Infrastructure and inflation protection - how does it work?

Inflation can impact the performance of an asset through several channels illustrated here by the simplified cash flow statement of an infrastructure asset (Figure 4)

In this above example, we apply a 5% inflation increase. For a traditional core infrastructure asset, there is a corresponding 5% increase in revenue as contracted CPI adjustments are made. As we shall discuss further, not all infrastructure assets have an inflation pass through mechanism as direct as this example.

Similarly, operating expenditures (Opex) in this model also increase at the same rate but its impact is limited, as infrastructure assets typically benefit from high operating margins. For example, the leading 65 toll road operators globally boasted an average operating margin of 85.4% in 2019.[3] As such, if labour costs, which represent 50% of the above toll road operators’ average costs move +/-5%, the cash impact would be marginal, and the operating margin remain unchanged.

The generally high leverage nature of infrastructure assets implies that changes in interest rates in response to rising inflation would have a significant negative impact on returns. However, there are several potential mitigants for investors. The first would be the timing effects of hedging, loan structuring and duration meaning that in the case of short-term changes in inflation, there may be no actual impact on debt servicing. The other two factors, which we will discuss later in this diary are potential increases in regulated returns in the case of longer-term inflation and the impact of discount rates on valuation.

Of course, these various mechanisms do not flow in a similar pattern across each infrastructure asset. The following table (Figure 5) illustrates the basic categories under which most infrastructure investors group the range of infrastructure strategies; Core, Core-Plus/Value Add, and Opportunistic. This framework provides a useful overview of the varying risk characteristics of these groups but, in terms of understanding the inflation revenue recovery mechanisms for individual asset types, there can be considerable overlap in characteristics as they apply to the traditional Core, Core-Plus/Value Add, and Opportunistic model.

We will take an extra step to understand how each of the cash flow items react in an inflationary shock. In general terms, there are 3 inflation protection mechanisms as illustrated in Figure 6.

• Contracted adjustments, typical of regulated or concession assets where the offtake party will commit to a mechanical adjustment around CPI through a pre-agreed formula. Assets with highly contracted revenues and generally the most sought after infrastructure assets given the predictability, if not stability of their cash flows. These types of assets are also generally described as core infrastructure.

• Monopolistic positions, where an infrastructure asset has an implicit pricing power to adjust prices to inflation across shorter-term contracts of 3 – 7 years. These assets are usually core-plus, or value add assets with examples including airports, broadcast towers, or midstream transmission assets through geographic monopolies, and finally,

• GDP-linked assets, generally identified as either expanding and existing platforms exploiting new technology or industry growth. Examples would include data centers, fibre networks, or some midstream horizontal drilling assets. Contracts here are often short term (2-3 years) but with strong user demand and low substitution risk, the operator generally sees strong implicit pricing power and a low customer churn rate.

Out of the three revenue characteristics, contractual agreements of Core infrastructure are considered the strongest. For all of these protections, we also note that they do not operate like inflation linked bonds, which may adjust overnight to inflation movements. Here even contractual agreements may take a year or more to be reflected in revenues.

Digital Infrastructure is an interesting example of a GDP-linked opportunistic asset as it operates in both economies with a lower level of GDP per capita, as well as the higher GDP per capita, but more mature, economies. Figure 7 shows the percentage of internet users per a given GDP level. The green trend line shows the strong linear relationship between internet penetration and GDP growth which provides strong pricing power protection to telco and digital infrastructure operators. The orange trend line, mostly OECD countries, shows this this relationship becoming weaker and even asymptotic and average incomes increase. However, these modern economies live in an age of data, where digitalization is a key driver of economic growth. In these markets it is not so much the number of users, but the volume of data of these users and businesses demand that is driving GDP growth, ensuring that in a scenario where inflation surges due to increased economic activity, digital infra will be able to increase revenues commensurately, either as a cause or consequence of that GDP growth.

Toll roads represent another interesting sector to study as they can seen at the intersection of having long term contracted, monopolistic and GDP-linked revenue structures. If we take the example of Transurban, the Australian listed toll road operator, we understand that 81% of its toll road contracts are CPI-linked or fully uncapped, with the small remainder being generously capped (Figure 8). The toll road segments are monopolistic in the contest of the travel time savings that drivers achieve relative to public roads. Finally, as we show in Figure 9, a strong GDP-linkage with a low price elasticity of demand.

