Lending against essential assets is not really a yield play. You are buying a lower loss rate: contracted, regulated cash flows that default less and recover more when they do.
Key takeaways
- Infrastructure debt lends against essential assets through an SPV, repaid from the asset’s own contracted or regulated cash flows rather than a company’s earnings.
- The defensiveness is measurable. Project finance bank loans have recovered about 76.8% on default versus roughly 52% for unsecured corporate debt, and the marginal default rate falls below 0.01% by around year 17 (Moody’s).
- The demand driver is structural, not cyclical: a ~US$94 trillion build-out to 2040, ~US$2 trillion a year of clean energy investment, and a data centre boom now financing itself increasingly through debt. It is a genuinely global market: Europe and Australia are two of the deepest pools, not just North America.
- Most direct infrastructure debt is institutional, but listed vehicles give ordinary investors an access route. On the London Stock Exchange, renewables and infrastructure investment trusts such as The Renewables Infrastructure Group (TRIG), NextEnergy Solar Fund and Greencoat Renewables hold contracted energy assets and carry asset-level debt you can buy in a normal brokerage account.
Infrastructure debt sits inside private credit, but it does not behave like a normal corporate loan book. The borrower is usually not a company with a product and a growth plan. It is a single asset that people have to keep paying for: a regulated electricity network, a contracted solar farm, a water utility, a port, a fibre backbone. The loan is repaid from that asset’s own cash flows, ring-fenced at project level and shaped by contracts and regulation rather than by quarterly earnings.
That structural difference is the whole point, and it is measurable. When you lend against an essential asset with contracted revenue, you default less and you recover more. What infrastructure debt finances, how the deals are put together, where the returns and risks sit, and how the loss numbers compare with ordinary corporate lending are all covered below. Our view is that what you are really buying is not yield. It is a lower loss rate and a demand pipeline that runs for decades.
What infrastructure debt actually finances
Infrastructure debt is lending to fund, build, refinance or expand essential physical assets. It is typically arranged as senior secured lending to a project company, a special purpose vehicle (SPV) that owns the asset and its contracts, rather than to a general corporate borrower. Repayment is expected to come from the project’s own cash flows, backed by asset-level security.
Corporate private credit works the other way. It lends to an operating business, and repayment depends on enterprise-wide earnings, refinancing access and management execution. It can be very well protected, but the core exposure is still to how a company performs. Infrastructure debt is closer to asset-backed lending with contractual cash flows, even when the borrower sits inside a larger sponsor group.
The practical implication most allocators skip over: in infrastructure debt you are underwriting cash flow resilience and documentation, not EBITDA growth. The question is not “will this business win its market” but “will this asset keep getting paid, and what do I control if it does not”.
The assets themselves fall into a few broad sectors:
- Energy: renewables, transmission and distribution networks, storage, district heating
- Transport: roads, bridges, airports, ports, rolling stock
- Digital: data centres, fibre, telecoms towers, where the cash flows are contracted
- Utilities: water, wastewater, waste management
- Social: hospitals, schools and civic buildings, often financed through availability-based public contracts
What infrastructure debt tends not to fund, or funds only at much higher risk and return: early-stage development with no permits and no offtake contract, merchant projects selling into spot markets with no contracted revenue, and assets whose income is fully discretionary and behaves more like a trading business than an essential service.
Why this matters now
Two forces are pushing institutional money toward infrastructure debt, and both are structural rather than cyclical.
The first is the sheer scale of what needs building. The Global Infrastructure Hub’s outlook, produced with Oxford Economics across 50 countries and seven sectors, put global infrastructure investment need at roughly US$94 trillion by 2040, rising to US$97 trillion once you add the UN goals for universal water and electricity access (GI Hub, 2018). On current spending trends about US$18 trillion of that stays unfunded. That gap is a multi-decade financing pipeline across new build and refinancing.
The second is the energy transition, which is capital-intensive and increasingly debt-financed once a project moves from development into construction and operation. Global energy investment passed US$3 trillion for the first time in 2024, with about US$2 trillion of it going to clean energy technologies and grids (IEA, 2024). Equity does not carry that alone. Debt becomes the workhorse the moment a project is de-risked into contracted or regulated cash flows.
