Limited recourse in project finance is a lending structure where, if the project fails, the lender’s recovery is capped at the project’s own assets and cash flows plus a narrow, pre-agreed set of extras: time-limited sponsor guarantees, funded reserve accounts, and contractual triggers that can expand the lender’s claim if the borrower acts in bad faith. It is the structure that lets a sponsor raise billions of dollars to build a power plant, pipeline, or toll road without pledging the rest of its corporate balance sheet behind the loan.
Where Limited Recourse Sits Between Full and Non-Recourse
Every loan sits somewhere on a spectrum defined by how far the lender can reach when the borrower stops paying.
Full recourse is the broadest. The lender can pursue almost anything the borrower owns: bank accounts, equipment, receivables, real estate. That breadth of claim is why full recourse loans usually carry lower interest rates.
Non-recourse is the opposite. The lender is limited to the specific collateral pledged for the loan. As the IRS puts it, a non-recourse loan “does not allow the lender to pursue anything other than the collateral.”1Internal Revenue Service. Recourse vs. Nonrecourse Debt If the collateral is worth less than the debt, the lender absorbs the difference.
Limited recourse occupies the middle. The lender’s primary claim is against the project’s assets and cash flows, but the loan documents carve out specific additions: a completion guarantee for the construction phase, a reserve account, a personal liability trigger if the borrower commits fraud. Each of those additions is negotiated, time-limited, or tied to a defined event. None of them hands the lender a general claim on the sponsor’s other businesses. That middle position is what makes the structure the default for large infrastructure, energy, and resource projects.
Why Project Finance Needs This Structure
A petrochemical plant, an offshore wind farm, or a cross-border pipeline can cost billions and generate no revenue for years. A sponsor does not want a construction setback on one project to pull down its whole balance sheet. A lender, on the other hand, is not willing to finance a venture backed by projections alone with no fallback at all.
The limited recourse structure addresses both concerns through a special purpose vehicle. The sponsor forms a new, legally separate entity whose only job is to own, build, and operate the project. The debt sits inside that entity. If the project fails, the lender’s claim is confined to the SPV’s assets and the specific extras the sponsor agreed to provide. The sponsor’s other divisions and assets remain outside the lender’s reach.
That separation only works if the SPV is genuinely independent. It must keep its own bank accounts, its own books, and its own decision-making. It cannot commingle funds with the sponsor or take on debt unrelated to the project. Lenders often require that the SPV’s shares be held through an independent trust and that independent directors sit on its board with veto power over any bankruptcy filing. These “separateness covenants” exist so a court cannot later pierce the SPV and consolidate its debts back into the sponsor’s estate.
This is where project finance diverges from ordinary corporate lending. A corporate loan looks at the borrower’s whole financial picture. Project finance looks at the project itself: whether the toll road will attract traffic, whether the plant will sell enough electricity under its contracts. The project must stand or fall on its own economics, and the limited recourse structure makes that isolation legally enforceable.
Sponsor Guarantees and When They Expire
Limited recourse does not mean the sponsor is unconditionally off the hook. During the riskiest phases, sponsors typically give the lender specific guarantees. What matters is that these guarantees are narrow in scope and duration, not open pledges of corporate assets.
The most common is the completion guarantee. During construction, the project earns nothing, and the lender faces the risk that the asset is never finished. The completion guarantee obligates the sponsor to fund whatever it takes to bring the project to operational status by an agreed date, meeting defined technical and financial benchmarks. In practice, this is a full recourse obligation for the construction period. Once the project hits those milestones and starts generating revenue, the completion guarantee falls away, and the lender’s recourse narrows to the project’s cash flows and contractual protections. That handoff from broader to narrower recourse is one of the defining moments in any project finance deal.
Cost overrun guarantees work in a similar way. If construction costs exceed the budget by more than an agreed threshold, the sponsor must inject additional equity. The guarantee protects the lender from the scenario where a half-built project has consumed all of the loan proceeds with nothing operational to show for it.
