Thus we address the following question: How should we gauge the relationship between ESG and the market value of infrastructure investments? This is a key question that institutional investors and prudential regulators need answered in order to integrate ESG into their financial decision-making process, e.g. to assess sustainability risks under the Sustainable Finance Disclosure Regulation of the European Union (SFDR).

We review existing ESG reporting and assessment schemes used in the infrastructure sector and find that they are not designed to answer this question, but can provide a basis for a robust scientific framework that would create genuine ESG investment knowledge.

ESG reporting and investors’ demand for monitoring

As stated, many investors now include non-financial aspirations in their mandates and mission statements, in which ESG reporting plays a growing role. Beyond the affirmation and realisation of such non-financial objectives, ESG reporting and assessment schemes on infrastructure assets have also developed in a financial context. This is a response to an increasing demand for monitoring from investors (Holmström et al., 1993) who need more information than what can be gleaned from the market price of assets to make investment decisions.

There are three motivations for this fast growing interest in the ESG characteristics of infrastructure investments:

• The fact that these characteristics have known consequences on the present value of investments e.g. an infrastructure that pollutes will be fined in most jurisdictions. This knowledge is priced by investors and already reflected in asset values;
• The belief that the ESG characteristics, while they are not currently priced by markets or regulators (i.e. economic externalities), are likely to become priced. In other words, conditional on certain regulatory changes such as a carbon tax or systematic changes of attitude amongst consumers, etc., the present value of infrastructure investments is also a function of their ESG characteristics. The deeper the knowledge about how exposed infrastructure investments might be, the more ESG characteristics can be reflected in asset prices;
• The aim to ’do good’ and meet non-financial objectives irrespective of their impact on asset values. This could include excluding certain countries, sectors or assets from a portfolio, irrespective of the likelihood of an impact on asset values or its potential magnitude.

These motives are not exclusive, but they are different in nature. The first two are concerned with the relationship between ESG and investment outcomes, whereas the third motive requires investors to implement non-financial objectives, that is, it imposes constraints on their ability to meet financial objectives. This choice of constraints, while it can be popular and laudable, is necessarily ad hoc and does not help investors to understand how to utilise ESG characteristics to meet their financial objectives.

Investors’ motives should not be confused or equated with those of public policy bodies such as the European Commission. Policy makers are primarily concerned with public policy objectives e.g. preserving the natural environment for human populations to live in, preventing or mitigating climate change, etc. Investors making decisions based on the third motive may well espouse similar objectives as public policy makers, but the two other motives are at least as important. It should also be noted that for prudential regulators, whose role it is to preserve the stability of the financial system by preventing large cascading losses amongst market participants, the question of the relationship between ESG and asset prices is paramount. Indeed, so-called ‘sustainability risks’ i.e. the risk of loss of value of underlying assets due to environmental or social events, are one of the main concerns of the SFDR (Regulation 2019/2088, L317/9).

Hence, investors in infrastructure have two sets of decisions to make relating to ESG: First, which assets should they exclude or focus on because of their ESG characteristics? Second, given the characteristics of the acceptable universe, what risks are they exposed to that they should manage within their portfolio, through various forms of diversification, hedging or insurance?

If the impact of ESG characteristics on infrastructure asset prices is not easily gauged from market prices, then additional investment knowledge is needed to decide how to invest. Thus, beyond the societal demand for greater ESG content and outcomes of the investments made by investors, the demand for ESG reporting and benchmarking also springs from the second motive described above as does the need to manage risks related to ESG within the portfolio.

In the end, the relationship between ESG and the fair market value of assets is determined by the extent to which the ESG profile of a firm creates exposures to risks that materially (systematically) drive the discount rates of the future cash flows of its financial assets.

This focus on risk may seem at odds with the frequent insistence on the role of the ‘impact’ of a business or project in the ’green investing’ literature or marketing. Of course, any economic activity has an impact and infrastructure companies can have very significant impacts, both positive and negative, on their natural and economic environments. However, these impacts do, in turn, create risks i.e. they increase or decrease the payoff uncertainty of the investment. In fine, impacts contribute to the discount rate or expected return that investors require to buy or hold the asset.

Existing ESG reporting frameworks do not create investment knowledge

Next, we ask if existing ESG reporting tools create the investment knowledge that investors require i.e. do ESG schemes created and used by infrastructure investors help clarify the relationship between ESG and infrastructure asset prices? Over the past decade, in response to the increasing appetite of investors for understanding and measuring the ESG characteristics of infrastructure investments, numerous tools and standards have appeared to facilitate the reporting and assessment of ESG metrics. ESG schemes for infrastructure investors are still at the ‘proliferation’ stage of standard development and a degree of consolidation, as well as integration of these soft rules into more stringent and mandatory regulatory frameworks can be expected.

