Pain Spots and Opportunities Regarding Environmental, Social, and Governance (ESG) Data

Abstract

All economic activity has an impact on society and the environment. Better knowledge about these non-financial consequences is being increasingly required of company managers, finance sector professionals, regulators and public policy makers, and ultimately by individual people when they purchase everyday items. However, relevant information is not universally available at the point of making critical decisions because the reporting of non-financial data is laborious, costly, and time consuming. This paper introduces a new data science construct, referred to as the 1n∞ model (pronounced “innate” model), to harness linked data protocols for distributed data. It shows that meaningful analysis of non-financial information (environment, social, and governance data/ sustainability data/ impact data) could be easy, accessible, and real-time. The industry of non-financial data could be disrupted by an “Internet of Impact”.

Full text

1
PAIN SPOTS AND OPPORTUNITIES REGARDING ENVIRONMENTAL, SOCIAL,
AND GOVERNANCE (ESG) DATA
Imagine a future in which meaningful analysis of non-financial information is
easy, accessible, and real-time
Karl H Richter ( 瑞可诚 )
8 January 2021
Original version written in English and translated into Chinese

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
2
Author’s Note
Special thanks to David E Wilton1 and Menglu Zhuang2 for reviewing this text and
helping to shape it with their valuable knowledge and expertise (I take full responsibility if I
have misinterpreted any of their comments). I would also like to thank Kate Ruff3 and
Willem Schramade4 for energising collaborations in advance of writing this paper, many of
their insights have found their way into this text. Finally, I would like to thank Angela Bai5
and the team at CASVI6 for inspiring discussions, and for sharing information about the
CASVI-3A model, without which this paper would not have been possible.
1 https://www.linkedin.com/in/david-wilton-impact-emerging
2 https://www.linkedin.com/in/menglu-zhuang-044a0358
3 https://www.linkedin.com/in/katherineruff
4 https://www.linkedin.com/in/willemschramade
5 https://www.linkedin.com/in/angela-bai-014a4212
6 China Alliance of Social Value Investment https://www.casvi.org

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
3
Abstract
All economic activity has an impact on society and the environment. Better
knowledge about these non-financial consequences is being increasingly required of company
managers, finance sector professionals, regulators and public policy makers, and ultimately
by individual people when they purchase everyday items. However, relevant information is
not universally available at the point of making critical decisions because the reporting of
non-financial data is laborious, costly, and time consuming. This paper introduces a new data
science construct, referred to as the 1n∞ model (pronounced “innate” model), to harness
linked data protocols for distributed data. It shows that meaningful analysis of non-financial
information (environment, social, and governance data/ sustainability data/ impact data)
could be easy, accessible, and real-time. The industry of non-financial data could be disrupted
by an “Internet of Impact”.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
4
Table of Contents
1. Introduction 5
2. Global Challenges in Sustainability Assessment, ESG Ratings, and Impact
Evaluation 9
3. Providing Context Through a High Level Comparative Review of the CASVI-3A
Model 13
4. Potential to Disrupt the Provision of Non-Financial Information 21
5. A Different Technological Mindset Could Unlock New Opportunities for Agile
Data Providers 28
6. Conclusion 34
References 38

