Infrastructure investing: equity and debt, revenue models, valuation and research

In depth12 min read · 11 sources

Infrastructure investing finances roads, power grids, fiber networks or hospitals, either as equity (infrastructure equity) or as loans (infrastructure debt). It is considered stable and protected against inflation. Here you will learn which revenue models lie behind it, how core and value-add assets differ, how to value an infrastructure company with a DCF, why interest rates 100 basis points higher can cut its value by around 11%, and what the research leaves of the promises.

What counts as infrastructure

Infrastructure consists of assets that tie up a lot of capital and run for decades; competitors can hardly replicate them. This creates the typical profile: a high initial investment, low running costs, often a natural monopoly and therefore regulation or long-term contracts. The usual classification is:

• Transport: toll roads, bridges, tunnels, ports, airports, railways and stations.
• Energy and networks: electricity and gas grids, storage, pipelines, power plants, wind and solar farms.
• Utilities: water, wastewater and waste.
• Digital infrastructure: fiber networks, cell towers and data centers.
• Social infrastructure: schools, hospitals or administrative buildings, often built and operated as a public-private partnership (PPP).

For investors, more important than the sector is where the money comes from: from the regulator, from a contract with the state, or from users, who may also stay away.

Revenue models: who bears the risk?

A current example of a regulated return comes from Germany: for the fifth regulatory period of the gas grids (2028 to 2032), the Bundesnetzagentur (2026) (Germany's Federal Network Agency) proposes setting the cost of capital as a weighted average of the cost of equity and the cost of debt, with a flat 40% equity and 60% debt:

The less demand risk, the calmer the cash flow. According to Moody's (2020), social infrastructure projects, which are mostly paid for through availability payments by public-sector offtakers, had the lowest cumulative default rate after ten years, at 1.1% (broad Basel definition).

Core to opportunistic, brownfield and greenfield

Fund managers classify infrastructure by risk. The boundaries are not standardized; the classification is an industry convention (core to opportunistic):

• Core: existing assets that are regulated or secured by long-term contracts, with little demand risk, e.g. grids or PPP projects. Most of the return comes from ongoing distributions.
• Core-plus: similar assets with more market or demand risk or a need for expansion, e.g. ports or airports.
• Value-add: assets whose value is meant to rise through expansion, conversion or better management; more of the return comes from capital growth.
• Opportunistic: new construction in new markets, new technologies or turnaround cases. This is close to private equity.

Cutting across this is the construction stage (brownfield and greenfield). Brownfield means the asset is already running and has a revenue history. Greenfield means it still has to be built, with construction, cost and schedule risk. Moody's (2020) shows why this matters: the annual default rates of project loans fall as projects age and reach the level of A-rated companies around the seventh year. Differences by country and use largely disappear once a project has an operating history.

Infrastructure equity: funds and listed stocks

There are three ways to get equity exposure to infrastructure. Closed-end funds collect capital commitments, invest over a few years and sell the holdings before the end of the fund's term. They come with a J-curve and fees like private equity. Open-ended and semi-liquid funds hold assets with no fixed end date and accept only limited redemptions. Listed infrastructure stocks (utilities and operators of toll roads, airports, cell towers or pipelines) and funds that hold them can be traded daily, but fluctuate with the stock market.

The reputation as a stable investment only partly fits the funds. Andonov et al. (2021) analyzed the cash flows of institutional investors: infrastructure funds fluctuate about as much and are about as cyclical as other private equity funds, and their success depends just as much on quick exits. On a risk-adjusted basis they performed poorly; public investors such as state pension funds achieved less than private ones. For public investors, according to the study, preferences for sustainability (ESG) and regulation explain 25 to 40% of the increased allocations and 30% of the underperformance.

Inflation protection is not automatic either. Using data on 824 listed infrastructure companies from 46 countries over 37 years, Rödel and Rothballer (2012) found that they hedge inflation just as well or as badly as other stocks; only companies with strong pricing power did slightly better over five years. For the USA from 1991 to 2013, Wurstbauer and Schäfers (2015) find only a partial hedge in the short term, and only for direct, unlisted infrastructure, but in the long term all the series examined move in line with inflation.

Infrastructure debt: project finance, senior and mezzanine

Debt is usually provided as project finance: a special purpose vehicle builds and operates the asset, and lenders are repaid only from its cash flows, without recourse to the owners' other assets. The debt is tiered. Senior loans are serviced first and are usually secured; mezzanine or subordinated loans come next and demand higher interest; equity gets whatever is left. Protection comes from covenants: if the debt service coverage ratio falls below a threshold, no distributions flow to the owners (lock-up), surplus cash pays down the loan faster (cash sweep), and reserve accounts cover a few installments in advance. More on lending away from the exchanges in the article private credit.

How often project finance loans default is shown by the database of Moody's (2020), which covers bank loans since 1983:

• Defaults: in the analysis of 7,354 projects from power, infrastructure and oil and gas, the cumulative default rate after ten years was 5.0% under the broad Basel definition and 3.2% under Moody's own definition. For infrastructure in the narrower sense (mainly transport, social and environmental projects), Moody's gives 4.2% (Basel).
• Recoveries after default: the average ultimate recovery rate was 81.5% (Basel); the median was 100% for most subgroups, i.e. no economic loss. However, the distribution has fat tails: around 9.3% of cases recovered only 0 to 24%.
• Comparison: in the European Economic Area and the OECD, the default rates correspond to those of companies with a low investment-grade rating. According to Moody's, however, the definitions are not directly comparable.

