2.6 Infrastructure Debt and Equity — Who Invests and Why
Roads, power lines and pipelines attract pension funds, insurers and sovereign funds, because contracted or regulated assets pay long, predictable cash flows that are often linked to inflation. How bond-like an asset behaves depends on who sets its income: a contract, a regulator or its users. The more the income depends on how many people use it, the more it behaves like a business.
Why it matters: Check who sets the revenue before assuming an infrastructure asset is safe.
Summary: Infrastructure attracts pension funds, insurers and sovereign funds because contracted or regulated assets pay long, predictable, often inflation-linked cash flows. How bond-like an asset is depends on who sets its revenue: a contract, a regulator or its users.
- A US regulated utility earns revenue = r × rate base + expenses + depreciation + taxes, reset at each rate case.
- With 50% debt, a one-point cut in the allowed return cut equity income by 22% in the worked example.
- Moody’s found a 5.6% ten-year default rate and 77.5% recovery for project finance loans over 1983–2017.
- Project default risk is front-loaded in construction and ramp-up, then falls below A-rated corporate levels.
Infrastructure assets typically generate long-dated, relatively predictable, often contractually fixed or inflation-linked cash flows (toll revenue, a government-guaranteed power purchase agreement) — a distinct risk-return profile from either pure fixed income or pure equities, as introduced in The Practitioner’s Codex Part 8. This makes infrastructure debt and equity particularly attractive to institutional investors with very long time horizons and a need to match long-dated liabilities — pension funds, insurers (this Volume’s Part 1: Insurance & Actuarial Finance), and sovereign wealth funds (The Practitioner’s Codex Part 8) are the dominant capital sources for large infrastructure deals worldwide.
A US regulated network earns revenue set by its state commission: revenue requirement = r × RB + E + D + T, where RB is the rate base (invested capital less accumulated depreciation), r the allowed return, E operating expenses, D depreciation and T taxes (Maryland PSC, 2019). Illustrative: a $1,000M rate base at 7% ($70M) + E $60M + D $40M + T $15M = $185M. The investor owns a claim on the rate base: income grows as approved investment enters the base and changes when a rate case resets r, so the asset trades like a long bond between rate cases and carries policy risk at each reset.
Investors sort infrastructure from core (operating, contracted or regulated) through core-plus to value-add and opportunistic (construction, expansion or market-price risk). Brownfield (operating) assets suit insurers and pension funds matching long liabilities (Part 1: Insurance & Actuarial Finance); greenfield assets carry completion risk and pay more.
What the default data show. Moody’s study of project finance bank loans from 1983 to 2017 found a 10-year cumulative default rate of 5.6%, an average ultimate recovery of 77.5% and a first-year marginal default rate of 1.3% that falls as projects leave construction and ramp-up; after about seven years the marginal rate is below that of A-rated corporate debt (Moody’s, 2019). Expected loss is roughly default probability × (1 − recovery): 5.6% × 22.5% ≈ 1.3% of principal over ten years. The shape matters more than the average: project risk is front-loaded, so a loan that survives completion becomes better credit every year. That is one reason long-horizon investors often buy operating project debt or refinance construction loans rather than lend from day one, and why construction lenders charge more for the same asset.
Before calling an infrastructure asset “bond-like”, ask who sets its revenue. If a contract with a creditworthy counterparty sets it, analyze credit; if a regulator resets it, analyze the reset; if users choose it, analyze it as equity. Then check inflation links: contractual indexing protects directly, while a US original-cost rate base passes inflation through only at the next rate case. Ignore the “core” label if large new capital spending or a contract expiry falls within your holding period.
Ignoring how leverage magnifies a regulatory reset. Fund the utility above with $500M of debt at 5% ($25M interest) and $500M of equity, ignoring tax. At a 7% allowed return equity earns ($70M − $25M) ÷ $500M = 9.0%. Cut the allowed return to 6% and revenue falls $10M to $175M; equity earns ($60M − $25M) ÷ $500M = 7.0%, a 22% cut in equity income from one regulatory decision. Model the next rate case, not only the current one.
Why do pension funds and insurers invest in infrastructure?
Their pension and annuity obligations run for decades, so they want long-dated assets with predictable cash flows to match them. Contracted or regulated infrastructure offers that, often inflation-linked, plus extra return for illiquidity, which they can bear because their own payouts are largely predictable.
Is infrastructure a good inflation hedge?
Partly, depending on the contract. Many toll concessions and availability contracts index payments to inflation. A US utility on an original-cost rate base recovers inflation only through later rate cases. Fixed nominal contracts offer no hedge, and rising interest rates can lower the value of all of them.
What is the difference between infrastructure debt and equity?
Debt receives fixed interest and principal first, protected by covenants, reserves and coverage tests such as DSCR. Equity receives what is left after debt service and lock-up tests, so it bears traffic, price, cost and regulatory risk along with the upside. Debt investors focus on the downside case; equity prices the whole range.
What is the difference between brownfield and greenfield infrastructure?