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
5
1. Introduction
Every type of economic activity and financial transaction contributes in some way to
the wellbeing of people and the ecology of our planet. These cumulative effects can be either
positive or negative, yet there is currently no easy way to know what this collective impact is.
Imagine a future in which it is possible to compare the non-financial impact of
groceries as easily as their price (Instans, 2020). Unless people are able to know the non-
financial consequences of their actions and decisions, at the time they are being made, then it
is not possible for them to have more than an educated guess about how they are affecting
society or the environment at large. This applies to the real economy as much as the financial
sector. It is not surprising therefore that across different cultures and political movements,
people are in their own way trying to interpret what “doing well and doing good” means7.
The non-financial consequences of economic activity are usually referred to as
externalities because the “costs or benefits on others … are not reflected in the prices charged
for the goods and services being provided” (Khemani & Shapiro, 1993). It has been argued
by Wilton (2019) and others that the remedy for this market failure requires a substantive
expansion of the theoretical underpinnings of capital markets and the economy. They suggest
7 Typically, this phrase is used to convey some form of dual strategy that achieves both financial
success and benevolent objectives at the same time.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
6
that non-financial impacts (i.e. externalities) can – and should – be internalised into financial
and economic considerations as a third theoretical pillar alongside financial return and risk.
While there are several notable initiatives that go some way in remedying this, such as the
sustainable finance regulation adopted by the European Union in 2020 (European
Commission, 2020) – ultimately, an overarching theoretical framework, or universally
harmonised regulatory requirements, are not yet established according to any globally
accepted norms.
The lack of standards presents both an opportunity and frustration for the several
information platforms providing ratings and data about the non-financial impacts of
companies and investment products. One of the most recent platforms is the Social Value 99
(SV99) Index8 developed and provided by the China Alliance of Social Value Investment
(CASVI). On the one hand CASVI has had the benefit of being able to study established
platforms such as FTSE4Good Index Series9, MSCI ESG Research10, and Sustainalytics by
Morningstar11. On the other hand, CASVI is already publishing its third annual index for
2019 and, like other established providers globally, needs to consider how best to react to the
8 https://www.casvi.org/
9 https://www.ftserussell.com/products/indices/ftse4good
10 https://www.msci.com/research/esg-research
11 https://www.sustainalytics.com

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
7
rapidly changing demands placed upon them whilst at the same time leveraging their existing
experiences and valuable trove of historic time-series data.
Within this context, this paper is written to stimulate discussion about how the future
of non-financial reporting could be. This paper incorporates the author’s knowledge of the
CASVI analytical process12 as well as broader academic literature and industry best
practices13 for the basis of comment and to understand the inherent challenges.
This paper also goes further to imagine what the key elements of next-generation non-
financial reporting might include, taking inspiration from the idea of an “Internet of
Impact”14. Deng et al. (2019) refer to the “Internet of Impact” as an idea that expresses
joined-up, searchable, and scalable data. It can help align local, national, international, and
global efforts towards data interoperability regarding the Sustainable Development Goals
(SDGs)15. Moreover, the vision of an “Internet of Impact” imagines a paradigm shift away
from impact data that is collected retrospectively, towards real-time data about non-financial
effects and better predictive analysis. This would be a notable change from the current
12 The author is an advisor to CASVI and Global Co-head of the CASVI Centre of Excellence.
13 The author is a part-time lecturer at the Frankfurt School of Finance and Management, in the field
of managing the non-financial impacts of sustainable finance.
14 The author coined the term “Internet of Impact” in 2016 whilst leading a practitioner working group
for the OECD. This working group was part of the G7 initiative that tasked the OECD build the
evidence base for social impact investing. See http://www.oecd.org/development/financing-
sustainable-development/development-finance-topics/social-impact-investment-initiative.htm
15 https://sdgs.un.org/goals