We propose a framework to integrate the role of ESG in the fundamental relationship between risk and fair value, which takes into account the role of each infrastructure company’s impacts on environmental, social and governance matters. To develop this framework, we conduct a comparative analysis of the existing ESG schemes used by infrastructure investors to determine the scope of ESG issues in relation to infrastructure investments, establish a common matrix or taxonomy of their ESG risks and impacts, and determine how the question of (financial) materiality i.e. what factors can be expected to systematically impact value, should be approached scientifically.

From the multiple standards available, we build a parsimonious taxonomy of ESG impacts and risks that, at the most general level, universally apply to any infrastructure company.

We use this taxonomy, which includes 10 super classes, 24 classes and 67 subclasses of ESG impacts and risks, to categorise 1,659 indicators, including 4,850 disclosures provided by existing schemes. This allows us to understand the scope, level of aggregation, and measurement difference of existing ESG schemes for infrastructure investment.

We find that despite current ESG standards being made for the purpose of infrastructure investing, the centrality of the firm and the importance of asset pricing are often ignored by or lost on existing schemes. These typically do not achieve a clear distinction between impacts and risks, in particular between the whose impacts and risks that ESG reporting and assessment should focus on. Instead, they tend to be lists of ’things that matter’ and do not necessarily focus on trying to measure the risks to which investors in infrastructure companies are exposed in the context of ESG. We argue that such lists, while very useful, fail to meet the standard of a genuine scientific framework: a list of concepts and categories that describe the relationships between them i.e. an ontology.

To define infrastructure, we follow the TICCS® classification system of infrastructure companies, which puts the firm at the centre of the approach. Infrastructure companies are what equity investors buy and debt investors lend to. Hence our focus is the ESG impacts of an infrastructure company, and what ESG risks it is exposed to. It follows that any ESG reporting or scoring, while it may spring from asset-level data, can be evaluated at the firm level, which is the correct unit of account for an investment reporting scheme.

There is little convergence between schemes in terms of scope (what the ESG perimeter includes), weights (what defines or constitutes materiality) and measurement (what data should be used to capture ESG characteristics). From one scheme to the next, the ESG performance of infrastructure companies is currently measured and presented in different and evolving ways. We find:

• Significant scope divergence between schemes as evidenced by the different biases, incomplete coverage and lack of overlap in terms of risk and impact classes, which is also a sign of measurement divergence;
• Measurement bias in the reporting of ESG information with the dominance of qualitative measures reported;
• Impact bias in the reporting of ESG information, and little attention to measuring risk exposures, especially not through quantitative risk reporting;
• Process and input indicator bias in the reporting of ESG information, highlighting the role of proxies in the various scoring and ratings methodologies used since actual impacts are not directly measured or reported.

Because of their lack of focus on the firm and its value, existing schemes focus almost entirely on ’impacts’, which may of course be indirect factors of risk, but also do not shed much light on the direct risks that arise from ESG. Some 88% of reviewed disclosures focus on impacts while only 12% aim to capture direct risks.

Our findings point to several likely developments in the area of ESG ratings and certification provision:

1. Infrastructure investment ESG standards will continue to change: the current absence of consistent definitions or approaches means that individual standards need to evolve and redefine their scope and methodologies;
2. This consolidation will be driven by end users: the degree of clarity and consensus around the objectives and the definitions used by ESG schemes, as well as the embedded assumptions that underpin these choices are likely to contribute to standard adoption, credibility and, eventually, dominance;
3. Schemes that also address the most pressing questions of policy makers and regulators are more likely to attract users. In the case of infrastructure investment, this is particularly the case with regards to climate change.

Creating an infrastructure ESG domain of knowledge for investors

To support the development of relevant ESG investment knowledge, we explicitly restrict the analytical framework to the link between ESG and asset prices.

Investors recognise that ‘externalities have consequences’ and, with rapid social and environmental changes over the past decades and the expectation of even more uncertain evolutions, they also anticipate these consequences by demanding better knowledge about their investment choices. This is what they and regulators need to understand in order to manage risks in the portfolio.

In the end, creating ESG investment knowledge does not change or remove economic externalities, it only makes them and their potential consequences for businesses more apparent and better documented. It is the knowledge of the uncertain consequences of externalities, including on future regulation or cash flows, that can influence asset prices.

