Types (% in AUS$) (% in AUS$) Risk
Social 9–11 4–12 Medium
Regulated 11–12 6–10 Low
Rail 12–13 8–12 Medium
Airports/Ports 11–13 5–10 Medium
Power Generators 12–14 4–12 High
Toll Roads/Greenfield 13–15 3–5 Medium/High
Note:IRR=internal rate of return.
Source:AMP Capital (2014b)
variety of assets found within a single sector. They suggest that the identity of the sector or the fact that a market is regulated provides sufficient evidence to determine the corridor for the expected risk-return profile of an asset or project. This is a severe mistake, which is still commonly made by (inexperienced) investors. Strictly speaking, the sector alone does not allow any conclusions to be drawn about the risk-return profile of an asset. An analysis of the individual case is always necessary. Also, a regulated market cannot be automatically equated with low risk. This is particularly apparent in the regulated telecommunications market, which embodies an extremely high level of market risk. There may also be a variety of regulations within the same sector across geopolitical boundaries (e.g. high-voltage electrical transmission, gas networks, rail transport networks).
Instead, the contractual structure has a major impact on the risk-return profile of an asset/project. This is because each contractual structure with its various contract partners represents a unique risk profile comprising risk factors pertaining to, among others, rev- enue and cost, credit, construction and operations, including maintenance, as well as politi- cal and regulatory risk, and has its unique risk allocation among the contractually involved parties.
Hence, it is the combination of the aforementioned influencing factors and characteristics – and the contractual structure in particular – that ultimately determines the risk-return profile of an investment. Therefore, it is possible for investments with simi- lar physical asset characteristics to appear identical on the surface (e.g. two road or power plant projects) yet to have entirely different risk-return profiles resulting from their underlying contractual structure and risk allocation (Weber, 2009). This is illustrated in Figure 2.1.
Figure 2.1 aims to demonstrate, once again, that the risk-return profile of an infrastructure asset is not predominantly determined by the sector but that it depends in large part on the geography, stage and contractual structure in which it is embedded and the risks that the private partners take on. As such, similar physical assets in a sector can deliver an IRR ranging from around 5 to well above 15%. The case with the least risk, categorised as ‘I. Operational – Availability/FIT based – No market risk’ in Figure 2.1 represents, for example, an operational asset with an availability-based PPP structure, not too highly leveraged, in which a recognised public body in a politically and fiscally stable country is the contractual partner of the private parties. As a consequence, the private sector takes little or no market (demand and price) risk.
From day one, the asset generates a long, stable and predictable cash flow for the duration of the contract period, which is reduced only if either the operator is not able to maintain
5 10 15 20
5–9%
5–11%
7–12%
9–14%
11–16%
12% plus Expected nominal gross return (%)
I. Operational Availability or FIT based No market risk II. Operational No or some market risk*
III. Greenfield No or some market risk
IV. Operational Market risk*
V. Greenfield Market risk
VI. Other non-regulated
– Assets with regulated, long-term availability based PPP or FIT contracts – no/minimal market risk (demand/volume or price) – Revenues/yield from day one
– Grantor of revenue is a recognised public body in stable country – No low or medium leverage
– Essential services, monopolies or negligible competition but no 20+ year contracts – some market risk (e.g. contract renewal) – Revenues/yield from day one
– Regulated by stable government regime – No low or medium leverage
– Same type of assets as under I. or II. but with development or construction risk (whereby little premium is paid for construction risk of, e.g., wind or solar parks)
– No/negligible yield at date of investment
– Monopoly character, regulated, but exposed to market risk (e.g.
partially/entirely user financed)
– Revenues yield from day one – possibly high leverage – Regulatory risk owing to periodic regulatory changes and/or
(possibly) unstable government (in emerging markets) – Same type of assets as under IV. but with development and
construction risk
– No or negligible cash flows at date of investment – Possibly high leverage
– Significant market/demand risk (user financed), no regulation – Assets may be operational or still under development/construction – Possibly emerging market countries
– Possibly high leverage
* not an availability based PFI/PPP or clean energy FIT based asset/project
F I G U R E 2 . 1 Drivers of assets’ risk and return profiles Source:B Capital Partners
and operate the asset as contractually agreed or the operating costs are higher than projected because, for example, the wear and tear of the asset dramatically increases – both rather manageable risks.