Digital infrastructure is the newest leg of the same story. Investment in data centres has nearly doubled since 2022 to around half a trillion dollars in 2024, and the IEA expects data centre electricity demand to more than double to about 945 terawatt-hours by 2030, driven mostly by AI computing (IEA, 2025). Contracted data centres and the fibre and power that feed them are becoming a mainstream infrastructure debt sector rather than a niche one.
This is a global market, and the deepest pools are not all in North America. Europe runs some of the largest dedicated infrastructure debt strategies anywhere: Allianz Global Investors closed its second Infrastructure Credit Opportunities fund at around €1 billion in 2025, and AXA IM Alts raised roughly €1.05 billion for a European senior infrastructure debt fund (AllianzGI, 2025; AXA IM Alts). Australia and the wider Asia-Pacific region are the other heavyweight market, where superannuation funds and managers such as Macquarie have financed infrastructure debt at scale for decades. Fund managers here quote in euros, pounds and Australian dollars as readily as in US dollars, and the borrower and the currency depend on where the asset sits.
Put those together and you get a market where lenders with structuring skill can stay selective. The deal flow does not depend on a risk-on mood; it depends on capital that has to be spent and debt that has to be refinanced.
How the deals are put together
Most infrastructure debt deals are built around the SPV. The vehicle owns the asset and the contracts and borrows under a credit agreement. Security is taken over the project’s assets and cash flows, and the lender’s rights are set by covenants and step-in mechanisms that are far more granular than mainstream corporate lending.
The structures you see most often:
- Brownfield senior debt: lending against operating assets with established revenues. Lower risk, tighter margins.
- Greenfield (construction) debt: lending through the build. Risk turns on contractor strength, the engineering and construction contract, contingencies and completion tests.
- Refinancing and acquisition facilities: used when a sponsor buys an asset and reworks its capital structure.
- Holdco debt: lending at a sponsor holding company sitting above the asset. Usually riskier, because the cash flows are one level removed and structurally subordinated to the asset-level debt.
The protection that makes infrastructure debt defensive is not a slogan, and it is worth being precise about where it comes from. Stability sits in some combination of contracted revenues (long-term offtake agreements, availability payments, capacity contracts), regulated returns (the regulated asset base models used for networks), and demand for an essential service that stays fairly steady through a cycle. On top of that sits structural control: cash waterfalls, debt service reserve accounts, locks on distributions to equity, and step-in rights that let the lender replace an operator and enforce against the asset. If you use infrastructure debt as a portfolio building block, you are buying contracted cash flow plus that structural control. The yield is only half of what you are paying for.
Where the returns come from
Returns in infrastructure debt are driven by contracted coupon income plus the spread you earn for complexity, illiquidity and specialist underwriting. In many strategies the base case is simple: get paid on time, get repaid at par. There is usually less upside participation than in the corporate private credit styles that carry fees, call protection and equity kickers.
That does not make it a low-return asset by default. The pricing range is wide because “infrastructure” spans very different risk buckets: investment-grade-like regulated networks at one end, construction-period renewables, digital build-outs and holdco debt at the other. Your realised return is mostly a function of three things: where you sit in the capital stack, how much risk is genuinely transferred through the contracts, and how tight the documentation is when a deal drifts off plan.
Inside a private infrastructure debt fund the return usually gets built from the interest margin (the core contracted yield), upfront and commitment fees for arranging and allocating capital, prepayment protection in some deals, and amortisation, where the loan repays principal over its life and reduces how much refinancing the asset ever has to do. Fund-level mechanics matter too. Some managers lift returns with portfolio-level borrowing; others run unlevered portfolios with a cleaner risk profile. It pays to know which one you are buying, because leverage at the fund layers financing risk on top of the asset risk you actually wanted. For the wider picture of how these funds generate returns and where they get into trouble, our private credit guide covers the mechanics.