In essence, the current demand for ESG reporting stems from two issues: 1/ a lack of knowledge regarding the ESG impacts and risks of infrastructure companies; and 2/ the fundamental uncertainty that the ESG aspects of their activity create for investors. Addressing the first issue amounts to documenting the exposure (or beta) of a company to certain risks. For the second, the consequences of ESG impacts and risks themselves for the firm remain uncertain, but can inform decision making and become a driver of the cost of capital in infrastructure investment.

We show that the scientific development of a body of ESG investment knowledge (or ontology) requires a number of key building blocks:

1. The clearly stated aim to create knowledge that relates the ESG characteristics of infrastructure companies – the entities in that investors decide to buy or hold – to investment decisions made on financial grounds i.e. considerations of risk and reward;
2. This helps clarify that the impacts of interest are those of an infrastructure company and the relevant risks are those to which the same company is exposed. Hence, the relevant domain of knowledge: instances of ESG risks and impacts of infrastructure companies. By grounding the approach in this manner, it becomes clearer that impacts are also sources of risks;
3. Next, a classification system is needed for the various objects of interest, including of course infrastructure companies and their ESG risks and impacts, but also standard classes of attributes and relations that allow the ESG characteristics of infrastructure companies to be described and create this knowledge. The definition of the attributes and relations that create this knowledge can then be science- and theory-based, using the most consistent assumptions or models in order to create a broad user base and maximise potential commitment by users.
4. Finally, this allows the question of materiality to be addressed. Materiality is a weak point in existing ESG schemes: they provide lists of potential material information to report or collect, but do not anchor this materiality in objective measures that would relate to the activities of infrastructure companies. Developing science-based materiality profiles for each of the 95 types of infrastructure assets captured in the industrial activity pillar of the TICCS® classification is a key step in the development of a body of ESG investment knowledge for infrastructure investment.

With this paper, we have laid out a roadmap for the scientific development of ESG knowledge for infrastructure investment. Future research and industrial efforts to consolidate and develop ESG investment knowledge can be expected as this knowledge becomes increasingly in demand from investors and their regulators.

Alignment with SFDR

The approach proposed in this paper is aligned with the work of the European Union’s SFDR expected to come fully into force in 2022. SFDR requires “financial market participants and financial advisers (…) to disclose specific information regarding their approaches to the integration of sustainability risks and the consideration of adverse sustainability impacts” (SFDR, L317/2).

While its primary public policy objective is to minimise adverse impacts on the environment and society, as mentioned above, SFDR is also about the risks to asset values. It requires the disclosure of so-called sustainability risks that pose “an environmental, social or governance event or condition that, if it occurs, could cause an actual or a potential material negative impact on the value of the investment” (SFDR, L317/9).

In effect, a taxonomy of the ESG impacts and risks of infrastructure companies is an essential step to address the concerns of SFDR. Moreover, one of the foundations of SFDR, is another taxonomy: the EU Taxonomy for Sustainable Activities describes the sustainability characteristics of various forms of industrial activities, including that of infrastructure companies. In other words, the EU Taxonomy is a first attempt at building objective materiality profiles that can be used to assess the ESG characteristics of an infrastructure company objectively.

Finally, this description of what matters from an ESG standpoint is to be documented using Regulatory Technical Standards (RTS) establishing a framework of reporting on principal adverse impacts and risks. A first draft describing the ESG data that will be required by the RTS was published in the Final Report on draft Regulatory Technical Standards, of the Joint committee of the European Supervisory Authorities in February 2021, and describes detailed indicators for environmental and social impacts.

To ensure compatibility with the SFDR, the EDHECinfra ESG taxonomy allows mapping of the required disclosures to respective impact and risk classes. Given that the EDHECinfra taxonomy is an exhaustive list of ESG impacts and risks for the infrastructure sector, 100% of the mandatory disclosures can be mapped to the subclasses of this taxonomy of risks and impacts. To enable measurement, each impact and risk can then be measured as indicators, which in turn will be informed by data collected according to the materiality profiles of each company and asset type as defined by TICCS®. In the RTS, these indicators are divided into a core set of 18 universal mandatory indicators that will always lead to principal adverse impacts of investment decisions on sustainability factors, irrespective of the result of the assessment by the financial market participant, and additional opt-in indicators for environmental and social factors, to be used to identify, assess and prioritise additional principal adverse impacts.

Future work by EDHECinfra focuses on supporting the implementation of the roadmap described in this paper, including documenting the ESG characteristics of infrastructure companies.