An asset classified as ‘IV. Operational – Market risk’, on the other hand, exhibits a riskier profile, because although the asset is operational and regulated it is exposed to market risk. In most instances, market risk is the single biggest risk for a private concessioner/owner followed by political/regulatory risk. Therefore, while these two types of assets are both operational – that is brownfield, regulated, and may look identical from the outside – their risk-return profiles differ significantly (see also Weber (2009) for further elaboration).
An additional, frequently used measure of risk is that of debt default rates. Infrastructure investments generally tend to benefit from low debt default rates. Moody’s has analysed and compared investment-grade5 cumulative default rates of corporate infrastructure6 debt vs.
non-financial corporate (NFC) debt issuers from 1983 to 2012 (see Figure 2.2). Although default rates were initially similar, cumulative default rates for infrastructure tend to level off as a project matures, whereas NFC cumulative defaults continue to increase, being around 50% greater than those for infrastructure debt after 10 years.
5The vast majority of corporate infrastructure debt (81%) is investment grade.
6The main sub-sectors, by sample size, within corporate infrastructure debt are: regulated utilities (63%) and unregulated utilities (9%) of corporate infrastructure, remaining other utilities (19%) and remaining other corporate infrastructure or non-utilities corporate infrastructure (8%).
F I G U R E 2 . 2 Investment-grade cumulative default rates, 1983-2012 Source:Moody’s (2012)
One often-cited reason for the superior performance of infrastructure over corporate bonds is the fact that banks analyse infrastructure bonds, which are based on project finance principles, more rigorously and cautiously than corporate bonds. That is, they have to base their credit decision primarily on the future cash flows of the (new) project company rather than being able to rely on the assets of a corporate issuer, with which they consider themselves to be familiar and/or have a longstanding relationship, which needs to be honoured (see also Section 6.1). Also, certain infrastructure project companies may be less exposed to economic fluctuations than ‘regular’ corporates that are (more) exposed to market risk, making them relatively crisis-resistant (e.g. natural monopolies). Last but maybe not least, average corporate infrastructure debt trading prices tend to be around 40–45% higher than those for NFC issuers, suggesting that the former may be ‘overpriced’ in the first place, in the sense of not fully reflecting all the risks of the corporates.
In any case, these results, which are based on long data series, are clearly an argument in favour of investments in infrastructure, be it in the form of equity or debt.
Moreover, when infrastructure debt does default, the recovery rates are much higher compared to those for corporate debt – with the exception of senior unsecured debt of unregu- lated utilities (Table 2.4) – confirming the sounder (credit) risk profile of project finance debt underlying most infrastructure transactions (see Chapter 6).
TA B L E 2 . 4 Recovery rates for defaulted corporate infrastructure debt, 1983–2012
Sector Senior Secured Senior Unsecured
Regulated utilities $82.52 $59.16
Unregulated utilities $60.96 $41.45
Others $65.93 $60.05
Average corporate infrastructure debt $68.72 $53.01
Average non-financial corporate issuers $49.30 $36.50
Source:Moody’s (2012)
Supporting the above, infrastructure debt has a lower volatility than corporate debt, as depicted in Figure 2.3, which compares notch-weighted rating volatility.7
Notch Weighted Volatility 0.8
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 Year
99 00 01 02 03 04 05 06 07 08 09 10 11 12 Non-Financial Corporate Issuers (upper line) Total Infrastructure Securities (lower line)
F I G U R E 2 . 3 Rating volatility for total infrastructure securities and non-financial corporate issuers Source:Moody’s (2014)
When comparing the performance of infrastructure against other asset classes, the lack of established performance benchmarks for infrastructure investments does not help. Given the risk-return bandwidth of infrastructure and the well-known general difficulties of bench- marking unlisted assets, this is not surprising. Defining the right benchmark depends heavily on factors relating both to the asset/liability profile of the investor and, of course, the invest- ment goals pursued with the infrastructure investment strategy (e.g. yield or capital gain, see also Section 1.3.8). The next section elaborates on possible approaches to benchmarking the performance of infrastructure assets, being well aware, though, that this is still early days.