While toll roads carry attractive pricing agreements, the asset owner carries demand risk. Despite this, toll roads usage appears to be mostly inelastic to toll increases, due to a combination of site-specific factors, time savings that toll roads offer, and broad acceptance of e-tolls. These assets also have a strong positive correlation to GDP growth. In the case of Transurban, we see the impact of the monopolistic element that most of these roads have as the manager almost doubled the toll price from AUD3.00 to AUD6.70 in the last two decades while traffic volume has more than doubled over the same period, very much in line with GDP (Figure 9).

Cost Protection

A common characteristic across many infrastructure assets is high economies of scale. The cost of one more car on a toll road is extremely low and generally results in high operating margins. For example, in our earlier toll road example (which noted an average operating margin of 85.4%), a rise in labour costs of 5%, or a similar increase in crude related asphalt costs, would be unlikely to have a meaningful impact on margins and, given the nature of the mostly CPI linked concessions, would quickly be recaptured.

Similar to the way that revenue protection varies across infrastructure assets under the earlier Contractual, Monopoly and GDP-linked classifications, the Inflation Recovery Mechanisms model also differs across infrastructure types for changes in costs. In the short to medium term, the response of infrastructure assets (and real assets in general) to inflation will critically depend on the underlying contractual nature of the asset. This will dictate the degree to which asset owners are able to pass through inflation via respective higher charges, tolls or utility rates. The underlying contracts or regulatory structures will influence what items may be linked to inflation (revenues, expenses, capital) and under what constraints (timing of adjustments, caps, other limitations). The following examples look at how margins can be preserved in inflationary periods in Monopoly and GDP-linked model.

Monopoly Assets

These models cover a range of assets including PPP’s and Availability /Social Infrastructure which may cover assets such as schools, hospitals, and other public buildings. A government entity enters into a long-term agreement (25-40 years) with an infrastructure manager to design, build, operate and manage facilities. Payments to the infrastructure manager are based on the infrastructure manager’s availability to meet the contractual performance standards. The infrastructure manager will structure its long-term contract around various sub-contracted service providers that will seek to maximize cost pass through.

The operating cash flow profile of these assets tend to be similar to contractual and regulated utilities in that revenue payments and cost recoveries over the life of the agreement are contracted in advance and incorporate inflation pass-throughs and a correlation to changes in CPI of close to 100%. For example, cost overruns at a construction or facilities management contractors are generally absorbed by the contractor under the terms of their contracts (and are expected to incorporate these risks in the earlier pricing for which the infrastructure manager has priced its overall bid). A recent purchase by Stafford’s SISF IV of a UK fund in which availability projects and PPPs made up 91% of the portfolio showed a 68% inflation sensitivity, i.e. a 1% increase in inflation is expected to result in a 0.68% increase in net IRR for the Fund. This is before the positive addition impact that high cash balances have from an associated increase in interest rates.

GDP-linked Assets

The Venn diagram in Figure 6 illustrates the range of assets that exist within and across these revenue models. Generally, the degree of margin (cost) protection is recognized as being strongest for assets that have pricing and or returns explicitly protected by contracts or through regulation. For GDP-linked assets, where margin protections are from implicit pricing power, investors need to closely analyse the relevant contract terms and a reliance its market strength.

In the case of GDP-linked assets such as midstream gathering and processing or in broadcast and cell phone towers, inflation protection from contracts or a perceived monopoly may not be as important as underlying sector or GDP growth.

In the case of US midstream operators, recent oil and gas price volatility, particularly the collapse on the oil price in early 2020 substantially weaken a number of upstream operators. As evidenced in the following chart (Figure 10), the US onshore rig count, used as a forward indicator of midstream throughput, has as a result been slow to recover at the same pace as the oil price, suggesting midstream operators remain in a weak position to enforce volume commitments or address inflation pressures.

