Real Assets Report 2011 — The Inflation Hedge That Depends on the Contract

Real assets became a major institutional allocation in this period for a coherent reason: the search for return described in the Global Investment Outlook 2012. With government bonds yielding almost nothing, institutions with fixed obligations needed income from somewhere.

Infrastructure and property offered a plausible answer — long-lived assets producing steady cash flows, with a claim to protect against inflation.

The inflation claim is where the analysis most often went wrong, and the correction is the report's central point.

A real asset does not hedge inflation because it is physical. A road, a pipeline or a building has no intrinsic mechanism that makes its revenue rise with prices. The hedge, where it exists, comes from the terms on which its output is sold:

• An explicit indexation clause in a concession or lease, linking tariffs or rent to a published price index. This is a genuine hedge and it is contractual.
• A regulated return framework that permits price increases reflecting costs. This is a hedge mediated by a regulator's decisions.
• Or pricing power in a competitive market — the ability to raise prices because demand permits it. This is a hedge only where the market structure supports it.

Where none of these applies, the asset does not hedge inflation. An infrastructure asset selling output under a twenty-year fixed-price contract has fixed nominal revenue, which is precisely what a bond has. Its value falls when rates rise, and inflation erodes its real returns exactly as it erodes a bond's.

The second theme is that the capital arriving compressed the returns it was seeking. Yields fell as competition for assets rose, so investors entering later paid more for the same cash flows — a structural feature rather than a mistake.

Where the hedge actually lives

The mechanism deserves precise development, because "real assets hedge inflation" is treated as a property of the asset class and is a property of specific contracts.

Start from what inflation protection requires: the asset's cash flows must rise with the general price level, so that real income is preserved.

Four revenue structures, four different answers:

Contractually indexed revenue. A concession specifying that tariffs rise annually with a published index. This is a direct, enforceable hedge, and its quality depends on which index, whether there are caps or floors, and how frequently it resets.

Regulated revenue on a real return basis. A regulator sets allowed revenue to deliver a specified real return on an asset base that is itself indexed. This hedges inflation through the regulatory formula, and depends entirely on the regulator continuing to apply it.

Market-priced revenue with pricing power. An asset whose output is sold at prevailing market prices, where those prices rise with inflation. This hedges to the extent the market permits, which varies enormously — an airport with capacity constraints has pricing power; a merchant power plant in an oversupplied market does not.

Fixed nominal contracted revenue. An availability payment or fixed-price offtake agreement with no indexation. This provides no hedge at all. The revenue is a nominal annuity.

The question is never "is this a real asset." It is "what index, if any, does the revenue move with, and who has the right to change that." Both answers are in the contract, and neither is in the physical asset.

The property equivalent follows the same logic. A lease with fixed rent for fifteen years is a nominal bond secured on a building. A lease with annual indexation, or with frequent market rent reviews, is a hedge. The building is identical in both cases.

The practical consequence for a portfolio is that an allocation labelled "real assets" may contain a mixture of genuine hedges and long-duration nominal exposures, and the label provides no information about which. The Real Assets Report 2018 develops this into a classification framework.

A long-dated nominal cash flow is a bond

The point deserves its own treatment because it determines interest rate sensitivity, which was the larger risk over the following decade.

Duration measures sensitivity to interest rates, and it is driven by how far in the future the cash flows arrive. An asset producing steady cash flows for thirty years has very long duration — longer than most bonds an institution holds.

So an unindexed infrastructure asset has two properties simultaneously:

• It looks defensive, with stable, contracted, predictable cash flows.
• And it is extremely rate-sensitive, because a small change in the discount rate applied to thirty years of cash flows moves the present value substantially.

These are not in tension. They are the same fact viewed two ways — the stability of the cash flows is exactly what makes the valuation a pure function of the discount rate.