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
8
paradigm in which data are primarily generated by analysts studying published reports,
requesting information from companies via laborious questionnaires, or by undertaking
customised assessments of specific activities.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
9
2. Global Challenges in Sustainability Assessment, ESG Ratings, and Impact
Evaluation
Investor interest in sustainable finance is growing rapidly year on year. In 2019, the
amount of new capital flowing into sustainable finance was more than the previous four years
combined (Iacurci, 2020). However, as if to moderate the enthusiasm of investors, the Chair
of the Securities and Exchange Commission (SEC) in the USA warned that “I have not seen
circumstances where combining an analysis of E [environmental], S [social] and G
[governance] together, across a broad range of companies, for example with a ‘rating’ or
‘score’, particularly a single rating or score, would facilitate meaningful investment analysis
that was not significantly over-inclusive and imprecise” (Mirchandani, 2020).
This scepticism about the meaningfulness of current approaches to environmental,
social, and governance (ESG) data, and sustainability assessment or impact evaluation
generally, is not isolated. The Japanese Government Pension Investment Fund (GPIF)
reported that the “correlation of ESG evaluations … is very low”, and that “the improvement
of ESG evaluation method[s] are essential” (GPIF, 2017). See Figure 1 below.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
10
Figure 1: Correlation of ESG ratings by FTSE and MSCI for Japanese companies
Note. From “Selected ESG Indices” by GPIF p. 13 (Slide 11).
https://www.gpif.go.jp/en/investment/pdf/ESG_indices_selected.pdf
Separately, the Aggregate Confusion Project at the Massachusetts Institute of
Technology (MIT) confirms that sustainability data are weakly correlated. According to
Rigobon (2020), who is co-founder of the project, “it is very likely … that [a] firm that is in
the top 5% for one rating agency belongs in the bottom 20% for the other. This extraordinary
discrepancy is making the evaluation of social and environmental impact impossible.”
The analytical framework developed by the Aggregate Confusion Project separates
the different methodologies and diverging scores in terms of “scope, measurement, and
weights” (Berg et al., 2020). According to this framework, Berg et al. (2020) found that
“measurement divergence is the most important reason why ESG ratings diverge, i.e.
different raters measure the performance of the same firm in the same category differently”.
Indeed, different preferences, values, and ethical perspectives will result in diverging
interpretations of the same underlying facts.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
11
Berg et al. (2020) argue that it is legitimate for different sustainability assessors to
have different opinions about scope and weights i.e. to decide what issues should be included
in the assessment and how each issue should be interpreted. They say that the resulting
variation in ESG scores is therefore not only understandable but desirable because the “users
of ESG ratings also have heterogeneous preferences for scope and weights”. In other words,
investors should be free to express a preference for sources of ESG information that
correspond with the same range of non-financial issues they also find materially relevant –
and equally to adopt the biases and interpretations that align most closely with their own
value judgements about how important those issues are.
This argument is consistent with arguments made decades previously in the context of
assessing the effects that organisations have on individual stakeholders. Gray et al. (1997)
said that social accounting must acknowledge and facilitate “polyvocal” voices (i.e.
pluralistic interpretations) of individual stakeholder opinions to enable full accountability in
terms of the organisation perspective, the societal perspective, and the citizen perspective.
This may be true when interpreting the ultimate effects on stakeholders, but it is
counterproductive if value judgements and biases distort source data that are expected to be
incontrovertible and unbiased.
Berg et al. (2020) say that their findings “demonstrate that ESG rating divergence is
not merely driven by differences in [scope and] opinions, but also by disagreements about
underlying data”. They say that such “divergence is problematic … if one accepts the view
that ESG ratings should ultimately be based on objective observations that can be
ascertained”.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
12
A similar conclusion can also be reached from a different perspective. According to
Wilton (2019), “incorporating impact as an additional objective in investment decisions is a
global trend which will continue to grow”. Instead of considering impact investing to be
merely a subset of the total investment universe, Wilton (2019) argues that the notion of
impact is a new third theoretical pillar of finance. Giving impact a similar theoretical status to
returns and risk “has major implications for the pricing of capital across all asset classes
globally” (Wilton, 2019). In order for this to be achieved, Wilton (2019) argues that a
“general theory” of finance is necessary, and that such a general theory requires simplified
data that are primarily quantified, and possibly qualified where necessary. The primary
requirement however is for data that are based upon objective observations, which are not
influenced by the individual mandate or scope of investors, and are free from the interpretive
bias or weights of individual analysts. Currently these kind of data are not universally
available, and they are impossible to extract from the majority of ESG information providers,
sustainability reports, or impact analysists.
Wilton (2020) summarises the problem into two categories: Firstly, that aggregated
data or ratings obscure detailed information that would be more effectively revealed via
disaggregated (unbundled) data – and secondly, that current methodologies suffer from
inherent biases and built-in preconceptions that limit their operational usefulness.
Before exploring the potential for disruption that these ideas suggest, the following
section first provides greater context by reviewing the analytical model of one of the newest
entrants in the ESG data and rating industry.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
13
3. Providing Context Through a High-Level Comparative Review of the CASVI-3A
Model
The CASVI-3A model is the primary tool for the China Alliance of Social Value
Investment (CASVI) to score the sustainability related effects of companies. The model
forms the basis of the CASVI Social Value 99 index16, the selection of stocks in the Bosera
CSI Sustainable Development 100 ETF investment vehicle17, and other CASVI products and
services. 3A is an abbreviation for Aim, Approach, and Action, which are the main category
headings of the CASVI evaluation framework. Figure 2 below summarises the sustainability
factors considered by the model, see Appendix 1 for more detail. The overall methodology
also includes sub-model elements for negative screening (see Appendix 2), which exclude
companies that are deemed to have harmful or undesirable sustainability effects (Ma, 2019).
16 https://www.casvi.org/en/h-col-103.html
17 https://www.casvi.org/en/h-col-240.html