2 . 1 . 3 B e n c h m a r k i n g i n f r a s t r u c t u r e i n v e s t m e n t s8
Institutional financial investors use benchmarks to compare the financial performance of their investments with the market. Accordingly, a benchmark can be defined as the relevant market with which the performance of an investment is compared. Typically, a financial benchmark is an index of similar listed securities. In the case of listed infrastructure assets, the indices dis- played in Table 2.2 in Section 2.1.2.1 may serve as suitable benchmarks for investors depending on their specific investments. Such indices, however, are of limited use for the performance measurement of unlisted infrastructure assets. As, for reasons mentioned in the previous sec- tion as well as the next paragraph, indices based on standardised, actual performance data of unlisted infrastructure assets do not (yet) exist. Therefore, investors lack industry standards for benchmarking and monitoring both the financial and sustainability performance of their
7Notch-weighted volatility: this metric is constructed by adding the notch-weighted downgrade rate and upgrade rate together. These metrics are computed by taking the number of notches (e.g. alphanumeric rating movement of a credit from Aaa1 to Aaa2) and dividing it by the number of outstanding ratings.
Because of this construction, the volatility rate can at times exceed 100%.
8Unless otherwise mentioned, sources for this section on benchmarking include: Weber B. (2013), EDHEC (2014), EDHEC (2013) and CFA Institute (2012).
infrastructure assets. In the absence of such standards, a range of approximate financial bench- marks is used by some investors, and yet none by others. The first sustainability benchmarks are just in the process of being developed. They are touched upon in Section 2.2.3.2.
2 . 1 . 3 . 1 S t r u c t u r i n g / d e v e l o p i n g a b e n c h m a r k In theory, there are various ways to structure a benchmark for unlisted infrastructure. In practice, though, investors are confronted with multiple challenges, of which the three most important ones seem to be that (1) bench- marks require a reliable data universe to work with, (2) benchmarks need to fulfil certain criteria, among which are replicability and investability and (3) returns need to be attributable.
First, insufficient data are due to (i) little and short publicly available performance data series (e.g. only Australia has performance data dating back over 15 years for unlisted infrastructure), (ii) a broad investment universe within infrastructure with an equally broad range of invest- ment strategies (comprising a wide range of different risk-return profiles, which would need to be itemised in order to compare the performance of individual investments/assets) and (iii) a lack of coherent performance measurements by investors. Second, replicability/investability (a typical requirement of a benchmark) is an equally difficult condition to meet because the above-mentioned insufficient data universe of publicly available, tradable assets make it chal- lenging, if not impossible, to construct a replicable, meaningful index/benchmark. Third, a classic return attribution to active management relative to an industry benchmark is merely impossible for unlisted infrastructure assets because infrastructure investment styles are often broad and the quantification of benchmarks subjective as well as backward looking.
2 . 1 . 3 . 2 B e n c h m a r k s c u r r e n t l y i n u s e Out of necessity, many financial investors use a simple, straightforward approach and benchmark their investments against absolute return expectations and/or inflation (CPI) plus a margin.9
However, some investors pursue a more sophisticated approach that considers overall asset allocation goals, investment strategy, risk profiles of underlying assets, performance expectations, etc., when specifying a benchmark. Table 2.5 illustrates selected benchmarks as used by selected financial investors.