Why the exposure was easy to miss in this period:

• Rates were falling, so the duration exposure was producing gains, which nobody investigates.
• Valuations were appraisal-based and infrequent, per the mark-versus-price analysis in the Private Equity Report 2008 — so the rate sensitivity did not show up as visible volatility.
• And the assets were classified as an alternative rather than as fixed income, so they were not aggregated with the portfolio's other duration exposure.

The last point is the practical failure:

An institution holding thirty-year bonds and unindexed thirty-year infrastructure holds one duration position recorded in two places. The risk report shows diversification because the labels differ.

This is the archive's recurring hidden-concentration finding, arriving through asset classification. The Global Investment Outlook 2022 documents what happened when the rate cycle turned and both legs repriced together.

When a regulator sets your return

Regulated assets are a large share of the infrastructure universe, and their return determination is unlike anything else in an institutional portfolio.

The typical framework: the regulator determines allowed revenue for a period — commonly five years — calculated to deliver a specified return on a defined asset base, after permitted operating costs.

What this means for the investor:

• The return is set administratively, not by the market. The regulator's methodology is the investment thesis.
• The allowed return is periodically reset, so there is a scheduled repricing event that is essentially political.
• Outperformance is possible by beating the cost assumptions, and is frequently shared with customers by design.
• And the asset base is defined by rules about what investment qualifies, which determines how growth translates into returns.

The risks are correspondingly unusual:

• Methodology change. A regulator revising how the allowed return is calculated changes the asset's value with no operational cause.
• Political pressure. Regulated assets provide essential services at prices the public experiences directly, so returns are politically visible in a way most investments are not. A period of high consumer prices makes utility returns a live political issue.
• And the reset is scheduled, so the risk has a known date — which is unusual and should be priced explicitly.

The analytical implication:

Underwriting a regulated asset means underwriting a regulator. The relevant research is the methodology, the consultation record and the political environment, not the asset's engineering.

The counterpart is that the framework's transparency is genuinely valuable — allowed returns are published, methodologies are consulted on publicly, and determinations are documented. This is more disclosure than most private assets offer, and it is free.

Illiquidity as a feature

Real assets are extremely illiquid, and how an investor should regard that depends entirely on their own liabilities.

Why these assets are hard to sell:

• Transactions are large, limiting the buyer universe.
• Diligence is slow and expensive — technical, environmental, legal and regulatory.
• Transfers frequently need consent from regulators, concession grantors or partners.
• And there is no continuous market. Sales are negotiated processes taking many months.

For an investor with genuinely long liabilities — a pension fund with obligations decades out — this is close to costless. They were never going to sell, and being paid an illiquidity premium for accepting a constraint that does not bind is straightforwardly good.

For an investor whose horizon is shorter or less certain, it is a serious risk. The Global Investment Outlook 2020 shows what happens when an institution needs liquidity and holds assets it cannot sell.

The structural point:

The illiquidity premium is a payment for a constraint. It is free money for an investor whom the constraint does not bind and an expensive trap for one it does. The same asset is priced identically for both.

The complication is that liability duration is often less certain than assumed. An institution's horizon depends on its funding position, its sponsor's health, and its members' behaviour — all of which can change. The Private Markets Outlook 2016 covers how allocation frameworks addressed this.

Buying the yield away

The final structural theme is one that applies to any asset class entering an institutional allocation wave.

The sequence:

1. An asset class offers attractive yields relative to alternatives, for a genuine structural reason.
2. Capital allocates toward it, attracted by exactly that.
3. Competition for a limited stock of assets raises prices, which lowers yields.
4. Later entrants pay more for the same cash flows and earn less.
5. Eventually the yield advantage that justified the allocation has been competed away, while the illiquidity and complexity remain.

Why the supply cannot respond quickly:

• The stock of existing infrastructure is fixed in the short run, and only a fraction is available for sale.
• New assets take years to build and carry construction risk that most institutional buyers do not want.
• And privatisation programmes, which release assets, are politically determined rather than responsive to investor demand.