In the case of cell phone towers, we recently noted industry commentary suggesting a weakening in tower operator pricing power. Telco mobile network operators (MNO’s) are the customer base of the private cell tower companies. Tower access lease costs have recently run up in the face of costs involved with roll out of 5G technology. Implied leasing costs have grown from under 5% of the major MNO’s wireless revenues to ~7% in 2020. In the face of flattening mobile revenues, cell phone towers operators are challenged to recover costs of the 5G transition having become a notable cost item for the MNO’s.[4]

Interest Rate Impact in Infrastructure

Infrastructure assets, which use high levels of debt to support a high level of long term fixed capital, should notionally see rising interest rates as a catastrophic risk. However, from a practical perspective, there are a three mitigants to interest rate risk that an experienced infrastructure manager can use to protect returns.

1. Capital modeling and asset refinancing

2. Changes in regulated and concession-based discount rates

3. Discount rates in valuations.

In the same way that revenue structures vary between core and opportunistic infrastructure in our earlier model, debt structures also vary across the broad spectrum of infrastructure assets in the way these assets structure the debt size, tenor and hedging to align with the underlying contract length and stability of cash flows.

1. Capital modeling and asset refinancing

As noted in Figure 11, infrastructure managers will look to match the leverage and debt maturities against the profile of the assets revenues (such as contract length and GDP sensitivity). While Opportunistic assets would appear to be most exposed to interest rate and refinancing risk, low interest rates and strong credit availability in recent years has given most infrastructure manager the ability to structure debt with a combination of asset refinancing and interest rate hedging. As a result, in a period of short-term inflation, most of the above groups should be a net beneficiary as revenues rise faster than operating and credit costs.

2. Changes in regulated and concession-based discount rates

Across Europe, North America, the UK and Australia, regulated utility equity returns are generally determined by the regulator using a formula that applies a defined premium on the regulated asset base relative to a weighted cost of capital (WACC). For regulated assets such as water or electricity distribution, these returns are subject to a defined pricing period, usually from 3 to 5 years. This format also forms a similar basis for a number of PPP and concession contract structures with government related counterparties but with longer duration periods.

In a period of a sustained increase in inflation, a meaningful increase in the risk-free rate (the basis of the WACC calculation) will have a flow on effect too increase revenues from the regulated asset following the WACC reset within the regulatory pricing period, usually 12 months.

3. Discount rates in valuations.

Investors in infrastructure funds benefit from external valuations of the fund assets in the annual accounts. The core methodology of these valuations is a discounted cash flow (DCF)[5], again, using a discount rate which generally comprises government bond rate plus an asset specific risk premium.

The potential negative impact of rising interest rates pushing up the discount rate is generally offset by the impact of smoothing generally adopted within the independent valuation process

The second factor is the general practice of “normalizing” of risk-free rate assumptions in independent valuations. Post financial crisis, the use of long-term averages for the risk-free rate has become the norm to both smooth out the effect of early post GFC volatility and offset the “abnormally” low government bond rates created by recent quantitate easing policies from the major central banks. This has created a substantial buffer of approximately 50-100bps[6] between the normalized and the spot risk free rate and in turn, a substantial buffer for valuations in a period of gradual increases in interest rates.

Valuation experts, Duff & Phelps (rebranded as Kroll) has published its recommended risk-free rate (RfR) along with its corresponding U.S. Equity Risk Premium since 2008. In June 2020 it lowered its normalized USD RfR estimate from 3% to 2.5% (also noting the same movement for CAD, and GBP). At the time of the adjustment through to December 2021, the normalized 2.5% represented at 100bps premium to the U.S. Treasuries at 20-year constant maturity. By March 2022, that rate had moved in line with the comparable Treasury rate. However, noting a spike in 20-year breakeven inflation rate from 1.5% in June 2020 to 2.8% in March 2022, it will be interesting to see if and additional normalization to the RfR is issued.[7]

[3] “Open Opportunity – a global benchmark of toll operator efficiency”, KPMG International, 2019
[4] Confidential consultant report associated with recent transaction.
[5] Complimented with additional data such as recent comparable transactions and peer valuations
[6] Stafford estimate
[7] The breakeven inflation rate represents a measure of expected inflation derived from 20-Year Treasury Constant Maturity Securities and 20-Year Treasury Inflation-Indexed Constant Maturity Securities The latest value implies what market participants expect inflation to be in the next 20 years, on average. Sourced from fred.stlouisfed.org