Worked example: DCF valuation and 100 basis points more

The value of an infrastructure company is the present value of its future cash flows. If they grow at a fixed rate over a limited term, such as a concession, the formula for a growing annuity applies:

Worked example: debt service coverage ratio (DSCR)

Valuation: discount rate, the turn in interest rates and smoothing

Unlisted infrastructure has no market price. Valuers usually estimate its value with a DCF model like the one above. The discount rate is the most sensitive assumption, and it is adjusted only slowly: estimated values follow the market with a lag and in muted form. This smoothing, and how to correct it with desmoothing and a public proxy, is explained in the article measuring private markets. It can be seen clearly in real estate in 2022 and 2023 (in the article real estate as an investment).

Regulated grids add a second lag: the regulator adjusts the allowed return only from one period to the next. In its draft for the gas grids from 2028, the Bundesnetzagentur shortens the period from which it derives the risk-free rate to five years, citing the turn in interest rates in 2022 (Bundesnetzagentur 2026). For your portfolio, this means: the volatility and correlation of unlisted infrastructure look lower on paper than the economic risk is. The stable cash flows are real; the stable valuations only partly so.

Regulation: relief for European insurers

Under Solvency II, the European Union's capital regime for insurers, insurers must hold own funds against market risks. For qualifying infrastructure investments (projects with predictable cash flows that can also withstand stress, and a contractual framework that protects investors), the European Commission lowered the requirements in 2015 (European Commission 2015): the shock for equity fell from 49% to 30%, and for a 20-year bond of credit quality step 3 from 30% to 20%. At the time, EU insurers held around €22 billion in infrastructure, less than 0.3% of their investments.

On the recommendation of the supervisor (EIOPA 2016), a category for qualifying infrastructure corporates was added. Their equity is now subject to a shock of 36% plus 92% of the symmetric adjustment, compared with 39% plus the full adjustment for listed type 1 equities (Solvency II Delegated Regulation 2015). These rules do not apply to private investors.

For advanced readers: duration, inflation and expected loss

Interest rate sensitivity is measured by duration. In the DCF example, the present-value-weighted average term of the cash flows (Macaulay duration) is 12.68 years; divided by 1.06, this gives a modified duration of 11.96. That is more than for the ten-year bond. For a perpetuity it is 1 / (r − g), here 25. Linearly, that would be a 25% loss for 100 basis points; because of convexity, it is 20%.

For inflation, what counts is the gap r − g. If the nominal interest rate rises because expected inflation rises, and the cash flows grow by the same amount through indexation, the gap stays the same. For the perpetuity, the value then remains exactly unchanged; for the concession in the example, it falls by only 0.51%. If, on the other hand, the real interest rate rises, infrastructure loses value like a long bond. This explains the mixed findings of studies on inflation protection.

For credit risk, the default rate and the recovery rate give the expected loss EL = PD × (1 − Recovery Rate). With Moody's figures (a 5.0% cumulative default rate after ten years and an 81.5% recovery rate), this gives a loss of roughly 0.93% over ten years, i.e. about 0.09% a year. This is a back-of-the-envelope calculation: recoveries often arrive only years after the default, and individual cases end in heavy losses. Blanc-Brude and Hasan (2016) and Blanc-Brude et al. (2018) therefore estimate the risk structurally from the path of the DSCR, the repayment schedule and the covenants, even when hardly any defaults have been observed.

What the research shows

• Funds are not a bond substitute: their cash flows resemble private equity, and on a risk-adjusted basis they perform poorly (Andonov et al. 2021).
• Listed infrastructure is not an asset class of its own: defined by sector, infrastructure stocks do not improve existing portfolios, according to Blanc-Brude et al. (2017). Defined by business model (regulated or contractually secured), a profile emerges that has improved the efficient frontier since the financial crisis.
• Inflation protection is limited: hardly better than for equities (Rödel and Rothballer 2012), most likely with direct infrastructure (Wurstbauer and Schäfers 2015).
• Credit risk falls with operating time (Moody's 2020). Blanc-Brude et al. (2018) calibrate the DSCR on 267 European projects over 17 years and consider the rating agencies' models for private loans to be biased.

Limitations and common mistakes

• One label, many risks: a regulated grid and an airport with passenger risk have little in common. Look at the revenue model, not at the word infrastructure.
• Smoothed valuations: low volatility in fund reports is partly a product of the measurement method.
• Underestimated interest rate risk: long cash flows react to the discount rate like long bonds. This also applies with inflation indexation when the real interest rate rises.
• Costs and liquidity: fees, the J-curve and redemption limits eat up part of the premium; a loan without a market can hardly be sold in a crisis.
• Political risk: regulators, concession grantors and legislators can change the rules.

This article explains relationships and research; it is not investment advice and does not recommend any funds.

How Portfolio Lab calculates it

Unlisted infrastructure is not available as a series in the app. Asset classes contains listed proxies as monthly series in euros; fund costs are already reflected in the price: “Listed infrastructure” (@INFRASTRUKTUR) with utilities, toll roads, ports and airports, plus “Energy infrastructure (MLPs)” (@INFRA_PIPELINES), “Renewable energy” (@INFRA_ERNEUERBARE), “Water” (@WASSER) and, for digital infrastructure, “REITs data centers & cell towers” (@REITS_RECHENZENTREN).

These series show the market risk of listed infrastructure, with its full volatility and its closeness to the stock market. What they do not show: the J-curve, redemption limits and smoothed valuations of unlisted funds, or the credit risk of project finance. In the look-through analysis, the app counts infrastructure funds as a whole toward the category “Infrastructure & natural resources”, even if their holdings list consists of equities. Where the series come from and how they are chained is described in the asset classes methodology.

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.