Produced in partnership with EDHEC
Global asset owners are expanding into infrastructure debt in a bid to lock in long-duration income, but the surge of money is also triggering questions about how to quantify risk in an illiquid and inherently opaque market.
The trend sits at the centre of a broader rotation from public to private markets as investors hunt for new sources of alpha and diversification. Infrastructure assets under management grew at a 19.7 per cent compound annual rate over the decade to 2024, with infrastructure debt expanding even faster at 23.1 per cent.
The shift is now visible in the growing number and size of mandates.
In September 2025, Dutch pension giant APG announced its first infrastructure debt allocation, awarding €425 million to an impact-focused mandate to Schroders Capital, and in May 2025, UK workplace scheme NEST seeded a new Europe-focused infrastructure debt fund run by IFM – in which it holds a stake – with an initial commitment of about €530 million.
Infrastructure debt has outperformed corporate and liquid bond peers at similar durations (yielding 4.9–5.1 per cent), with lower volatility (4.5–5 per cent) and with higher risk-return ratios (up to 0.98), according to data provider Scientific Infra & Private Assets (SIPA).
It is these type of risk-return characteristics that make private assets well suited to generating reliable income, according to NEST Invest chief executive Mark Fawcett.
“Private credit, infrastructure, etc – if you go to the right part of that asset class, you can generate the income [and] generate a superior pension income,” he said at the recent Fiduciary Investors Symposium at the University of Oxford.
Behind the veil of private infrastructure credit
However, the challenge of measuring risk and return becomes particularly pronounced within private market sub-sectors such as infrastructure debt.
Most infrastructure credit is project finance and so assessed by ratings agencies. It pushes investors towards an imperfect workaround: taking a listed bond proxy and adding an illiquidity premium.
“There is no like-for-like listed market for infrastructure credit – you just don’t have public bonds formed from wind farms,” says Abhishek Gupta, head of product at Scientific Infra & Private Assets (SIPA), the commercial arm of the EDHEC Infrastructure & Private Assets Research Institute.
Investment grade infrastructure debt and corporate bonds have a long-term monthly return correlation of 0.8-0.9 given shared macro drivers such as interest rates and credit cycles, but there are also periods when their performance decouples. For example, in the non-investment-grade segment, the 12-month rolling return correlation between infrastructure debt and corporate bonds turned negative between mid-2018 and late-2021, illustrating periods of decoupled performance despite exposure to shared macroeconomic drivers.
Another issue with benchmarking is the illiquidity premium itself, which is usually treated as a static number rather than a dynamic variable.
“This illiquidity premia also changes over time during the market cycles depending how active the market is,” Gupta says.
It can lead to fundamental risks, despite infrastructure debt consistently demonstrating overall low default probabilities.
“If you don’t assess the risk properly, you can’t estimate spreads in this market as effectively. That’s a problem if you’re not valuing infrastructure credit on your books fairly to the market.”
An assessment of private debt ratings using InfraMetrics® classifications compared against public credit ratings (such as S&P) for the same issuing entities where both private and public instruments exist, shows a consistent alignment over time.
On average, 13 per cent of private debt instruments classified as investment-grade by public agencies were assigned a non-investment-grade label by InfraMetrics® over the past five years.
Notably, in 2023 and 2024, alignment between the two rating approaches reached 100 per cent, coinciding with an observed decline in overall credit risk following the pandemic period.
A heterogeneous asset class
Between 2019 and 2023, the average one-year probability of default (PD) for infrastructure debt declined, stabilising around 1 to 1.3 per cent, while recovery rates have consistently remained above 75 per cent, according to SIPA. Yields are attractive at 4-5 per cent compared to riskier private infrastructure equity at 6-8 per cent.
But the distribution of risk is lumpy depending on lifecycle, leverage, and sector, according to Riazul Islam, senior quantitative researcher at SIPA.
“Assessing infrastructure credit risk requires a granular and forward-looking approach that captures both project-specific characteristics and broader macroeconomic conditions. The InfraMetrics® model incorporates time-varying factors such as the debt service coverage ratio (DSCR), leverage, cash flow available for debt service (CFADS), firm age, and benchmark interest rates (e.g., the risk-free rate). These inputs allow for dynamic recalibration of credit risk as both firm-level fundamentals and external environments evolve. This factor-based framework enables nuanced, data-driven evaluation of creditworthiness and supports informed decisions on restructuring scenarios or credit classification adjustments over time.”
Globally, energy and water resources have shown the highest average probability of default, reflecting commodity cycles, resource constraints, and higher operational volatility (1.91 per cent), while network utilities with monopoly-like models posted the lowest (0.19 per cent).
Young firms (0-5 years) were most vulnerable, with default probabilities around 1.8 per cent – often due to construction risks and lack of operational history – with mature stage firms dropping to just 0.6 per cent.
Accurate data is the key to finding new investment opportunities, accurately valuing existing investments, and to underpin capital market assumptions, Islam says.
“Infrastructure debt can be managed not as an illiquid, opaque niche, but as a mainstream credit allocation with measurable risk and return characteristics.”
More granular, project-level data is allowing investors to distinguish between fundamentally different sources of risk across lifecycle stages, sectors and capital structures, rather than treating infrastructure debt as a homogenous allocation.