2 . 1 . 3 . 3 A s t r a t e g i c a p p r o a c h t o b e n c h m a r k i n g While absolute performance and CPI- related infrastructure benchmarks are convenient and easy to use, they may be of little value or even misleading if investors select them based on the wrong parameters, for example physical characteristics rather than the risk-return profiles of the individual assets.
Professional investors eventually need to take a strategic, more informed approach to benchmark selection. A strategic approach to benchmarking infrastructure assets requires the specification of a benchmark, which is appropriate, i.e. that is able to measure if the main investment goals (e.g. return target given a certain risk profile) of the investment strategy are met. This means benchmarks need to match the investment strategy with its individual risk-return specifications. The physical characteristics of the asset itself are of minor impor- tance(see Section 2.1.2.3).
9In the early days of unlisted infrastructure investing, institutional investors often applied listed infrastruc- ture indices as benchmarks mainly because of a lack of alternatives. Nowadays, none of the experienced infrastructure investors does so. See also Section 2.1.4 for a discussion of correlations between listed and unlisted infrastructure.
TA B L E 2 . 5 Examples of benchmarks used in the industry
Investor Benchmark
Insurance companies
Australian insurance company Absolute return (8%)+coverage of internal management costs
German insurance (predominantly life insurance)
Government bond+margin (5%) South European Insurance (diversified) Absolute return (6.5%) unlevered, post tax Pension plans
British Columbia Investment Management Corporation (bcIMC)
8% absolute return with adjustments for asset, country, and currency risks
Caisse de D´epˆot et Placement du Qu´ebec 50% S&P 500/TSX+25% S&P 500+25%
MSCI EAFE Index California Public Employees’ Retirement
System (CalPERS)
CPI+margin (5%)
A Canadian public pension Government bond+margin (4–5%)
CPP Investment Board (CPP) Calculated on an investment-by-investment basis European Public Pension CPI (Harmonised Index of Consumer Prices
[HICP])+margin (n/a) Municipal Employees’ Retirement System
(MERS) of Michigan
Barclays Aggregate Bond Index OPSEU Pension Trust (OPTrust) CPI+margin (5%)
Ontario Teachers’ Pension Plan (OTPP) CPI+margin (4%)+sovereign spread (where CPI is based on country and currency of investment)
PSP Investments CPI+bond return+equity premium (inflation
adjusted infrastructure risk premium and infrastructure cost of capital – 2014 annual report) UK Public Pension Fund CPI (Retail Price Index (RPI))+margin (5%)
(used across all alternative assets) Borealis Infrastructure (invests on behalf of
OMERS and other institutional investors)
Absolute return set at the beginning of the year based on operating plan
Note:The authors do not know whether the institutions listed are still using the indicated benchmark by the time this book is published.
Source:Weber B. (2013) and CFA Institute (2012)
Benchmarking infrastructure assets, like any other asset, can be viewed as a three-step process:
i. Define the infrastructure investment strategy, specifically its desired target risk-return profile(s); it is possible for one infrastructure strategy to have two or three different sub-strategies with distinct risk classes and related return targets, if required or more suitable;
ii. Select a suitable benchmark and specify margin if appropriate – do so for every risk category, if applicable;
iii. Measure the performance of the infrastructure asset – after having assigned it to the respective risk category, if applicable – against the relevant benchmark.
i. Define the infrastructure investment strategy Investors ideally develop their infras- tructure investment strategy based on their overall long-term investment strategy and asset allocation. To this end, they are advised to consider the relevant investment criteria in the context of infrastructure investing and the related questions listed in Table 2.6. These criteria may help investors to define their infrastructure investment strategy and, most importantly, to carve out the related risk-return profile(s). Investors are encouraged to address and form an opinion about each of the investment criteria in order for their infrastructure investment strategy to be sufficiently precise while leaving enough flexibility (i.e. a large enough universe) for its actual implementation (i.e. the asset picking).
