PPA Credit Support: A One-Way Street Running the Wrong Direction

A project posts security in favour of its offtaker to cover a construction period of two or three years. The offtaker owes payments for fifteen to twenty. Security runs in one direction, and it is not the direction where the money sits.

That asymmetry is deliberate and, for an investment-grade utility, defensible. It also means the single largest exposure in a financed project — two decades of contracted revenue from one counterparty — is usually uncollateralised and held together by a credit rating. When the rating moves, a posting obligation fires, at precisely the moment the counterparty finds posting hardest.

This is the ninth post in the series, and the first of the three standalone contract topics that close out series D. You will get what the seller actually posts and how it is sized, why the buyer normally posts nothing, how downgrade triggers work and why they are procyclical, what the alternatives to a letter of credit now are, and how to size the uncollateralised exposure rather than assume it away.

ℹ️ Note: This describes how these mechanisms work in practice. It is not legal or investment advice — security packages are heavily negotiated and vary by offtaker type and market.

Who Posts Security to Whom?

Almost always the seller to the buyer. Stoel Rives puts it flatly: "the seller posts security in favor of the offtaker, but the utility offtaker almost never posts security in favor of the seller."

The reasoning is credit, not fairness. Utility offtakers typically carry investment-grade ratings, which makes them acceptable counterparties without collateral. The project company on the other side is a special purpose vehicle with no operating history, no balance sheet and an asset that does not yet exist. One party has a rating; the other has a plan.

Three forms are acceptable, per Stoel Rives: "cash deposited in escrow, a letter of credit from a highly rated ('A' or better) bank, or a guarantee from a creditworthy entity." They also note that cash is "virtually never posted as security," because a special purpose entity's liquidity is precisely what it does not have spare.

That leaves the letter of credit as the workhorse — which matters, because the LC market has changed.

How Is the Seller's Security Sized?

Differently before and after commercial operation, because the risks are different. During development the offtaker's exposure is schedule. After COD it is delivery.

Development period security is sized directly off the date ladder. Stoel Rives describes it as "the per diem amount of any delay damages... multiplied by the number of days between such target commercial operation date and the 'drop dead' date."

Read that against post 3 and the structure snaps into place. The development security is the PPA's delay damages, running for the full width of the window between the target COD and the longstop, converted into a posted instrument. It is the offtaker collateralising exactly the exposure the sponsor was modelling from the other side. Where the sponsor sees a dead zone and uncovered days, the offtaker sees a number it can draw on.

Two consequences follow, and neither is obvious from the PPA alone:

• A longer longstop is not free. Extending the drop-dead date widens the multiplier and increases the security the project must post — so schedule flexibility is bought with liquidity.
• Negotiating the PPA delay damages rate down reduces the posting obligation as well as the liability. The two move together because they are the same number.

Post-commercial operation security shifts basis entirely. Stoel Rives puts it "somewhere between six and 18 months of expected payments under the PPA," and notes an interesting wrinkle in levelized-price agreements: amounts occasionally increase until a "crossover" point around years 12 to 15, on the theory that utilities pay more than justified in the early years and are therefore more exposed to a seller that stops delivering.

That crossover logic is worth holding onto. It is the offtaker explicitly modelling its own unamortised exposure through the life of the contract. The seller's side of the same question — how exposed am I to this buyer over twenty years — is rarely modelled with the same rigour, and there is usually no instrument sitting against it at all.

Why Does the Buyer Usually Post Nothing?

Because a rating is treated as a substitute for collateral, and for a genuinely investment-grade utility it largely is. The offtaker's promise to pay is backed by a regulated business with predictable cash flows and a public rating that would suffer if it defaulted.

The substitution holds until two things stop being true.

The offtaker is not a utility. Stoel Rives notes that different standards apply where buyers "lack strong credit ratings or operate as subsidiaries without independent assets." Corporate PPAs are the growth area and they are precisely the case the convention was not built for — an unrated corporate, or a rated parent contracting through an unrated subsidiary, has the legal form of an offtaker without the credit that justified going uncollateralised.

The rating moves. Creditworthy buyers commonly avoid posting upfront but become obligated if their "credit rating falls below a negotiated threshold, such as investment grade levels."