So demand rose against near-fixed supply, which produces price rather than volume — the same structure as the property analysis in the Asia-Pacific Investment Report 2010.

The practical consequences:

• Entry timing mattered enormously. Assets acquired early in the allocation wave were bought at yields later entrants could not obtain.
• The definition of the category expanded to include assets with more risk, since the core assets had become expensive — which is the standard response to compressed returns and the point at which the original thesis stops applying.
• And leverage was used to maintain target returns on assets whose unlevered yields had fallen, which reintroduces exactly the rate sensitivity the asset class was supposed to avoid.

The Real Assets Report 2018 and Global Investment Outlook 2022 cover where this ended up.

Volume risk is the one that actually varies

Inflation and rates dominated the discussion; the risk that most often determined outcomes was how much the asset was used.

Infrastructure revenue has two components — a price per unit and the number of units — and the risk in each is completely different.

Assets with no volume risk:

• Availability-based structures, where the operator is paid for keeping the asset available regardless of use. A road paid by availability earns the same whether one car or a million use it.
• Regulated networks with revenue caps, where allowed revenue is fixed and tariffs adjust to deliver it. Lower volumes produce higher tariffs, so the revenue is protected.
• Take-or-pay contracts, where the offtaker pays for capacity whether or not it is used.

Assets with full volume risk:

• Toll roads paid by traffic.
• Airports paid by passengers.
• Ports paid by throughput.
• Merchant generation paid for output sold.

Why volume risk is systematically underestimated:

• Traffic and passenger forecasts have a well-documented optimism bias, consistently across projects, countries and decades. The forecasts are produced during a process whose purpose is to justify the project.
• The forecast horizon is very long — thirty years or more — so small errors in an assumed growth rate compound enormously.
• And demand is correlated with the economy, which means volume falls exactly when everything else in the portfolio is falling. The asset described as defensive is cyclical at the revenue line.

An availability-paid asset is a contract with a government. A traffic-paid asset is a leveraged bet on regional economic growth over thirty years. They are both called infrastructure and they have almost nothing in common.

The 2008–2011 period tested this directly. Volume-exposed assets saw traffic, passenger and throughput declines that most base cases had not contemplated, while availability-based and regulated assets performed as contracted. The divergence within a single asset class was larger than the divergence between asset classes.

The practical check is to ask what happens to revenue if volumes fall by a quarter. For an availability asset the answer is nothing; for a traffic asset it is a quarter of the revenue against unchanged debt service — which is where the leverage described above becomes decisive.

What an allocator could act on

Read the revenue contract, not the asset description. Inflation protection lives in indexation clauses, regulatory formulas or market pricing power, and an asset with fixed nominal revenue provides none of it.

Aggregate unindexed real assets with the portfolio's bond duration. Long-dated fixed nominal cash flows are a rate exposure regardless of what the underlying asset is made of, and separate labelling conceals a single position.

Underwrite the regulator when buying a regulated asset. Methodology, consultation history and political environment determine the return; the engineering does not.

Match illiquidity to genuine liability duration, and test that duration. The premium is free for an investor the constraint does not bind and a trap for one it does, and horizons are less certain than they look.

Ask when in the allocation wave you are arriving. Fixed supply against rising institutional demand produces price, so later entrants systematically pay more for the same cash flows.

Watch for category expansion and added leverage as return targets bind. Both are responses to compressed yields and both mean the original thesis no longer describes what is being bought.

What 2011 established

• Inflation protection is contractual, not physical, and varies from complete to entirely absent within the same asset class.
• An unindexed long-dated asset is a bond, with the duration exposure hidden by asset classification and appraisal-based valuation.
• Regulated returns depend on a regulator's methodology, with a scheduled, politically-determined reset.
• The illiquidity premium is free or expensive depending entirely on whether the constraint binds the holder.
• Institutional demand against fixed supply compresses yields, so entry timing determines much of the return.