Understanding ESG and sustainability measurement: from data to IT.
"Consider an IT architecture that supports ‘multiple versions of the truth’"
A wide range of ESG research information and data can be sourced from a large number of third party ESG and Sustainability providers, writes Chris Johnson, HSBC Securities Services (HSS).
Scores and ratings are the most widely known and recognised type of ESG data. These are designed to provide a grading of companies as to their performance using environmental, social and governance factors and methodologies. The providers of scores and ratings typically use a combination of company disclosed information and their own research. The underlying factors can also be adapted as a means to apply negative screening (e.g. alcohol, tobacco, arms) or positive screening (e.g. good corporate governance and working practices).
Climate data includes measures such as greenhouse gas (GHG) emissions which are usually presented as scope 1 (direct), scope 2 (indirect) and scope 3 (supply chain); additional metrics include carbon intensity and implied temperature rise.
The EU Sustainable Finance Disclosure Regulation (SFDR) includes a requirement for asset managers, in the European Union, to consider up to 20 Principal Adverse Indicators (PAIs), or to explain why these were not considered. There are up to 18 PAIs (mandatory) plus two to be selected from a list of 46 Additional (optional) adverse indicators. The PAIs include measurements of greenhouse gas emissions, biodiversity, water, waste and social and employee matters. So in total there are 64 adverse indicators (18 Principal and 46 Additional) which are expected to generate a significant amount of disclosed ESG data from companies in the EU from 2022. Non-EU asset managers managing EU domiciled funds, or marketing non-EU domiciled funds into the European Union, will also be impacted by requirements under the SFDR. The EU is also consulting on the creation of a database, called European single access point (ESAP) that would make the disclosed ESG data publicly accessible.
See also: Climate risk data is a hot mess. These open source pioneers want to set things straight.
The 17 Sustainable Development Goals (SDGs) were set in 2015 by the United Nations General Assembly and are intended to be achieved by the year 2030. The goals are wide-ranging and examples include poverty, hunger, health, education, gender, water, energy, work, innovation, inequality, climate action, life below water and life on land. An increasing number of ESG suppliers offer metrics for SDGs.
Other types of ESG data include controversies, which identify news about companies that could impact share prices, and forward-looking analysis in the form of various types of stress-tests and scenario analysis. Forward-looking ESG data is a growing area, supporting Value-at-Risk style analysis for specifics including the likely impact of climate change. The methods for gathering this data can be very advanced, and for example can include artificial intelligence techniques to obtain news from social media platforms.
Understanding ESG and sustainability measurement: The data provenance question
The majority of ESG data available on companies is unstructured and non-standardised. For example, ESG scores and ratings largely represent qualitative research opinions using proprietary methodologies.
This lack of consistency can result in significant variability in scores for the same company as measured by multiple ESG data suppliers, and can add complexity when creating comparable historical datasets. The proportion of formally structured and standardised ESG data is relatively low, at this stage, and is normally taken from annual company reports, or sourced from publicly available information. There can be variability between suppliers in the values returned, even for disclosed data points, and coverage levels can vary by region too. A case in point is GHG emissions data where approximately 4,000 companies disclose their emission levels, and so for the remainder of companies that don’t disclose several data suppliers will provide their own estimated levels. The imbalance between unstructured/qualitative and structured/quantitative is likely to be redressed as regulators mandate more company ESG disclosures, and as there is greater demand for consistent ESG metrics.
Regulation and Compliance
One body working to secure an increase in company disclosures is the Task Force for Climate-Related Disclosures (TCFD). Chaired by Michael Bloomberg, its 31 G20 member states are working towards its stated goal of achieving “Better information [that] will allow companies to incorporate climate-related risks and opportunities into their risk management and strategic planning processes”. The TCFD is being adopted more widely as a mandatory requirement, leading to greater company disclosures, as evidenced by announcements by the UK, New Zealand, Hong Kong and Singapore in recent months.
Among a range of regulatory frameworks being progressed, the EU has finalised a harmonised reporting template, which is used for SFDR, which details a very comprehensive range of adverse indicators as referenced earlier in this article.
Regulators in Asia, such as the Hong Kong Monetary Authority (HKMA) and Securities and Futures Commission (SFC) in Hong Kong, and the Monetary Authority of Singapore (MAS) in Singapore, are also consulting on ESG regulations.
There are also significant initiatives being proposed to achieve harmonisation and standardisation of ESG data. The European Securities and Markets Authority (ESMA) has written to the European Commission (EC) sharing its views on the need to ensure the quality and reliability of ESG data. And the International Financial Reporting Standards Foundation (IFRS) is pursuing an initiative towards achieving a global set of internationally recognised sustainability reporting standards to allow for consistency, comparability and transparency of reporting. This is supported by the International Organization of Securities Commissions (IOSCO).
Reporting opportunities and pitfalls
ESG scores and ratings are generally available for companies and so can be applied directly to equities, and for corporate bonds by linking issuers to the company ratings. ESG suppliers are constantly increasing their coverage levels but there is understandably a greater emphasis on companies with larger market capitalisation levels. So, reporting needs to take into account that scores and ratings might not be as readily available for companies with smaller capitalisation, or in regions where disclosure levels are lower, and for wider asset classes (e.g. collectives, derivatives (whether ESG-specific or not), government bonds, private equity and real assets).
Given the non-financial nature of ESG measurement, new techniques need to be considered. For example, a methodology and mechanism to capture historic realised gains for GHG emissions.
ESG scores and ratings, and carbon emissions data, can be quite static; in some cases, the data can be unchanged for 12–18 months. Therefore, it is possible that daily, or even monthly, tracking of changes in scores can have limited relevance. Scoring methodology changes in recent years, some leading suppliers have revised their scoring system for ESG. While this is positive, for continuous improvement purposes, if it were to be repeated, it could cause difficulties with tracking improvement and measuring historical performance.
There are some practical considerations before reporting of line-by-line ESG analysis and data. First, it is necessary to check whether the third-party supplier permits reporting onwards to other recipients, and also the fact that the scores and ratings can vary so much between suppliers, making it relevant to identify the supplier in order to enable identification of the source. It can be quite a complex undertaking to present ESG scores and ratings in a manner that is comparable. This is because the systems of rating/scoring can differ significantly, and the breadth of ESG information goes far and wide, with a multitude of factors and granularity possible should one delve beneath the E, S and G themes into the underlying factors.
Some technology considerations
Careful consideration is needed before automatically setting about adding ESG into IT systems.
It might be advisable to partition ESG data in a stand-alone environment, to enable flexibility while ESG business needs and requirements become more established, and until globally recognised ESG data standards have been defined and adopted.
It is also worth considering an IT architecture that supports ‘multiple versions of the truth’ as opposed to the more traditional ‘golden source’ concept that has been employed historically for much financial data. Ingesting third party ESG data can also pose challenges because ESG suppliers can have different approaches towards instrument and entity mapping, so an additional integration and translation layer, to provide data mapping between instruments and issuers, could be necessary.
