Research Reports 10 November 2020
The Operating Model to Enable Sustainable Investing
Improving ESG Data Quality, Frameworks, & Measurement Tools
In this excerpt from the Future of Sustainability in Investment Management report, explore how investment organizations manage data, technology, systems, and tools to enable sustainable investing practice.
How the Organization Manages Its Products and Services
In the operating model, we consider data, technology, systems, and tools.
Environmental, social, and governance (ESG) data were a minor part of all investment data 5–10 years ago and now are a very significant data source when looking across all data use at investment organizations. We note a long list of improvements that investors seek from their data and technology.
The operating model challenges in deploying sustainability capabilities can be split into problems with data quality and challenges related to organizational structure, culture, and focus in managing and using data.
ESG Data Quality
ESG data are substantial and fast growing but unwieldy. In the area of data quality, investors desire their data and the management of data to be material, to enable investment decision making to be more data driven and evidence driven; valid, to provide a consistent view for all assets and allow comparability; and usable and adaptable, to build an information edge from data and improve knowledge management.
We can characterize the data challenge generally for investment firms as creating a technology system that aims to process and channel relevant high-quality information adaptably, cheaply, and efficiently with security and resilience into the investment process and into reporting to clients. The test of quality in data will be substantially about the depth of inferences that can be made from the data and the connected judgements, heuristics, and algorithms that can be applied to the data.
Data quality requires two essential features — materiality and validity. Materiality tracks the degree of insight possible from any data point in addressing investment-relevant questions. Validity tracks the actual capture of that insight in that data point. Validity will reflect objectivity, accuracy, timeliness, granularity, and transparency. Objectivity carries particular weight in this list; it is how repeatable the results are if measured again or how much they derive from direct measures or from a model. Objective data are seen as “‘hard”; subjective data are seen as “soft.”
- Materiality is the degree to which the precise form of a measure reflects decision-useful insight about investment-relevant questions.
- Validity is the degree to which an actual data point is an accurate representation of the measure in question, where validity is reduced by subjectivity and various problems of accuracy, timeliness, granularity, and transparency.
Soft data are data that are hard to measure and express, contrasting with hard data, which are the traditional form and the opposite of soft data. Soft data generally come from assessment, opinion, experience, or interpretation or through modeling that is, therefore, relatively subjective and has validity issues. There is considerable soft data in ESG areas that have high materiality but validity issues.
For example, diversity data are material for identifying good corporate culture and effective decision making, but data on racial backgrounds are typically not adequately captured, relying on estimation. Therefore, such data are relatively soft. Good diversity practice is likely to be captured only by employee surveys that are subjective in that they are opinion based.
Investors tend to evaluate the benefits of a given level of data quality somewhat narrowly; they often unduly favor simple facets of data quality, such as objectivity and accuracy; and they usually do not sufficiently consider the full data context, in terms of its materiality and the natural scarcity of good quality data in complex systems where simple causality is not present.
Investors also tend to use data without sufficient regard to data drawbacks. By explicitly assessing materiality and validity, the data “provenance” can become a better input to how much weighting is justified. But the behavioral context is critical as well. This suggests the need to handle soft data with a full appreciation of its influences. For example, if soft data are made an explicit and “hard” target, there are likely “gaming” and other governance difficulties. It is possible to diminish these impacts if multiple data points are included as reference points instead of explicit targets.
Investors have growing opportunities to put soft data and alternative data, derived from non-traditional sources such as financial transactions, sensors, mobile devices, satellites, public records, and the internet, alongside traditional data as enhancements to overall data quality or to use such data in artificial intelligence algorithms.
The challenges of all types of data, but soft data in particular, produce large-scale problems of consistency in issuer reporting of ESG data, as well as in understanding the outputs from the ESG rating companies, where large differences can exist from the data models used.
Investors face overlapping challenges because of these problems: distributing data efficiently, using data efficiently, knowing the levels of data quality provided in the current systems and tools, and contributing to better data quality.
These issues make reporting standards particularly important to improve company and issuer disclosures.
ESG Data across the Investment Ecosystem
Analysis and data on ESG factors are critical at three points in the investment ecosystem:
- Company reporting
- Companies’ obligations to report on ESG factors in their financial and statutory reporting are relatively light.