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
14
Figure 2: CASVI-3A Model (Aim, Approach, Action) – see Appendix 1 for more detail
Note. Source: CASVI
CASVI developed their methodology from the ground up after undertaking extensive
research of global best practice (see Figure 3 below), as well as the Sustainable Development
Tier One Indicator Tier Two Indicator Tier Three Indicator
1. AIM
Driving Force
1.1 Value Driven 1.1.1 Core Values
1.1.2 Business Ethics
1.2 Strategic Driven 1.2.1 Strategic Objectives
1.2.2 Strategic Planning
1.3 Business Driven 1.3.1 Business Positioning
1.3.2 Target Customers & Users
2. APPROACH
Innovation Force
2.1 Technical Innovation 2.1.1 Research and Development
2.1.2 Products & Services
2.2 Model Innovation 2.2.1 Business Model
2.2.2 Industry Impact
2.3 Management
Innovation
2.3.1 Corporate Governance
2.3.2 Information Disclosure
2.3.3 Risk Control
2.3.4 Incentive Mechanism
3. ACTION
Transformation
Force
3.1. Economic
Contribution
3.1.1. Profitability
3.1.2. Operation Efficiency
3.1.3. Solvency
3.1.4. Growth Capability
3.1.5. Financial Contribution
3.2. Social Contribution
3.2.1. Value to Customer & Users
3.2.2. Employee Rights and Interests
3.2.3. Business Partner
3.2.4. Safe Operations
3.2.5. Contribution for Public Good
3.3. Environmental
Contribution
3.3.1. Environmental Management
3.3.2. Utilization of Natural Resources
3.3.3. Pollution Prevention and Control
3.3.4. Ecology and Climate

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
15
Goals (SDGs) adopted by the United Nations in 2015. The model is based upon the extent of
data that are regularly published and readily available about listed companies in China.
Organisation Model/ Rating/ Standard
1 MSCI MSCI ESG Rating
2 FTSE Russell FTSE ESG Rating
3 Thomson Reuters Thomson Reuters ESG Score
4 S&P S&P ESG Rating
5 KLD KLD ESG Rating Qualitative
6 KLD KLD ESG Rating Exclusion
7 RobecoSAM (Dow Jones) SAM Corporate Sustainability Assessment
8 GRI GRI Guidelines
9 Bloomberg Bloomberg ESG Rating
10
Syntao Syntao ESG
11
ISO ISO 26000
12
Chinese Academy of Social Sciences (CASS) CASS-CSR
13
Just Capital Just Capital 100
14
SASB SASB Standards
15
Sustainalytics ESG Risk Rating
16
Refinitiv Refinitv ESG Score
17
ECPI ECPI ESG Rating
18
Vigeo-EIRIS ESG Assessment
19
ISS-oekom ISS-oekom Corporate Rating
Figure 3: References and precedents used by CASVI in developing the 3A model
Note. Source: CASVI
In comparison with other internationally used sustainability assessment frameworks,
the CASVI-3A model departs from the usual convention of strictly organising sustainability
data or ratings under the headings of Environment, Social, and Governance (ESG) factors.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
16
However, ESG factors are included within the CASVI-3A model, and the relevant scores
could be extracted and reformatted according to an ESG framework if required.
The Action part of the CASVI-3A model includes Social and Environmental factors,
whilst Governance factors are included within the Approach part of the model. Additionally,
the Action part of the CASVI-3A model includes Economic factors, which are normally
excluded from ESG reporting. Recently however, the International Business Council (2020)
of the World Economic Forum (WEF) also identified Economic factors alongside
Environment, Social, and Governance factors as the four key pillars of sustainable value
creation. The International Business Council (2020) departs from the conventional ESG
terminology, and instead, uses the categories of Principles of Governance, Planet, People,
and Prosperity. Nevertheless, the effect is the same, which is to emphasise that the four
pillars are aligned with the essential elements of the SDGs. Similarly, the Positive Impact unit
of the United Nations Environment Programme Finance Initiative (UNEP FI) has aligned its
22 Impact Areas with the SDGs and arranged them according to the pillars of Social,
Environmental, Economic, and Governance (Positive Impact, 2020).
It is noteworthy that these three separate initiatives have followed unrelated processes
and yet all independently adopt a similar high level structuring of sustainability information
in terms of the conventional ESG factors plus another E for Economic factors i.e. ESG+E,
even if the detailed interpretation of economic factors is not always exactly the same.
However, the type of information that these various initiatives include under the
heading of economic factors is not always the same. Some, like CASVI and the International
Business Council, tend to include more traditional financial factors, whereas the UNEP FI