The second is the one to model, because it is a contingent obligation on a counterparty whose capacity to meet it is correlated with the trigger. A downgrade below investment grade is not a random event — it happens when the business is deteriorating, when funding is tightening and when bank lines are least available. The mechanism designed to protect the project fires at the moment it is hardest to satisfy, and the consequence of a failure to post is typically a default and a termination right the project may not want to exercise.

ℹ️ Note: Check what happens if the buyer fails to post after a downgrade. A termination right against a deteriorating offtaker is a right to lose your contracted revenue and go merchant — which is why the practical remedy is usually a renegotiation, and why the rating trigger's real function is to force one.

Why Have Letters of Credit Become Harder?

Because bank capacity for them has tightened while demand has grown. Energetic Capital attributes the squeeze to "tighter regulatory capital requirements" and "internal credit concentration rules," against a rising volume of requests as the market expands beyond investment-grade corporates.

The cost side matters as much as the availability side. An LC is not a one-off: it is a facility carrying an annual fee for as long as it stands, which Energetic Capital describes as something that "can quickly become an expensive annual item paid by the project." Post-COD security running six to eighteen months of PPA revenue, held for years, is a standing charge against the same cash flow the debt is sized on.

That has pushed sponsors toward alternatives. Energetic Capital lists five, each with a real trade-off:

Option	Advantage	Drawback
Credit insurance / risk transfer	"Doesn't burn up scarce bank capacity or tie up collateral"	Premium cost
Parent company guarantee	"No direct out-of-pocket costs"	"Uses up the parent's guarantee capacity (hard to scale!)"
Bank guarantees / alternative facilities	Flexibility on draw conditions	Still consumes bank capacity
Tripartite / public-sector backing	Government or quasi-governmental credit	Availability is programme-dependent
Portfolio-level credit insurance	Pricing improved by "risk diversification benefits"	Requires scale to justify

The parent guarantee line is the one sponsors underestimate. It appears free because no fee is paid, and it is not free — it consumes a finite resource. A sponsor with ten projects cannot post ten parent guarantees against the same balance sheet without eventually being told so by its own treasury or its lenders. The cost is simply deferred to the eleventh project.

The instrument choice also changes the draw mechanics, which is the point post 2 made about EPC security and which applies identically here. An on-demand LC pays against a compliant demand. A guarantee may require establishing default first. The gap between those two is measured in months of litigation at exactly the time the project needs cash.

When Does the Seller Get Its Security Back?

On milestones, and the release schedule is worth as much attention as the amount. Security that is posted early and released late is a longer, more expensive obligation than the headline figure suggests, and the difference is pure carrying cost.

The natural break is commercial operation. Development security exists to cover the offtaker's schedule exposure, and that exposure ends when the plant delivers — so the instrument should be released at COD and replaced by the post-COD performance security. Two failure modes recur.

Overlap. If development security is not released until the post-COD instrument is in place and accepted, the project briefly carries both. On the numbers below that is roughly $12.8m of simultaneous posting, against a special purpose entity with no spare liquidity. The overlap is usually administrative rather than intended, and it is fixable by drafting the release as automatic on the same event that triggers the replacement.

Drawn-down security is not restored security. If the offtaker draws on development security for delay damages, the instrument is reduced. Most PPAs require the seller to replenish it to the full amount within a short period. That obligation arrives during a delay — when the project is already absorbing carrying cost and has already established that it is behind schedule. The replenishment covenant is a liquidity call at the worst possible moment, and it belongs in the downside case rather than in the legal summary.

Step-downs over the operating period follow the same logic in reverse. Where the crossover mechanic applies, the instrument grows for twelve to fifteen years before declining, so the highest posting requirement can fall in the middle of the term rather than at the start.

Why Do Lenders Care About Collateral Assignment?

Because the PPA is the security. The debt is sized against contracted revenue, so the contract producing that revenue is the most valuable asset the lenders take — more valuable, in a distressed case, than the hardware.

PPAs therefore contain provisions permitting the seller to assign the agreement as collateral, together with protections for the lenders taking it. Those protections matter more than the assignment itself, and they run in two directions.

