Although there has been a significant amount of work undertaken by both practitioners and academics on the (beneficial) impact of adding listed real estate to an unlisted real estate portfolio (Moss and Farrelly 2013), and indeed the role of real estate in a multi-asset portfolio, there has been little work undertaken thus far on the impact of combining listed and unlisted infrastructure funds.
We believe that now is the time to look at this topic, given the noticeable increase in allocations to infrastructure by both institutional and retail investors over the past five years. This has been driven by a desire from investors for income, inflation-hedging characteristics, a low correlation with other asset classes and, where possible, liquidity. Intuitively, we would expect that listed infrastructure, like REITs can act as both a return and liquidity enhancer when combined with unlisted infrastructure, and also as diversifier in a mixed-asset portfolio.
Bachher et al (2012) describe the asset class as consisting of a heterogeneous collection of core physical assets and essential service businesses that support an economy’s day-to-day functioning. Most private investment in the area has gone into civil infrastructure, which is generally defined to include the communications, energy, transportation, and water and waste sectors. In the UK and Europe, significant private investment has also gone into social infrastructure, including schools, government buildings and healthcare facilities.
The AMP Global Listed Infrastructure team defines the infrastructure investment universe by splitting the industries into four categories, based on risk and reward, with ‘ring 1’ being the ‘purest’. Their definitions are:
• Ring 1. Water utilities, oil and gas storage/transport, transmission and distribution, social infrastructure.
• Ring 2. Toll roads, airports, ports, communications, diversified, and master limited partnerships (MLPs).
• Ring 3. Integrated utilities, contracted generation, postal, railways.
• Ring 4. Construction, engineering, infrastructure services, shipping logistics, and merchant generation.
As can be seen, the ‘purest’ level provides lowest risk. It is important to note that in infrastructure (unlike real estate) it is the risk-adjusted return characteristics that are key to investment decision-making, rather than the underlying asset itself. As such, it is possible to distinguish clearly between positive characteristics inherent in an asset (those in ring 1) and characteristics that increase risk and decrease attractiveness.
In summary, the positive characteristics of infrastructure can be classified as:
• Assets with monopoly-like characteristics;
• Long-term (10 years-plus) contracts;
• Clear inflation linkage;
• Track record;
• High visibility of cash flow.
Similarly, the characteristics that deviate from ‘pure’ infrastructure plays would include:
• Participation in highly competitive markets;
• Significant turnover sensitivity to volume changes in variables such as traffic;
• Significant negative sensitivity to commodity prices;
• Lack of income certainty – for example, greenfield assets;
• Volatile cash flow.
The key required investment characteristics
Unlike other asset classes, investors use infrastructure for a variety of different roles in portfolio management. One of the best ways of understanding the requirements from an infrastructure allocation is to look at the institutional benchmarks for that asset class. Bachher et al (2012) looked at the available benchmark families for the infrastructure asset class and found that investors have different goals, leaving no single ‘right’ way to benchmark the asset class.
In a survey of the benchmarks used by institutional investors, they found that there was not even a consensus on the style of the benchmark. The types used, which by definition reflect the investment characteristics of infrastructure, were:
• Absolute return;
• Inflation based (CPI-plus);
• Inflation plus risk premium (CPI plus bond and equity premiums);
• Fixed-income index;
• Hybrid approaches.
The variety of benchmarks used reflects not only the lack of long-term time series data for the asset class but, more particularly, the different roles infrastructure can be expected to play in portfolio management. Unsurprisingly, different strategic approaches necessitate different benchmarks. Despite the differences, there does appear to be a consensus among institutional investors that infrastructure fits somewhere between regular equities and fixed income on the risk-reward spectrum.
The researchers found that some investors were looking to infrastructure to provide equity-like returns with bond-like risks and serving as a first-order proxy for long-dated liabilities.
In summary we can identify the main required investment characteristics of infrastructure as:
• Attractive risk-adjusted returns, complementing a diversified portfolio;
• Reliable inflation-linked returns;
• Low correlation and volatility compared with traditional asset classes;
• Stable long-term yields, with the potential for capital growth;
• Defensive characteristics emanating from the provision of essential services;
• Potential for value enhancement through active management of the assets.
How listed/unlisted sectors differ
As with real estate, the underlying returns of listed and unlisted infrastructure can be expected to converge over the medium term, but over the shorter term there are obvious structural differences relating to liquidity and volatility. These are summarised in table 1.