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
17
Positive Impact includes socio-economic factors. The merits of distinguishing between socio-
economic factors and financial factors, and therefore the separation of the data, has also been
emphasised by the Social Progress Imperative. Precisely because of this strict separation, the
Social Progress Imperative is able to undertake more revealing and insightful regression
analysis, for example, to compare the scores from their Social Progress Index with Gross
Domestic Product (GDP) data (Green, 2015).
Such a broader interpretation of economic factors also resonates positively with Triple
Bottom Line accounting, which seeks to encourage “businesses to track and manage
economic (not just financial), social, and environmental value” that they create or destroy
through their activities (Elkington, 2018). Elkington (2018) emphasises the distinction
between narrow financial factors (such as turnover and profit) from broader economic factors
(such as circular economic models and inclusive economic prosperity). Even though Triple
Bottom Line is framed as an accounting or reporting framework for non-financial value, it
intends to inspire a more aspirational ideal of an economic model that is inherently
regenerative rather than only extractive. This regenerative paradigm is akin to a balanced
natural system that is in a continual cycle of renewal, where harmony is maintained because
the outputs and waste of one part become the inputs and essential nutrients of another part.
At least at a high level, it is encouraging that there is convergence in information
structuring by CASVI, WEF International Business Council, UNEP FI Positive Impact, and
Triple Bottom Line. Although each initiative has distinctly different strategic objectives,
methodological intensity, and ultimately philosophical approach that would affect analytical
bias – which is for a separate discussion – the common organising structure lends itself well

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
18
to evolving a common data architecture for organising and reporting information about
sustainability issues and non-financial impacts.
Furthermore, Elkington (2018) argues that “radical intent” is required to achieve a
regenerative economic model. Indeed, the idea of impact investing is typically defined by the
“intention to generate positive, measurable social and environmental impact alongside a
financial return” (GIIN, 2019). From Figure 2 (above), it can be seen that the CASVI-3A
model does assess factors within the Aim (Driving Force) category that correspond to the
strategic intentions and purposefulness of an organisation. By explicitly including these
factors in the scoring model, CASVI arguably differentiates itself from other approaches that
only implicitly require intentionality as a definitional characteristic.
Another feature of the CASVI-3A model that is relevant to the international discourse
is the Approach (Innovation Force) category. The Approach category acknowledges how
much effort and resources a company invests in transforming itself towards being more
sustainable. This is an important signal to help redress a cognitive dissonance that can appear
between the short-term requirements of quarterly financial reporting and the longer-term
priorities of the sustainability agenda. The investment community and capital markets
typically do not yet reward companies for making long-term strategic investments in

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
19
transforming their organisations towards being more sustainable. These efforts are often
treated as unproductive costs instead potential drivers of value.18
After high-level analysis, the CASVI-3A model compares favourably with the global
best practice and leading trends. The model includes discrete scoring in terms of
environmental, social, and economic factors (within the Action part of the model) as well as
governance factors (within the Approach part of the model). Moreover, the model goes
further than many of CASVI’s contemporaries by scoring companies in terms of their
intentionality (Aim) and how they go about being more sustainable (Approach). Seen from
another perspective, the Aim and Approach categories seek to offer forward looking signals
about non-financial impacts, whereas the Action category considers historic performance as it
actually occurred.
A critical observer may note that it is difficult to isolate individual factors within the
CASVI-3A model, for example governance factors, and extract them from the scores in order
to compare on a like-for-like basis with the related scores provided by other sustainability
rating agencies; or that the financial and socio-economic factors could be more clearly
delineated into two separate sub-categories should a data consumer require that distinction.
18 These arguments were made by panelists in 2019 when the author moderated a discussion at a
conference in Geneva. Panelists represented multi-national corporations and capital markets
practitioners headquartered in Switzerland and Europe. The conference was held under Chatham
House Rules, which requires that the identity or the affiliation of the speaker(s) may not be revealed.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
20
This presents an opportunity for CASVI. There is an increasing trend for ESG specialists to
position themselves as both a provider of a curated sustainability ratings (or scores) as well as
to meet the rising demand for raw data at different levels of detail and granularity (Foubert,
2020). The added-value benefit of providing raw data to the market is that the consumers of
data can then aggregate and customise their collection of data in a way that is relevant for
their own decision-making purposes.
Another challenge that CASVI faces, as do all ESG data and rating organisations, is
that much of their raw data gathering and compiling is still very labour intensive and
therefore costly, as well as often being dependent upon the interpretation of junior analysts or
the limited time availability of senior analysts. With the advent of new digital technologies,
there is increasing potential to overcome these challenges by harnessing new developments in
web tools, machine learning, natural language processing, and artificial intelligence to
increase coverage and improve methodologies (Foubert, 2020).