- The SASB and GRI initiatives have been setting standards that extend these obligations. The “Statement of Intent to Work Together Towards Comprehensive Corporate Reporting,” published by SASB and GRI, as well as CDP, the Climate Disclosure Standards Board (CDSB), and the International Integrated Reporting Council (IIRC), marks an important step forward.
- Investor analysis and decisions
- Investors rely on a mix of internal and external resources for their analysis.
- ESG ratings by such organizations as MSCI and Sustainalytics are used widely as inputs to analysis.
- The ESG analysis is turned into active portfolios and, via rules-based methods, into ESG indexes.
- Investor reporting
- Asset managers reporting on their ESG-related products with product disclosures will be subject to industry standards in the future (the full report highlights the forthcoming CFA Institute standards).
- Asset owners reporting on their portfolios are subject to increased regulation in certain jurisdictions, particularly in Europe. Climate reporting is increasingly expected, consistent with TCFD disclosure standards.
It is important to recognize how company reporting plays a key part in the sourcing of ESG data. But in sustainable investing we have areas where regulations and standards are still developing. Although many practitioners see substantial opportunity for standards to play a bigger part in better data and expect that standards will converge, one-quarter are still unsure about how the data challenge will be resolved, as shown in Exhibit 21.
Note: Exhibit 21 can be found within the main report.
The consistency and comparability of ESG data from companies is poor. There are very limited national requirements for companies to report on most ESG data, with companies left to determine for themselves which ESG factors are material to their business performance and what information to disclose to investors.
But there are signs that certain initiatives will pay off over time. The major global ESG information disclosure frameworks/standards are as follows: the GRI Standards developed by the Global Reporting Initiative, the SASB Standards developed by the Sustainability Accounting Standards Board, the International Integrated Reporting Framework developed by the International Integrated Reporting Council (IIRC), and the recommendations prepared by the Task Force on Climate-Related Financial Disclosures (TCFD) at the request of the Financial Stability Board.
Each has its part to play at this point in the developing ecosystem. Both “rule-based” (GRI and SASB) and “principle-based” (IIRC) methods are adopted. On one hand, rules have the advantages of being clear and easy to understand as to what should be done; on the other hand, there are many instances where strict rules and regulations end up with inflexible and superficial responses. Hence with TCFD, which has both rules and principles, for example, explicit data are sought on greenhouse gas emissions, but in scenario testing models, the TCFD framework requires companies to just illustrate their approach.
ESG data usage varies by context, and the range of standards may enable a degree of bespoke reporting that is helpful for firms. The alignment of ESG use cases with the different reporting approaches is illustrated in research by Nissay Asset Management. Although the convergence of standards may seem attractive, the underlying issues are complex and the specialisms in the current arrangements seem at present to function adequately.
The current reporting ecosystem has been a work in progress for more than a decade. There is a wide base of dissatisfaction among investment industry stakeholders with the present position on data and reporting. Going forward, we can expect a greater degree of urgency from the industry to make significant progress, with a stronger framework likely to emerge, albeit slowly. In January 2019, the World Economic Forum (WEF), in collaboration with Allianz SE and the Boston Consulting Group, suggested the following actions are necessary:
- Improved transparency of the entire ecosystem (such as alleviating duplication of activity and unintentional conflicts)
- Effective and active cross-system interactions (such as incorporating more of the end user’s needs)
- Stricter harmonization of methodologies for measuring key performance indicators (KPIs) related to ESG (such as enhancing the comparability of KPIs to help the decision making of investors and others)
Investor Analysis and Decisions
Valuation approaches lack consistency, and investment professionals report various ways of incorporating ESG into equity analysis, as shown in Exhibit 23. (Exhibit 22 was intentionally omitted from this excerpt.) ESG company ratings are widely used by practitioner survey respondents, with 63% using them as a part of their data analysis.
ESG data providers, such as MSCI and Sustainalytics, play an important role in sustainable investing by gathering and assessing information about companies’ ESG practices and then scoring those companies accordingly. These ratings systems are widely used by investment organizations in both analysis and reporting.
