordination and routinely buys more capacity than the price negotiation does.
As of September 2026
Sources: corporate clean power purchase agreement volumes, regional splits and buyer concentration are from BloombergNEF, drawn from its 1H 2026 Corporate Energy Market Outlook, published 19 February 2026. Quarterly power purchase agreement price direction is from LevelTen Energy, North American PPA Price Index, Q2 2026, published 21 July 2026, using only the percentage moves LevelTen publishes openly; the absolute price levels are subscriber-gated and are not quoted. Ten-year cumulative default rates for green energy and conventional energy project loans are from the Global Infrastructure Hub, Infrastructure Monitor 2023, published December 2023, which republishes Moody's project finance bank loan data as at 2021. No contracted-versus-merchant leverage, margin or coverage differential is quoted, because no named institution publishes one as a current general series; the only quantified statement located was a 2008 power-sector rating methodology since superseded, whose contracted and merchant measures are not comparable. Nothing here describes a transaction or engagement.
This position sits within our project finance practice.