Brownfield assets already operate and have a revenue history; greenfield assets are still to be built. Greenfield investors bear construction, completion and ramp-up risk, which Moody’s data show is where most project defaults occur, so they demand higher returns. Many investors buy brownfield stakes after completion instead.
Infrastructure suits long-horizon investors because contracted or regulated assets pay long, predictable cash flows, but how bond-like an asset is depends on who sets its revenue: a contract, a regulator resetting the allowed return, or users. Default data show project risk concentrated in construction and ramp-up, with operating project debt performing like strong investment-grade credit.
Four questions on this chapter. Decide on your answer first, then click “Reveal Answer.”
1. A utility has a $500M rate base, an 8% allowed return, $30M of operating expenses, $20M of depreciation and $8M of taxes. What is its revenue requirement?
- $540M
- $90M
- $98M
- $58M
Reveal Answer
Answer: C. Revenue requirement = r × RB + E + D + T = $40M + $30M + $20M + $8M = $98M.
2. That utility is funded with $250M of debt at 5% and $250M of equity, ignoring tax. A rate case cuts the allowed return from 8% to 7%. What happens to the return on equity?
- It falls from 11.0% to 10.0%
- It falls from 11.0% to 9.0%
- It falls from 8.0% to 7.0%
- It falls from 16.0% to 14.0%
Reveal Answer
Answer: B. Equity earns ($40M − $12.5M) ÷ $250M = 11.0% before and ($35M − $12.5M) ÷ $250M = 9.0% after: leverage turns a one-point cut into two.
3. What pattern did Moody’s find in project finance loan default rates over time?
- Lowest early, then rising steadily as the assets age
- Highest in the final few years before loan maturity
- Roughly flat across the whole life of the loan
- Highest early, then falling after construction and ramp-up
Reveal Answer
Answer: D. First-year marginal defaults were 1.3%, and after about seven years the marginal rate fell below that of A-rated corporate debt.
4. A toll bridge competes with free routes, and drivers choose whether to pay. How should an investor analyze its cash flow?
- As equity, because revenue depends on demand
- As a sovereign bond, because the asset is public
- As credit, because a contract fixes revenue
- As a regulated asset with periodic resets
Reveal Answer
Answer: A. When users decide the revenue, there is no contract or regulator standing behind it, so it carries equity-like demand risk.
2.7 Public-Private Partnerships (PPPs)
In a public-private partnership, the government hires a private group to design, build, finance and often run public infrastructure for decades. The group is paid by user fees or by availability payments from the government. The government hands over construction and operating risk, but pays more for private finance than it would for its own borrowing. So it is worth doing only when the value of the risk it transfers is bigger than that extra cost.
Why it matters: A PPP only makes sense when the risk handed over is worth more than the extra finance cost.
Summary: A public-private partnership hires a private consortium to design, build, finance and often operate public infrastructure for decades, paid by user fees or by government availability payments. It transfers construction and operating risk at the price of more expensive private finance, so it is worth doing only when the risk transfer is worth more than that premium.
- Availability payments leave demand risk with the government; toll concessions move it to investors.
- In the worked example, transferring demand risk needs more than twice the equity and defaults at a 28.1% traffic shortfall.
- The UK NAO found PFI delivered more projects on time but at investor returns 2–4 points above government borrowing.
- The Indiana Toll Road concession, bought for $3.8 billion in 2006, filed for Chapter 11 in 2014.
A Public-Private Partnership (PPP) is a specific project finance structure in which a government contracts a private consortium to design, build, finance, and often operate a piece of public infrastructure — a toll road, a hospital, a school — over a long-term concession period, in exchange for either toll/user-fee revenue or a stream of availability-based payments from the government itself. The core rationale: transferring construction and operating risk to a private party with a direct financial incentive to deliver on time, on budget, and to a high operating standard, in exchange for the government paying (directly or through user fees) over the life of the concession rather than funding the entire upfront capital cost itself.
A $260M road expects $40M of tolls and $10M of operating costs a year (CFADS $30M); debt runs 15 years at 6%. Under an availability payment, the government pays a fixed $40M a year while the road is open to standard, keeps the tolls and so keeps demand risk, the risk that usage falls short. Under a demand-risk concession, the private party keeps tolls and traffic risk. Illustrative sizing: availability at 1.30× (debt $224.1M, equity $35.9M); demand risk at 1.60× (debt $182.1M, equity $77.9M, debt service $18.75M).
| Traffic vs forecast | Tolls | Demand-risk DSCR | Availability DSCR | Government net |
|---|---|---|---|---|
| −30% | $28M | 0.96× (default) | 1.30× | −$12M a year |
| Base | $40M | 1.60× | 1.30× | $0 |
| +20% | $48M | 2.03× | 1.30× | +$8M a year |
Flip points: with fixed costs, the demand-risk project hits a 1.20× lock-up at a traffic fall of 18.75% and defaults at 28.1%. The availability project ignores traffic; the government takes the swing. Transferring demand risk needs more than twice the equity.
Governments test a PPP for value for money: its risk-adjusted lifetime cost against a public sector comparator, the estimated cost of the same asset built with government borrowing. Private finance costs more, so the PPP wins only if risk transfer and cost discipline are worth more than the extra financing cost.
