Infrastructure investors should abandon absolute return benchmarks: Lessons from the Covid-19 lockdowns

Published:  June 2020
Author(s):
Download the Paper
Image
Share:
This paper argues that there is no reason for investors in unlisted infrastructure to continue using absolute return or ‘cash +’ benchmarks. It calls for investors to abandon them and adopt market-relative benchmarks based on fair value and representative data.

Summary

From the viewpoint of the state of the art in financial research, there is no reason to use absolute return benchmarks: backed by empirical evidence, we argue that infrastructure invest- ments are not market-neutral and that investors in infrastructure cannot escape the fact that the prices they pay for assets are formed in a market, where systematic risk exists. In effect, it is for this very reason that we can speak of an infrastructure asset class. However unlisted and illiquid, infras- tructure asset prices and returns, just like other listed or unlisted assets, are first made of beta.

In some cases, unlisted infrastructure investments can also deliver alpha.

To demonstrate the added value for investors and managers alike to switch to benchmark- relative infrastructure investing, we build peer- group portfolios of large asset owners and large asset managers and compare their performance to that of the broader market.

We show that investors can use market bench- marks to understand their performance during the Covid-19 lockdown episode meaningfully, but also to measure out-performance over the long run. With absolute return benchmarks, all investors will under-perform in 2020, which is just as uninformative as the pre-Covid-19 situation, when all infrastructure investors beat their absolute return benchmark every year.

We find that the two main peer groups of infrastructure investors have actually been able to deliver alpha and as such have significantly outperformed the unlisted infrastructure equity market as a whole.

This alpha, however, only exists relative to a market beta. It is illusory to pretend that it can exist by itself. And the only way to measure it is to use a market benchmark.

In practice, absolute return benchmarks have dominated unlisted infrastructure investment so far because there was little alternative. In spite of their understanding of the usefulness of using a relative benchmark, many investors in unlisted infrastructure have continued to use absolute benchmarks because they considered that available relative benchmark solutions were not adequate.

Indeed, we also show that benchmarks created using appraisals or listed proxies proved to be so unrepresentative, biased and generally inade- quate that investors were left with the option of using absolute return benchmarks as the lesser evil.

However, recent progress has changed this situation and a modern approach to benchmarking unlisted infrastructure portfolios is now

possible. More representative data and mark-to- market asset pricing are possible and have been developed in recent years.

The Covid-19 crisis has drawn attention to the risks to which particular types of infrastructure, like motorways or airports, are exposed. This has led to a stronger demand from investors and regulators to measure the impact of the crisis on the profitability of infrastructure and to better understand the impact of the crisis on different types of infrastructure assets or managers.

Naturally, absolute benchmarks cannot answer such questions since they are indifferent to market events, or the risks to which each segment of the class is expected to be exposed. It is in this context that the use of a fair, repre- sentative, market-relative benchmark for the unlisted infrastructure asset class becomes even more important.

Our results highlight the range of risk-profiles and drivers available to investors in the unlisted infrastructure space.

They also show that the impact of the Covid-19 lockdowns have been very different in different

segments of the sector: returns have been impacted negatively by much lower cash flows in the transport sector, but also by interest rate movements in all sectors and by a varying uptick in the risk premia of each asset depending on their sector and individual characteristics. We detail all these effects in the paper.

We also show that equipped with proper bench- marks, investors can tell their unlisted infrastructure market beta from their portfolio alpha. Using data for the actual investments of two key peer groups (large asset managers and large asset owners) we find substantial market outperformance especially for large asset managers, as well as different impact of the Covid-19 lockdowns for each peer group.

The realisation amongst investors that infrastructure assets represent significant risk exposures and that these should be understood and managed will determine the coming of age of the infrastructure asset class.
For asset owners, a better understanding of the risks related to infrastructure assets:

  • requires documenting the risk exposures created by their infrastructure investments;
  • requires benchmarking performance relative to the market index or customized benchmark that best represents these risks and creates better aligned incentives in terms of fees; and,
  • allows for a better integration of infrastructure assets in the total portfolio, including for asset-liability management purpose.

For asset managers, showing which systematic sources of risks (and returns) their investment strategy embodies can:

  • explain what part of their performance is driven by risk factors within or beyond their control;
  • demonstrate their ability to deliver access to a well-defined infrastructure portfolio in terms of risks and rewards; and,
  • help demonstrate their ability to outperform the benchmark that best represents their strategy.

With proper benchmarks numerous applications are possible that will bring unlisted infrastructure forward as a fully-fledged asset class. Courageous and insightful investors will opt for transparency and relevance by letting go of absolute bench- marks that are now outdated.