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
21
4. Potential to Disrupt the Provision of Non-Financial Information
This section considers again the broader global context, and builds upon the analytical
framework presented in Section 2 by the Aggregate Confusion Project at MIT (Berg et al.,
2020). This framework categorised the different ESG assessment methodologies, and the
resulting divergence of ESG scores, in terms of scope, measurement, and weights.
In separate research, it has been suggested that to compute impact data efficiently, the
data should be delineated according to three layers of complexity (Richter & Schramade,
2019). Each layer corresponds to increasing levels of philosophical complexity as well as
information complexity. This delineation can be aligned with the framework presented by
Berg et al. (2020) by re-ordering the categories as follows, with complexity increasing from
top to bottom:
1. Measurement – Objective and incontrovertible raw data that is sufficiently
granular and well-structured [1]
2. Scope – Methodological frameworks, accounting rules, or analytical processes
that aggregate and organise the raw data in a way that corresponds with the
requirements and mandate of the data consumer [n]
3. Weights – Opinion, value judgements, and calibration of the organised data in
a way that represents individual stakeholder interpretations of the reality they
experience [∞]
It can be argued that for every single raw data point of measurement (represented by
1 above), there might be several methodological approaches that could be materially relevant

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
22
based upon the scope of analysis (represented by n above), and for every methodological
approach there would theoretically be an infinite number of weights that can be assigned for
each potential interpretation by every individual stakeholder, whether they are an individual
person, organisation, or another entity (represented by ∞ above). For the sake of simplicity,
this three layer information construct will be referred to as the “1n∞” model (pronounced
“innate” model19).
This newly proposed 1n∞ model for impact data aligns with the longer standing
arguments of Gray et al. (1997) who developed a comparable schema when considering the
potential conflicts between the requirements for polyvocal voices in social accounting and the
likely desirability to business organisations of non-financial data. This tension inherent within
the 1n∞ model has also been emphasised by Nicholls (2018) who, in paraphrasing the
philosophers Habermas and Foucault, suggests that “facticity and validity” need to be
blended in non-financial accounting. In other words, that factual correctness needs to be
balanced with the ethical perspective and corresponding interpretative bias of individuals
(people or organisations).
A pragmatic approach to managing this dichotomy is presented by Ruff and Olsen
(2018), who suggest that the notion of “bounded flexibility” can serve as “a middle ground
19 “Innate” also refers to an inherent attribute or essential character of something.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
23
between ‘anything goes’ and ‘only one right way’. This approach creates comparability by
focusing on the commonality of the construct itself, rather than on differences in the
indicators used to define and measure it”. Ruff and Olsen (2018) argue that successful
standards strike a balance between “uniformity (one size fits all)” and “relevance (customized
to specific needs)”.
It is suggested that the 1n∞ model offers a useful data science construct for managing
the philosophical tension between facticity and validity, whilst at the same time harnessing
the practicality of bounded flexibility. It is posited that data providers may find the 1n∞
model helpful to identify where they can focus their research and development, which will
enable them to build more customised and cost-effective solutions. Ultimately, these
solutions should support end-users undertake more meaningful analysis of impact and
sustainability issues.
It is anticipated that the 1n∞ model could also inform new business opportunities for
a next-generation of data providers to better serve end-user needs regarding sustainability
issues, non-financial impact, and ESG factors. Currently this data market is very congested
with bundled offerings that aggregate data into summary scores or ratings. These combine the
source data via proprietary data handling processes in ways that are often not transparent, nor
do they allow end users to easily disaggregate the underlying source data. The results are
inherently biased (Wilton, 2020) and consequently present conflicting signals (GPIF, 2017;
Berg et al. 2019).