In commercial real estate, limited personal guarantees cover specific risks that sit outside normal operations: environmental contamination, misuse of construction loan proceeds, failure to pay property taxes. A developer who runs the project properly never triggers these guarantees, but they give the lender a direct personal claim if something goes seriously wrong.
How the Lender Is Protected After Construction
Once the project is operating, the lender’s protection shifts from sponsor guarantees to the project’s own cash flow, backed by a network of financial covenants designed to catch trouble early.
Debt Service Reserve Account
The debt service reserve account is a dedicated bank account holding enough cash to cover principal and interest if revenue dips. Lenders typically require the DSRA to hold six to twelve months of debt service, funded from project revenues during the early operating period. If the project draws on the reserve, covenants require it to be replenished before the sponsor can take any distributions. The DSRA buys time to recover from a bad quarter without an immediate default.
Debt Service Coverage Ratios
The debt service coverage ratio measures cash flow against debt payments. A DSCR of 1.25 means the project earns 25% more than it needs to service its debt. Most project finance lenders require a minimum DSCR between 1.20 and 1.50, depending on the industry and how predictable the revenue is. A toll road with volatile traffic will face a higher requirement than a power plant with a twenty-year purchase agreement.
When DSCR drops below specified levels, the consequences escalate. A mild dip can trigger a distribution lock-up, preventing the sponsor from pulling cash out of the project until the ratio recovers. A sharper decline can activate a cash sweep, redirecting all available cash flow to accelerate debt repayment rather than building reserves or paying equity. As the project’s health deteriorates, more of its cash moves toward the lender.
Other Operating Covenants
Beyond DSCR, limited recourse loan agreements impose tight operating controls on the SPV. Capital expenditure limits stop the entity from taking on unplanned costs. Restrictions on additional debt keep new creditors from jumping ahead of existing lenders. Mandatory insurance protects the physical assets. Quarterly or monthly reporting obligations give lenders early visibility into performance. Taken together, these covenants substitute for the general credit backstop that a full recourse loan would carry.
Bad Boy Carve-Outs
The sharpest protection in a limited recourse loan is the “bad boy” carve-out. These clauses identify specific borrower actions that, if they occur, blow the limited recourse structure apart and convert the loan into a full recourse obligation against the borrower, the guarantor, or both.
The logic is simple. Limited recourse is a privilege the lender extends on the assumption that the sponsor will act honestly and preserve the collateral. If the borrower commits fraud, diverts project funds, lets the property deteriorate, or files a voluntary bankruptcy, the lender’s willingness to accept a capped recovery disappears. Common triggers include:
- Fraud or misrepresentation, such as false financial statements or concealed material facts.
- Misapplication of funds, including diversion of project revenue or loan proceeds to unauthorized purposes.
- Unauthorized transfer of collateral, such as selling or encumbering project assets without lender consent.
- Voluntary bankruptcy by the borrower, or collusion with a third party to force an involuntary filing.
- Failure to maintain the asset, including physical waste, unpaid taxes or insurance, or allowing mechanics’ liens to attach.
- Violation of separateness covenants, including commingling SPV funds with the sponsor’s other entities in a way that could undermine the project’s legal independence.
The conversion from limited to full recourse is a powerful deterrent. A sponsor facing a struggling project might otherwise be tempted to strip assets, stop paying insurance, or file a strategic bankruptcy to stall the lender. Bad boy carve-outs change that calculus, because the sponsor’s own assets suddenly come into play. Negotiations get intense here. Sponsors push to narrow the triggers and cap any resulting liability to actual damages; lenders push for broad triggers and exposure up to the full loan balance.
The Revenue Contracts That Carry the Debt
Because the lender cannot fall back on the sponsor’s general credit, the certainty of the project’s future revenue matters enormously. The contracts around that revenue often matter as much as the physical asset.
Off-take agreements are the backbone. A power plant will typically sign a long-term power purchase agreement with a utility before construction financing closes. Under a take-or-pay arrangement, the buyer commits to paying for the project’s output regardless of whether it needs the electricity, provided the plant meets minimum availability standards. Payments are sized to cover the project’s fixed costs, including debt service, even if market conditions change.