Cure rights before termination. As post 3 covered, Skadden describes a lender right to step in and cure a project company default before the offtaker can terminate as a near-universal feature of project financings, typically documented in a direct agreement with the offtaker. Without it, an offtaker could terminate for a curable seller default and extinguish the lenders' principal security before they had an opportunity to act.

Notice and standstill. The direct agreement normally requires the offtaker to notify the lenders of a seller default and to hold off on termination for a defined period. That standstill is what converts a theoretical cure right into a usable one — a cure right without notice is a right to fix something you were never told about.

The practical consequence for the security package is that the seller's posted instruments and the lenders' direct agreement protect different things. The LC covers the offtaker's monetary exposure to delay or non-delivery. The direct agreement covers the lenders' exposure to losing the contract entirely. A project can have ample posted security and no meaningful lender protection, or the reverse, and the two are negotiated by different parties at different times.

One diligence check is worth doing early: confirm the offtaker has actually agreed to the direct agreement terms the lenders will require, before the PPA is signed rather than after. Retrofitting consent from a counterparty that has already secured the commercial terms it wanted is materially harder.

How Do You Price the Exposure the Buyer Never Collateralised?

As an expected loss, not as a binary. Pexapark frames PPA counterparty risk through the standard credit decomposition: expected loss is driven by "Probability of Default (PD)," exposure at default, and loss given default.

That framing is more useful than it first appears, because it converts an argument about whether a counterparty is "bankable" into a number that can sit in a model next to everything else. The project is not asking whether the offtaker will default. It is asking what the twenty-year contracted revenue stream is worth after adjusting for the chance that it stops.

Pexapark also lists the mitigations that operate before any instrument is posted: "advance payments, MACs (Material Adverse Clauses), increased payment frequency," alongside third-party enhancements including credit insurance, LCs and parent guarantees, and collateral held as "cash, liquid securities" in escrow.

Increased payment frequency is the quietly effective one. Moving from monthly to semi-monthly settlement roughly halves the exposure at default on receivables at no capital cost to either side. It does nothing about the loss of the contract itself, but it meaningfully reduces the amount sitting unpaid when a default arrives.

How Do You Model Credit Support in Excel?

Three blocks: what you must post, what you are owed and cannot collect, and what the gap is worth.

The inputs

Assumptions, labelled as such:

Contract capacity                     150 MW
PPA price                             $45.00 /MWh
P50 annual energy                     250,000 MWh
Annual PPA revenue                    $11,250,000
PPA delay damages                     $12,000 /day
Target COD                            2027-06-01
Drop dead date                        2028-06-01
Post-COD security requirement         9 months of expected payments
LC annual fee                         1.40%
Offtaker one-year PD                  0.60%
Loss given default                    65%

Block one — what the seller posts

Development_Security = PPA_Delay_Damages_Per_Day × (Drop_Dead − Target_COD)
                     = 12,000 × 366                            = $4,392,000

Post_COD_Security    = Annual_Revenue × (9 / 12)               = $8,437,500

LC_Annual_Cost       = 8,437,500 × 1.40%                       = $118,125

Note what the first line is doing. Widening the longstop by 90 days adds 90 × 12,000 = $1.08m to the posting obligation. The schedule flexibility negotiated in the PPA has a direct liquidity price, and it is rarely quoted alongside the drafting concession that creates it.

Block two — what the buyer does not post

Buyer_Security_Posted = $0    (investment-grade offtaker, no posting until downgrade)

Receivable_Exposure   = Annual_Revenue × (Settlement_Days / 365)
                      = 11,250,000 × (45 / 365)                = $1,386,986

Contract_Exposure     = NPV of remaining PPA margin over term

The receivable is the small, visible number. The contract exposure — the value of losing a twenty-year offtake and replacing it at merchant or at whatever the market offers a distressed seller — is the large invisible one, and it is the reason a rating downgrade is a financing event rather than an administrative one.

Block three — the expected loss

Annual_EL = PD × Exposure_At_Default × LGD
          = 0.60% × 1,386,986 × 65%                            = $5,409

Downgrade_Contingent = Buyer posting obligation triggered at sub-IG
                     = 9 months of payments                    = $8,437,500

The annual expected loss on receivables is trivial — which is exactly why nobody collateralises it, and exactly why the number is misleading on its own. The risk is not the receivable. It is the contingent event: a downgrade that either produces $8.4m of buyer-posted security or produces a default and a termination right nobody wants to use.