The rating approaches naturally vary and are not the subject of standards. These differences in approach arise in what data are collected, what research is conducted, and the models then applied to produce ratings, including scoring methodologies and weightings attached to various ESG issues. As a result, the rating for a single company can vary widely among different providers. Research by State Street Global Advisors demonstrated a correlation of 0.53 for MSCI and Sustainalytics ratings across the MSCI World Index of companies. As a point of comparison, correlations in company credit ratings have generally been around 0.9.
This inconsistency, alongside issues about transparency, has been the source of adverse industry comments. It is inevitable that a range of ratings would emerge, given that so many ESG data sources are intrinsically soft, the data purpose can vary, and the source data can be structurally weak. Nevertheless, we expect correlations to increase somewhat over time. But the lack of correlations can be viewed as an opportunity to differentiate investors’ approaches by adding value with proprietary research. Among practitioner survey respondents, 73% expect the influence of ESG ratings on firms’ cost of capital to be greater in the next five years.
Another area of data growth has been in data related to climate, and the following exhibits highlight data from the recent report “Climate Change Analysis in the Investment Process.” As seen in Exhibit 24, about 40% of investment professionals incorporate climate risk into their analysis, and the primary reason is that it is material.
Client demand is another motivating factor, with a third of Asia-Pacific (APAC) investors and half of Europe, Middle East, and Africa (EMEA) and Americas investors looking for their investment firm to incorporate climate change.
Exhibit 25 shows that the most common types of risk considered are physical and transition risks.
However, there are data needs, and 78% of practitioner survey respondents want at least one type of climate information that is missing; details are provided in Exhibit 26.
ESG research and ratings from specialist data providers are currently a key part of the ESG ecosystem, and we expect this position to strengthen further in the next 5–10 years.
The challenges of data in sustainable investing have allowed growth in the practice of “greenwashing.”Greenwashing means conveying a false impression or providing misleading information or a misleading narrative about how a company and its products are environmentally sound or positive in an ESG context. There is some concern that the recent significant inflows to ESG products may not be based on full information and that if outcomes disappoint, there could be a backlash.
In Exhibit 27, we have evidence of a strong conviction that the issue is problematic and a wish to introduce standards to diminish greenwashing; 78% of practitioner survey participants support such standards.
Greenwashing exists across the chain in the ecosystem. One particular area where standards appear to be timely is in the confusion surrounding investment products that offer one or more ESG-related features. The EU taxonomy for sustainable activities and the Sustainable Finance Disclosure Regulation (SFDR) will help reduce greenwashing, and CFA Institute is developing a voluntary global industry standard that would establish disclosure standards for investment products with ESG-related features. The purpose of the standard is to provide greater product transparency and comparability for investors by enabling asset managers to more clearly communicate the ESG-related features of their investment products.
Organizational and Cultural Aspects of Data and Technology
In the organizational structure and culture of data management, organizations naturally seek data to be easy to access and consistent at all parts of the organization or wherever it has significance. The degree to which there is external dependency for data interpretation militates against this goal, as do inconsistent data architecture for different parts of the portfolio and investment process, an inability to view data in a total portfolio context, and investors’ ability to be agile in innovation and adopting new technology. In each of these four areas, current data practices fall far short of “fit-for-purpose” status.
Critical to good practice is for investors to be able to analyze ESG data their way to reach their conclusions. This is not an argument for excluding the views of other experts but is a complement to the investor’s analysis.
Investment organizations seeking to achieve a competitive advantage with their ESG analysis will recognize that technology is a necessary foundation. The discussions on technology and data in our roundtables demonstrated that the technology opportunity was growing for all, with more data sources becoming available and more differentiation possible. As shown in Exhibit 28, 71% saw alternative data as benefiting the robustness of sustainability analysis, and 43% saw sustainability as benefiting from the application of artificial intelligence. The proportion that saw a particular edge from proprietary data in the ESG area was 27%.
Views were spread out on the best technology strategies. Many asserted that larger-scale change projects were necessary. Others were more supportive of adding to technology incrementally. All expressed a view that ESG data needed to be managed better going forward. Most expressed the view that effective changes to technology were extremely difficult to execute and that cost and time overruns were routine.
The technology issues were in large part cultural in nature and started with many organizations working in siloed and fragmented organizational structures. This separation problem produces disparate management views on technology and limits investment organizations’ capabilities with and capacity for technology.