The numbers behind “defensive”
This is where infrastructure debt earns its reputation rather than claiming it. The clearest long-run evidence comes from Moody’s, which has tracked project finance bank loans since 1983.
Two figures do most of the work. Ultimate recoveries on defaulted project finance bank loans have averaged about 76.8% over 1983 to 2020, and in almost two-thirds of default cases the eventual recovery is 100%, meaning no economic loss at all (Moody’s, via GI Hub 2021). On the default side, the marginal default rate on infrastructure project loans falls below 0.01% by around year 17: once a project is built and operating, the risk of it defaulting keeps dropping the longer it runs (Moody’s, via IPFA).
Assembled next to corporate debt, the contrast is the whole investment case:
| Metric (Moody’s data) | Infrastructure / project finance debt | Corporate debt |
|---|---|---|
| Average ultimate recovery on defaulted debt | ~76.8% (project finance bank loans, 1983 to 2020) | ~52% on senior unsecured; lower than the historical average in 2020 |
| 10-year cumulative default rate | ~6.7% (unrated project finance loans) | Consistent with low investment-grade corporate credit |
| Default risk over time | Marginal default rate falls below 0.01% by ~year 17 | Broadly stable to rising with the credit cycle |
| Credit quality by year 11 | Comparable to Baa3-rated corporate loans (5.11% vs 5.13% cumulative default) | Baa3 benchmark |
Sources: Moody’s via GI Hub, 2021; Moody’s via IPFA. Figures cover rated and unrated project finance bank loans and are indicative of the asset class, not any single fund.
The pandemic tested it in real time. Moody’s rated infrastructure universe saw eight defaults in 2020 and seven in 2021, barely above the 1983 to 2021 average of four a year, and only one of those was directly attributable to Covid (GI Hub, 2021). Corporate recoveries in 2020, by contrast, came in worse than their long-run average. The label held up when it was stress-tested.
Now the worked version, because a recovery rate only matters once you turn it into a loss. Expected credit loss is default probability multiplied by loss given default, where loss given default is one minus the recovery rate. Take a book of senior secured infrastructure loans with a broadly investment-grade 10-year default profile of around 5% and Moody’s 76.8% recovery. Loss given default is 23.2%, so expected loss over the period is roughly 5% × 23.2%, about 1.2%. Run the same 5% default rate through a 52% corporate recovery, and loss given default is 48%, so expected loss is about 2.4%. Same headline default rate, double the loss, purely because of what you recover when a deal breaks. That recovery gap is what you are actually paying the illiquidity premium for, and it is why Macquarie frames infrastructure debt as roughly 70% recovery against about 52% for unsecured corporate debt (Macquarie, 2024). (Worked figures are illustrative, using published class-level inputs, not a forecast for any fund.) The Moody’s dataset spans project loans across the Americas, Europe, the Middle East, Africa and Asia-Pacific, so the defensive case does not belong to one country. Wherever the asset sits, the source of the resilience is the same: contracted or regulated cash flows and asset-level security, not the jurisdiction stamped on the loan.
Where the risk actually sits
Infrastructure debt is not risk-free fixed income. The risks are different, and they tend to arrive through the contract and the asset rather than through a sponsor’s income statement.
Construction and completion risk. Lend into a build and your exposure is cost overruns, delays, permitting problems and contractor performance. It is managed with fixed-price engineering contracts, performance bonds, contingency budgets and completion tests that have to be passed before the long-term debt terms fully apply.
Revenue and volume risk. Contracted revenue is powerful but not uniform. An availability-based public-private partnership is a different animal from a merchant power plant selling into spot markets. Even with a contract, you need to understand the offtaker’s credit quality, the termination provisions, the inflation indexation, and what happens if the asset performs below its thresholds.
Regulatory and political risk. Regulated assets are stable until the regime shifts: an allowed-return reset, political intervention in tariffs, a change to what costs can be recovered. This is less about daily volatility than about diversifying across jurisdictions and not assuming today’s regulatory settlement lasts forever.
Refinancing and duration risk. Infrastructure debt is often longer-dated than corporate direct lending. That can be a feature, matching long-term liabilities, but it raises sensitivity to rate moves and refinancing conditions. Amortising structures and conservative maturity profiles reduce how much rides on being able to refinance at exactly the wrong moment.
