Part 2: Infra performance benchmarks wanting

Related resources

Part 2: Infra performance benchmarks wanting

4 minutes
May 29, 2019 11:30 pm

A version of this blog was first published on

The EDHECinfra/G20 survey of infrastructure benchmarking practices included representatives of 130 asset owners accounting for $10 trillion AUM. It revealed that that existing performance monitoring benchmarks are self-defeating for asset owners and managers. But we have made great progress on tacking this problem at EDHECInfra. We have been creating more representative, better defined infrastructure benchmarks.

This is the second in a three-part series examining the results of the survey.

Infrastructure investment requires customised performance benchmarking

Performance-monitoring benchmarks differ from the asset-allocation benchmarks. They need to represent actual investment choices made when implementing a fund’s investment policy.

In the case of infrastructure, the difference between policy and performance-monitoring benchmarks is all the more significant. The ability to implement any given style or tilt is itself uncertain. Why? Well, infrastructure markets are notoriously illiquid and in part driven by public procurement and other policy decisions that are not easily predicted.

The implementation of a broad policy allocation to infrastructure may take multiple incarnations. Different levels of geo-economic, industrial, or business-risk exposures frequently require dedicated sub-allocations. Often these can be fully gauged only after the fact. For instance, Investors need a high degree of specialist industrial knowledge within any infrastructure sector. This usually militates for individual sub-strategies or mandates.

Perhaps even more importantly, building large, well-diversified positions in any segment of the unlisted-infrastructure space remains difficult today, given the average time and size of individual transactions.

As a direct result, while policy benchmarks focus on long-term rewarded risks, performance-monitoring benchmarks may require being tailored to an investor’s or their manager’s actual portfolio. Moreover, achieving sufficient granularity is very important to benchmark the investments made fairly and accurately.

The core-satellite approach

As discussed in the first part of this series, in a core-satellite context (no relation with ‘core infrastructure’), investors can monitor and manage the performance of asset managers and investment teams. They achieve this by defining a core portfolio, representative of the expected behaviour of a given investment style or strategy. Then they can define a satellite portfolio in terms of its tracking error relative to the core.

In the case of highly illiquid asset classes like unlisted infrastructure in which a well-defined ‘core’ is not directly investible, this distinction remains valid. It still gives investors a way to monitor the dual objective given to asset managers. They aim to deliver the core strategy (deal by deal) and to outperform the average as captured by the core benchmark.

An implementation of this approach to monitoring unlisted infrastructure managers can make use of the tracking error given to a manager as a representation of the construction of the infrastructure portfolio. The younger the portfolio, the larger the tracking error. As a portfolio of infrastructure debt or equity increases in size and representativeness, the tracking error should be reduced to only represent the space within which the manager can deliver alpha.

Investors acknowledge serious issues with infrastructure performance monitoring benchmarks

In the 2019 EDHECinfra/G20 survey, 50% of respondents declared using the same benchmarks for performance monitoring as they do for strategic asset allocation. Some 75% of infrastructure equity investors reported using absolute benchmarks for performance monitoring.

In light of the comments above, this is highly problematic. Absolute benchmarks can be a good indicator of target returns. However, to monitor performance adequately, investors should use a benchmark that represents their choice(s) of investment policy explicitly defined in terms of risk profile.

In effect, the practices described by investors in this survey correspond more to the definition of a hurdle rate rather than a benchmark.

  • About 70% of respondents acknowledged that the benchmarks they use for performance monitoring do not allow investors to measure risk-adjusted performance.
  • When we posed the same question to asset owners only, more than 75% of respondents reported similar concerns.
  • Almost 40% of respondents also agreed that the use of another asset class as a proxy for unlisted infrastructure equity is a challenge.
  • Close to 30% of respondents also acknowledged that current private benchmarks tend to report smoothed returns i.e. not to capture risk exposures.
  • Around 30% of asset owners also said that current industry-peer, money-weighted benchmarks do not allow for a fair comparison of asset managers. Indeed, such indices are sensitive to the timing of cash flows. Such flows can vary across fund managers and can even be manipulated to achieve higher returns.

In this survey, nobody liked their performance monitoring benchmark. All of them wanted to use a more representative, better defined benchmark. They want one that is truly informed by actual market movements and risk factors.

Better benchmarks can reconcile asset owners and managers over infrastructure investment

As in other alternative asset classes, asset owners and managers have long been at odds when it comes to demonstrating value and performance with illiquid asset classes like infrastructure. Managers say they have better deal making skills, while asset owners point to their high fees.

This has pushed a number of large asset owners to internalise infrastructure investment decisions. They believe they would be better off investing directly in infrastructure than through third parties.

However, as our survey demonstrates, asset owners still do not know where they stand when it comes to infrastructure investing. The problem is that they do not use representative, mark-to-market benchmarks. The ones they have available fail to take their actual strategy and risk exposures into account. Their internal investment team continue to report the same ill-suited metrics they used to get from external managers.

In effect, existing benchmark practices are self-defeating for infrastructure asset owners and managers alike.

The right tools open up opportunities

Delivering a portfolio of unlisted infrastructure asset in line with a given mandate is hard work. It requires skills, time and commitment, whether executed through a manager or internally. As discussed above, delivering the ‘core’ portfolio (as in ‘core-satellite’) creates value and justifies rewarding managers or investment teams.

Furthermore, the lumpiness and average size of infrastructure assets can add complexity. Even a reasonably well-delivered core portfolio still offers opportunities to create alpha through asset selection and operation.

Hence, a customised benchmark that would represent a well-defined investment policy or strategy would help investors. They could better determine the value created by the manager or team that creates this portfolio. They could also measure any outperformance relative to this strategy. In this context, managers can demonstrate the value they create twice. They achieve this first by creating a portfolio that tracks a well-defined benchmark and then by outperforming it.

The EDHECinfra indices released quarterly from June 2019 onwards can help achieve the objective of better defining and benchmarking individual investors goals and asset managers value creation in unlisted infrastructure investments.