There are also, however, differences in terms of the investable universe of listed and unlisted infrastructure, by asset and industry type, whereby listed infrastructure can provide opportunities not available in the unlisted sector, in addition to its structural benefits, such as liquidity and diversification.
Benchmarks that can be used for listed and unlisted infrastructure
In determining whether there are benefits to investors from combining listed and unlisted exposure, we need to identify the most suitable benchmarks. There are a number of possibilities analysed in this article. It should be noted that this study was completed prior to the launch of the IPD Global Infrastructure index.
Unlisted infrastructure indices: An Australian Unlisted Infrastructure index is now available from IPD and goes back to 2001; Preqin Private Infrastructure index. This is global and goes back to 2007.
Unlisted infrastructure fund actual returns: Using Preqin data we are able to group global infrastructure indices returns by vintage
Unlisted infrastructure target returns: Looking at the data provided by Property Funds research we are able to group together target IRR returns to provide an unlisted benchmark. This currently equates to a Net IRR target of 10% to 12%, representing a gross IRR target of 15% to 20%.
Listed infrastructure NAV indices: These indices provide the underlying NAV progression of the constituent members rather than the total return profiles of the units in the Funds.
Listed infrastructure indices: There are a number of benchmarks available, including Brookfield, Macquarie and S&P.
Listed infrastructure fund: Using the Consilia Capital database we can construct a representative sample of listed infrastructure funds returns. Table 2 shows the equal-weighted average of 14 of these funds. Returns are all rebased in US dollars.
It can therefore be seen that there are a wide range of benchmarks that could be used, both absolute and relative, global and domestic, but there are limitations (particularly in terms of the length of time series available ) that investors need to be aware of in all of them.
Which methodology is most appropriate to understand the impact of combining listed and unlisted infrastructure? Given the variety of benchmarks available, and the lack of global comparisons, it is important that we make our analysis as rigorous and explicit possible. We therefore decided to split the study into two sections.
1. Combining global unlisted and global listed. There is no broadly acceptable global unlisted infrastructure index available prior to 2007. Therefore, the initial part of the study will look at the impact of combining listed and unlisted global infrastructure since 2007.
2. Combining domestic unlisted and global listed. There is, however, an IPD-based Australian index available from 2001, so for the second stage we can look at a domestic unlisted portfolio combined with a global listed exposure (which has been done for real estate) over this longer period.
Combining global listed with global unlisted funds since 2007
Turning to the impact of combining listed and unlisted exposure, table 2 shows the initial results. We show initially a 100% unlisted portfolio, then a 100% listed portfolio, and then the impact of combining a 70% unlisted weighting with a 30% listed weighting. The final column shows the difference between this 70/30 blended return and the unlisted return in that calendar year.
What is noticeable is that – with the exception of 2008 where the returns of the unlisted funds were at odds with all liquid asset classes – adding listed infrastructure to unlisted infrastructure has a neutral-to-positive impact on raw returns, before taking into account any benefit of improved liquidity or diversification of risk.
Combining global listed with domestic unlisted funds since 2002
With a longer time series, we can now start to look at the impact of combining listed and unlisted over the cycle. Unlike the previous period, this captures the run up to the GFC and the outperformance of listed in that period.
Undertaking the same analysis as previously, we see again the increased volatility of listed on a quarterly basis, and interestingly the correlation is only 18.6% over the period.
Impact of the cycle on blended returns
Turning finally to the annual changes and the impact of adding listed, we observe the following: in only one year out of the period (2008) would the addition of listed have been a negative factor on absolute performance (table 3).
Applications in practice
Given the positive contribution of adding listed infrastructure to an unlisted portfolio, we believe there are a number of potential applications of this strategy in providing a liquid source of long-term income, which has inflation-hedging properties and low correlation with other asset classes. These include:
• A simple 70/30 fixed blend of global unlisted and global listed infrastructure to enhance risk-adjusted returns and liquidity.
• Similarly, using a fixed blend of domestic unlisted (which typically is easier to access than global) and global listed infrastructure to reduce risk correlations while gaining exposure to the asset class and enhancing liquidity.
• Using an actively managed listed portfolio to provide exposure to asset and industry types not accessible through the unlisted market.
• Varying the weighting of an actively managed listed portfolio from an initial 30% to take account of differing valuations in the listed
and unlisted sector to enhance risk-adjusted returns.
• Using a listed infrastructure portfolio to gain access to the sector pending investment into specific unlisted funds/projects.
Alex Moss is managing director at Consilia Capital