The comparison worth reporting

Seller_Posts     = 4,392,000 (development) → 8,437,500 (post-COD)
Buyer_Posts      = 0, contingent on downgrade
Seller_Exposure  = ~20 years of contracted margin
Buyer_Exposure   = ~2 years of construction schedule

Four lines. The party posting security is the party with the shorter exposure, and the party with the longer exposure posts nothing. That is not necessarily wrong — it is a considered allocation based on relative credit — but it should be stated rather than assumed, because it is the structure the whole financing rests on.

ℹ️ Note: Model the LC fee as an operating cost across the full period it is held, not as a closing cost. At 1.4% on $8.4m held for fifteen years it is roughly $1.8m of undiscounted cash, which is the kind of item that disappears into "other" in a sources and uses table.

To run the full version — development security sized off your own date ladder, post-COD security stepped through a levelized-price crossover, LC fees carried as a period cost, and the downgrade trigger modelled as a contingent obligation — prompt Dezzmond with your PPA security provisions and price curve.

What Do Lenders Actually Check?

Lenders read the security package twice: once for what the project must fund, and once for what protects the revenue they are lending against.

• What must the project post, and when? Development security and post-COD security are both uses of funds and belong in sources and uses, not in a footnote.
• What form, and what is the draw mechanic? On-demand instruments are treated very differently from guarantees requiring proof of default.
• What is the LC fee over the full holding period? A standing annual charge against the cash flow the debt is sized on.
• Is the offtaker rated, and by whom? An unrated corporate or an unrated subsidiary of a rated parent are different credits from a rated utility.
• What is the downgrade trigger, and what happens if the buyer fails to post? Including whether the resulting termination right is one the project would ever want to exercise.
• How frequent is settlement? It sets exposure at default on receivables directly.
• Does a guarantee reach the full obligation? A guarantee limited to defined obligations may not cover a termination payment.

Frequently Asked Questions

Why does the seller post security when the buyer owes far more money?

Because the credit positions are asymmetric. A utility offtaker typically holds an investment-grade rating and is treated as an acceptable unsecured counterparty, while the seller is a special purpose vehicle with no balance sheet and an asset that does not yet exist.

How is development security sized?

Stoel Rives describes it as the per diem PPA delay damages multiplied by the number of days between the target commercial operation date and the drop-dead date — which means it is the date ladder from post 3 converted into a posted instrument.

How much post-COD security is typical?

Stoel Rives puts it at somewhere between six and 18 months of expected PPA payments, occasionally increasing until a crossover point around years 12 to 15 in levelized-price agreements.

What happens if the offtaker is downgraded?

A negotiated threshold, commonly investment grade, triggers a posting obligation. The difficulty is that the trigger correlates with the counterparty's ability to satisfy it, so the practical outcome is often a renegotiation rather than collateral.

Why is cash almost never posted?

Because a special purpose entity's liquidity is scarce and expensive. Stoel Rives notes cash is "virtually never posted as security," leaving letters of credit and guarantees as the practical options.

Closing: The Instrument Is Not the Exposure

Credit support is the part of a PPA that looks most like a solved problem. There is a defined amount, a defined form, a defined trigger, and a bank standing behind it. Everything is quantified.

What is quantified is the seller's obligation. The project's own largest exposure — two decades of payments from a single counterparty — sits on the other side of the same section, uncollateralised, supported by a rating and a contingent posting obligation that fires when the rating fails. That is a reasonable structure against a regulated utility and a considerably less reasonable one against an unrated corporate, and the drafting frequently looks the same in both cases.

The series pattern holds for the ninth time. The development security is the date ladder from post 3, priced. The LC draw mechanics are the security question from post 2, repeated with a different counterparty. And the exposure that matters is the one with no instrument against it — which is how every post in this series has ended, because that is where the money is.

Next: deemed generation, and how an offtaker pays for power it refused to take.

Sources: Stoel Rives — The Law of Solar: Utility-Scale Power Purchase Agreements · Energetic Capital — Credit Support Options for Renewable PPAs · Pexapark — Pricing Credit Risk in the PPA Market · Skadden — Lenders' Relationships with Project Counterparties