The outcome in the ESG area is that data management tends to be segregated and disjointed and struggles to support the integrated management of ESG insights alongside the traditional sources of data. In current practice, a combination of siloed structures and large-scale, heterogeneous datasets means that ESG data tend to be uneven and fragmented across the organization and lack search and research ease.
The organizational issues include the all-pervading presence of legacy technology — spreadsheets, e-mail, and electronic shared file structures — as the key infrastructure for ESG data. We are still some way off from bespoke software systems and cloud computing innovations in this area.
The proliferation of new data sources and analytic technologies that are likely to be a feature of the ESG growth phase could potentially overwhelm current data-governance practices by greatly increasing fragmentation. We suggest that ESG success will hinge heavily on how well an organization’s technology adapts to the new circumstances of more data alongside more complexity in organizational structure.
These problems are exacerbated by the cost–benefit disconnect that inevitably exists in an area where the core problem of a strategy for effective technology is so ill defined and contested. We are accustomed to robust risk budgets, governance budgets, and financial budgets, but technology budgets do not seem to be very effectively managed because of the challenges from silos and the communication gaps that are at the center of these problems.
Can you imagine a world in which
- material ESG factors for a company are as easily obtained as an income statement?
- all inputs and outputs related to business processes, including externalities, are priced accurately and equitably?
- performance attribution reports include impact attribution alongside risk and return?
- internal technology end users don’t have to rely on others for data inference or analysis?
- legacy technology is fully replaced by bespoke software systems and cloud computing innovations?
About the Author(s)
Rebecca Fender, CFA, is the Chief of Staff for the Research, Advocacy, and Standards area of CFA Institute. She provides strategic analysis and partners with key stakeholders to develop and implement initiatives, processes, and metrics that will improve the team’s effectiveness.
Previously, she was a member of the industry research team and the founder and leader of the Future of Finance initiative. As the thought leadership platformforCFA Institute, her group published studies to help investment professionals build their careers and serve their clients more effectively.
Ms. Fenderhas testified before the U.S. House Financial Services Committee AI Task Force on the impact of artificial intelligence on investment roles. She speaks regularly at industry events and has been quoted in the Financial Times, Bloomberg, and the New York Times, among others.
Prior to joining CFA Institute, Ms. Fender was a vice president at BlackRock working with pension funds and endowments, and she also worked at Cambridge Associates, where she published research about manager selection. She earned her undergraduate degree in economics from Princeton University and holds an MBA from the Darden School at the University of Virginia.
Robert Stammers, CFA, is director of Investor Engagement on the Future of Finance team for CFA Institute. Prior to joining CFA Institute, Mr. Stammers was the principal for his founded company, where he consulted to aide real estate owners, lenders, and syndicators, develop and analyze structured real estate investments.
As a senior executive for several institutional fund managers, Mr. Stammers was the portfolio manager for a $1 billion enhanced real estate fund, a $1.2 billion private timber fund, and several pension fund separate accounts.
Mr. Stammers has his bachelor of arts in economics from Connecticut College, his masters in business administration from Emory University, and was awarded the CFA designation in 1997.
Roger Urwin is global head of investment content and advisory director at Towers Watson. Previously, he was global head of its investment practice and worked at William Mercer and Gartmore Investment Management. Mr. Urwin is the author of a number of papers on asset allocation policy and manager selection and serves on the board of the Institute for Quantitative Investment Research and as advisory director to MSCI. He holds a master's degree in applied statistics from Oxford University and has qualified as a fellow of the Institute of Actuaries.
Rhodri Preece is Head of Industry Research for CFA Institute. He is responsible for building and maintaining the global thought leadership function at CFA Institute, including leading the planning, coordination and creation of research content across CFA Institute research platforms including the Financial Analysts Journal, the Research Foundation, and the Future of Finance initiative.
Rhodri formerly served as head of capital markets policy EMEA at CFA Institute, where he was responsible for leading capital markets policy activities in the Europe, Middle East and Africa region, including content development and policy engagement.
Prior to joining CFA Institute, Mr. Preece was a manager at PricewaterhouseCoopers LLP where he specialized in investment funds.