Pass-through structures add another layer by linking input costs to output prices. If fuel prices rise, the energy payment under the off-take agreement adjusts upward by a matching formula, shielding the project from margin compression. Lenders look hard at these contracts, because the credit quality of the off-taker and the enforceability of the agreement are, in practice, more important than the project’s physical plant. A modern facility without a purchase agreement is a far riskier bet than a modest one locked into a twenty-year contract with a creditworthy buyer.
What Limited Recourse Costs the Sponsor
Risk isolation has a price. Lenders charge higher interest rates on limited recourse project finance debt than they would on a comparable full recourse corporate loan, because their recovery options are capped. The premium depends on the project’s risk profile, the strength of the revenue contracts, the sponsor’s track record, and market conditions. Spreads vary across sectors and geographies, but pricing always reflects the fact that the lender is betting primarily on the project, not the sponsor.
Lenders also compensate through structural terms. Loan-to-value ratios tend to be conservative. Amortization is sculpted to match the project’s expected cash flow profile rather than a standard repayment curve. Loan agreements require detailed ongoing reporting, independent engineer inspections, and compliance certificates. Lenders active in this space develop deep industry expertise, because underwriting a desalination plant demands different technical knowledge than underwriting a highway concession.
For the sponsor, the higher cost of capital is the price of containment. A company with a $10 billion balance sheet can pursue a $3 billion infrastructure project knowing the worst case is the loss of its equity and whatever limited guarantees it provided, not the other $7 billion. That isolation is what allows sponsors to run several large projects at once without any one of them threatening the enterprise.
Tax Treatment: The At-Risk Rules
Limited recourse debt raises a specific issue under the federal at-risk rules. Under 26 U.S.C. ยง465, a taxpayer can deduct losses from an activity only to the extent they are personally at risk in it. A taxpayer is at risk for amounts contributed directly and for borrowed amounts where they bear personal liability for repayment.2Office of the Law Revision Counsel. 26 U.S. Code 465 – Deductions Limited to Amount at Risk
The statute explicitly excludes amounts “protected against loss through nonrecourse financing, guarantees, stop loss agreements, or other similar arrangements.”2Office of the Law Revision Counsel. 26 U.S. Code 465 – Deductions Limited to Amount at Risk So if a sponsor’s exposure on a limited recourse project is capped at its equity contribution plus a $50 million completion guarantee, its at-risk amount for tax purposes is capped accordingly. Losses beyond that at-risk amount cannot be deducted in the current year, though they may be carried forward.
There is an important exception for real property. Qualified nonrecourse financing secured by real estate and borrowed from a bank or government entity is treated as an at-risk amount, even though no one is personally liable for repayment. The practical effect is that real estate project finance often enjoys more favorable loss deduction treatment than infrastructure or energy project finance using a limited recourse structure. Sponsors and their tax advisors need to model the at-risk limitations early, because they feed directly into the after-tax returns that make a project viable.
Keeping Project Debt Off the Sponsor’s Balance Sheet
One strategic reason sponsors use this structure is the possibility of keeping project debt off their consolidated financial statements. Under U.S. accounting standards, the key question is whether the sponsor has a controlling financial interest in the SPV.
The analysis runs through two models. Under the variable interest entity model, the sponsor must avoid being classified as the “primary beneficiary,” which applies when the sponsor both has the power to direct the activities that most significantly affect the SPV’s economic performance and has the obligation to absorb losses or the right to receive benefits that could be significant to the SPV. Under the voting interest model, a sponsor holding more than 50% of voting interests would normally consolidate; substantive participating rights held by other equity holders can prevent that result even with a majority stake.
Structuring the SPV to avoid consolidation is a balance. The sponsor needs enough control to run the project effectively but not so much that the project’s debt lands on its balance sheet. Getting this wrong undercuts one of the main financial reasons for choosing a limited recourse structure, because consolidated project debt raises the sponsor’s reported leverage and can affect its credit rating and borrowing costs across every other activity it runs.