Valuation and liquidity reality. This is private credit. Liquidity is limited and marks can be model-driven. Fine if you size it as patient capital; a problem if you treat it as a liquid ballast you might need to sell in a hurry.
How to think about it in a portfolio
Infrastructure debt tends to make sense when you want contracted income with structural protection and you can accept lower liquidity for it. It sits alongside corporate private credit rather than replacing it. A workable way to frame the choice:
- If you want smoother cash flows, weight toward operating assets with contracted or regulated revenue and conservative leverage at the asset level.
- If you want higher return, accept that the extra yield comes from construction exposure, complexity or sitting higher up the structure in subordinated tranches. That is paid risk, not free yield.
- If you care most about downside control, spend your diligence on covenants, distribution locks, reserve accounts and step-in rights. In infrastructure debt the outcome is often decided in the documentation.
Access depends on where you sit. Most infrastructure debt is bought directly by institutions through private funds run by managers such as AllianzGI, AXA IM Alts and Macquarie, and those funds are typically closed to smaller investors. If you cannot write an institutional cheque, the practical route is a listed vehicle. On the London Stock Exchange, renewables and infrastructure investment trusts, including The Renewables Infrastructure Group (TRIG), NextEnergy Solar Fund and Greencoat Renewables, own contracted energy assets and use asset-level debt, and their shares trade like any other stock. That is equity exposure to leveraged infrastructure rather than a pure debt position, so read it as a different risk, but it is investable from a normal brokerage account. Whatever the vehicle, its home-country tax (withholding at source, for instance) is a property of the vehicle, while your own tax treatment depends on where you are resident, which is a question for your adviser rather than a rule that travels with the asset.
One structural point sharpens all of this. Private infrastructure debt assets under management passed US$168 billion in 2024, growing at about 14.4% a year since 2015 inside a private credit market that has reached roughly US$3.5 trillion overall (Preqin via Macquarie, 2024; AIMA, 2025). As more capital arrives, the distance between disciplined lenders and generalists reaching for deals widens rather than narrows. The asset class delivers its low loss rates only when the structure is built to earn them, which is why manager selection does most of the work. For the pillar view of the sector, see our infrastructure investing guide.
FAQs
Is infrastructure debt the same as project finance?
Project finance is a common way to structure infrastructure debt, but they are not identical. Project finance usually means non-recourse or limited-recourse lending to an SPV, with repayment tied to the project’s cash flows. Infrastructure debt also covers lending to established infrastructure corporates and to holding companies, where the recourse and the risk are different. What matters is where the cash flows sit and what security the lender actually holds.
Why can infrastructure debt feel less volatile than corporate private credit?
Many infrastructure assets run under long-term contracts or regulated frameworks, which dampen cash flow swings and tend to produce fewer covenant surprises. That showed up in the pandemic, when Moody’s rated infrastructure universe recorded only slightly more defaults than its long-run average. But “less volatile” depends on the segment: construction-period and merchant-revenue deals can be highly sensitive. The stability is earned through revenue design, not the label.
What does “senior secured” mean here?
It means the lender has first-ranking claims on the project’s assets and cash flows and sits at the top of the debt stack. In practice it comes with control rights: cash waterfalls, reserve accounts and restrictions on paying money out to equity. Senior secured status is only as strong as the covenant package and the enforcement mechanics behind it, so read it as a legal position rather than a guarantee.
How do interest rates affect infrastructure debt returns?
Many deals are floating-rate, so coupon income can rise as base rates rise, subject to any caps or borrower hedging. Fixed-rate deals give more predictable cash flows but can look less attractive if rates move sharply after origination. Separately, rate moves feed through to asset valuations and refinancing conditions, which matter more for long-duration assets. Maturity, amortisation and hedging together decide how exposed a given book actually is.
Next read
- Private credit, explained covers how these funds generate returns and where they run into trouble.
- Infrastructure investing maps the wider opportunity set across equity and debt.