In addition to the risk and return profiles, reflected in the selection of certain countries, sectors, stages of entry, currencies and the like, the cash flow profiles of any particular strategy are usually of paramount importance for investors and are closely linked to their investment horizon. Essentially, investors can be categorised into (primarily) yield-driven investors (‘yield investors’) and IRR-driven investors (‘IRR investors’), see Section 1.3.8 for further information.
TA B L E 2 . 6 Infrastructure investment criteria – questions and considerations Investment Criteria Questions/Considerations
Risk profile Which kinds and scales of risks am I able or prepared to take?
Return profile Given the risk profile, what are the short and long-term return expectations for both annual running yield and IRR?
Cash flow profile Timing of cash flows
Preferred cash flows (e.g. capital gains at exit vs. annual yield)
Interest vs. dividend payments (consider tax and/or regulatory implications) Investment horizon Long-term vs. short-term holding period
Inflation protection Relevance of real vs. nominal returns, which is predominantly driven by elasticity of demand and structure of investors’ liabilities (e.g. pension fund vs. life insurance vs. reinsurance liabilities)
ESG criteria (Industry specific) exclusion criteria (e.g. military) to be applied, if any Stages of
investments
To be considered from the point of view of both the cash flow profile (greenfield generally has no immediate cash flow) and/or the risk profile (greenfield has additional risks)
Currencies Hedging costs
Leverage To be considered from both a risk and a return point of view Geographies Political and regulatory risks, currency exposure
Liquidity If liquidity is of utmost importance, unlisted infrastructure may be the wrong investment approach to start with
Diversification Specify across which aspects/categories of risk and for which purpose Control of assets Kind of control usually required for listed investments
Which control rights, if any, for which purpose?
Under which circumstances are more control rights required and of which kind?
When addressing these questions and considerations and when specifying the investment criteria, investors may want to pay attention to and/or have a clear opinion about the fol- lowing selected, seemingly sophisticated but possibly important, issues, which are frequently overlooked:
Post tax vs. pre-tax returns:equity returns in the form of dividends or capital gains tend to be measured after tax, while bond returns (interest payments) are discussed based on their coupon value. For tax-exempt investors (as many are, e.g. pension funds), a distinct pre- and post-tax consideration is not relevant. For the majority of investors, however, it is.
Leverage:equity returns are mostly geared, while debt returns are not.
Inflation protection of debt vs. equity: the vast majority of debt – infrastructure and corporate alike – is not inflation linked. Equity returns of infrastructure assets, in contrast, very often are (even up to 100%), hence providing a significant protection against inflation risk, which debt does not.
Absolute vs. relative return targets:the structure of the liabilities of an investor usually determines its desired/required profile of future returns – absolute vs. relative. Insurance companies, for instance, typically need to lock in absolute returns because in their, for example, life insurance contracts with customers they commit to absolute amounts to be paid out. In a possibly deflationary environment, inflation-linked returns would backfire.
The liabilities of pension funds, that is the pensions, in contrast, are linked to inflation, which can go up or down. Therefore, pension funds should have a strong preference for inflation-linked returns. In a deflationary environment, their liabilities would decrease accordingly.
Country risk: depending on the country risk, a higher or lower return for a certain investment may be appropriate. In case the country in which an investment is located differs from that of the investor, it needs to be decided which country determines the benchmark. Is it the geographic location of the asset or of the investor? Good arguments for both may be possible.
ii. Select a suitable benchmark and margin Once the most important parameters of the investment strategy have been defined, a suitable benchmark must be identified or developed that captures the most relevant characteristics of the investment strategy. If appropriate, in addition a margin needs to be specified that reflects the risk level.
Investors may have creative and unusual ideas regarding benchmarks. As a starting point, possible benchmarks are listed below, which may be more or less suitable for individual investors, depending on the infrastructure strategy they pursue:
Selected benchmarks
Absolute return
CPI+margin
Real estate (index)+margin
Bond yield+margin
Peer group
Listed infrastructure+margin
Listed equity+margin
Asset mix+margin
For additional information on these benchmarks, please refer to Table 2.7.