PPP structures shift financing risk off a government’s immediate balance sheet, but they do not eliminate the underlying cost — the government (or its citizens, through user fees) still ultimately pays for the asset over the concession’s life, typically at a somewhat higher total cost than direct public financing would have required, in exchange for the risk transfer and, often, faster delivery than the public sector could achieve directly.
For: the UK National Audit Office’s census of 37 PFI projects found 76% delivered on time or early and 78% with no construction price increase after contract award, against a 1999 government survey in which 70% of conventional projects were late and 73% over budget (NAO, 2003).
Against: the NAO’s 2018 review found PFI investors were forecast to earn 2–4 points above government borrowing costs, a group of PF2 schools cost about 40% more than government borrowing would have, and HM Treasury had collected no outturn data to measure the claimed benefits (NAO, 2018). The UK ended PFI for new projects in its October 2018 Budget. The IMF warns that PPPs let governments invest without reporting spending or debt up front, and that crystallized PPP contingent liabilities cost about 1.2% of GDP on average over 1990–2014 (IMF, 2018).
What survives: PPPs tend to build more reliably but finance more expensively, and the published data rarely show whether the trade pays. The evidence is mostly British; US evidence is thinner.
Give each risk to the party best able to control it. If neither party can influence usage, keep demand risk public and pay by availability; if the private party sets prices and service quality (a terminal, a parking concession), demand risk can go private. Accept a PPP only if it passes value for money with the extra financing cost counted. Ignore the rule where the government cannot credibly commit to pay for 25–30 years; then availability payments are only as good as its credit.
Pricing a demand-risk concession on hopeful traffic forecasts. Above, a 28.1% shortfall turns a 1.60× project into a default. In the real case, a consortium paid $3.8 billion in 2006 for a 75-year lease of the Indiana Toll Road; its concession company filed for Chapter 11 in September 2014 as traffic fell short (Indianapolis Business Journal). The state had been paid up front; the investors and lenders took the loss. Bid on a downside traffic case.
What is the difference between an availability payment and a toll concession?
Under an availability payment, the government pays a fixed annual sum while the asset is open and meets standards, and keeps the tolls and traffic risk. Under a toll concession, the private party collects tolls and bears traffic risk. Availability deals need less equity and cheaper debt.
Are PPPs cheaper than public financing?
Rarely in financing cost: private capital cost 2–4 points more than government borrowing in the UK NAO’s 2018 review. A PPP can still cost less overall if risk transfer and lifecycle discipline cut overruns and maintenance by more than that premium, but this is project-specific and hard to verify.
Why do governments use PPPs?
To transfer construction and operating risk, tie payment to performance and spread a large capital cost over the asset’s life. A weaker reason is accounting: a PPP can keep the commitment off the reported deficit and debt at the start, which the IMF warns can undermine fiscal rules.
A PPP transfers construction and operating risk to a private consortium paid by user fees or availability payments over a long concession, and availability structures leave demand risk with the government while toll concessions move it to investors. The evidence suggests PPPs build more reliably but finance more expensively, so each deal must pass a value-for-money test with the extra financing cost counted.
Four questions on this chapter. Decide on your answer first, then click “Reveal Answer.”
1. A toll road collects $50M a year with $15M of operating costs and $20M of debt service. If traffic falls 25% and costs are fixed, what is its DSCR?
- About 0.88×
- About 1.31×
- About 1.13×
- About 1.75×
Reveal Answer
Answer: C. Tolls fall to $37.5M, CFADS to $37.5M − $15M = $22.5M, and DSCR to $22.5M ÷ $20M ≈ 1.13×, down from 1.75×.
2. Under an availability-payment PPP, who bears the risk that usage falls short of forecast?
- The private consortium’s equity holders
- The construction contractor and its insurers
- The project’s senior lenders and bondholders
- The government that makes the payments
Reveal Answer
Answer: D. The government pays a fixed sum while the asset is open to standard and keeps the user fees, so it keeps demand risk.
3. When does a PPP pass a value-for-money test?
- When its risk-adjusted lifetime cost beats the public sector comparator
- When its private financing rate is below the government’s borrowing rate
- When the private consortium earns at least its target equity return
- When it keeps the project debt off the government’s reported balance sheet
Reveal Answer
Answer: A. The comparison is with the estimated cost of building the same asset with government borrowing, including the risks the government would keep.
4. What did the UK National Audit Office’s 2018 review find about the cost of private finance in PFI deals?
- Private finance was cheaper than government borrowing
- Investor returns ran 2–4 points above government borrowing costs
- Investor returns ran about half a point above government borrowing
- Investor returns ran 8–10 points above government borrowing
Reveal Answer
Answer: B. The NAO also found a group of PF2 schools cost about 40% more than government borrowing would have, and no outturn data on the claimed benefits.
- Maryland PSC, Ratemaking Overview (Jan 10, 2019) — Revenue requirement = r(RB) + E + D + T
- IMF, How to Control the Fiscal Costs of Public-Private Partnerships (Oct 2018) — PPP contingent liabilities averaging 1.2% of GDP, 1990–2014