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
24
Wilton (2020) argues that the bundling of sustainability information and the “search
for the ultimate [ESG] methodology may be a journey up a blind alley”, asserting that
investors primarily need “better raw data [unprocessed and unweighted] combined with
flexible investor-driven analytics”. This should be made available as a public good, which
can be achieved by updating regulations, “accounting rules, and stock exchange listing
requirements, so that the analysis … can develop in a manner analogous to equity analysis”
(Wilton, 2020). The reason, according to Wilton (2020), is that the degree of uncertainty
associated with non-financial impacts is more analogous to equity analysis than debt analysis.
Equity values can be affected by any number of factors and therefore analysts typically use a
myriad data sources to inform their opinion. Whereas debt analysts rely heavily on the clarity
of concise credit ratings to determine the probability that a future cashflow will be paid
according to the predefined contractual terms.
The bundled approach (providing a single composite ESG rating, or summary ratings
for E, S, and G factors) leaves users no choice but to accept the underlying methodological
nuances, analytical assumptions, and weights of the provider – without any opportunity to
check the quality of the source data nor compare raw data from different providers on a like-
for-like basis. This is not necessarily the fault of existing ESG rating providers. They are
limited by the way source data can be obtained, which is usually gleaned from reading what
companies have written in their reports, or submitted via manual questionnaires to the ESG
rating provider. Unfortunately this approach captures data indirectly, resulting in second-hand
information. In an ideal world, data would be sourced directly via raw data feeds from the
management information systems (MIS) of companies. This would reduce the laborious

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
25
efforts for both the companies providing information as well as the analysts trying to make
sense of the information.
In recent years, some ESG rating providers have started to provide unbundled ESG
data (i.e. disaggregated data) alongside bundled ESG scores or ratings, which confirms that
there is a demand for more granular data. However, based upon the current operating models
of ESG data providers, it may be difficult for them to take advantage of this trend without
incurring prohibitive additional costs of personnel time, technology, and specialised expertise
in providing the reliable unweighted data that users increasingly require. Therefore the need
for more granular unbiased data may create alternative market opportunities for innovative
data intermediaries to specialise in providing certain types of raw data. The 1n∞ model
shows how these raw unweighted data can in turn be aggregated, organised, and weighted in
any number or ways, allowing end users to customise or enrich the data according to their
specific needs.
This should not be confused with the existing process whereby some ESG data
specialists combine the data feeds from several existing ESG rating providers in order to try
extract more refined signals. If the existing data sources already embed assumptions and
biases, which they generally do, then arguably the results of further analysis based upon these
data feeds will be either distorted or vague (unless the biases and assumptions can somehow
be fully reversed through advanced data science techniques). Some end users may find value
and success in these approaches, although self-evidently sophisticated approaches like this
could be expensive.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
26
It is envisaged that these diverse needs could be better served by adopting the data
science construct of the 1n∞ model. This will allow some providers to specialise and operate
within a niche to address one part of the market, whilst others may choose to provide an all-
encompassing (bundled) offering across all three tiers:
1. Provide unbundled raw data with an emphasis on objectivity, accuracy,
verifiability, and granularity.
2. Accounting and reporting frameworks that organise data according to relevant
scope and mandate, in ways that avoid or minimise any initial conditions bias as
well as interpretative value judgements as much as possible. This approach can be
augmented with independent audit to reduce the unambiguity of accounts and
reports.
3. End-user tools that allow individual users to codify their values, priorities, and
preferences in order to assign their own personal weights to the results of tiers 1
and 2 above. This will empower them with more customisable analytical
solutions to interpret non-financial information in a more meaningful way, which
is therefore likely to be more closely aligned with their world-view, and therefore
more contextually relevant to them. Further innovation within this layer is likely
to support the development of highly customisable products and services, as well
as FinTech solutions focused on sustainability and non-financial impacts.
It is asserted that if a market for data can be delineated in such way, and if a vibrant
ecosystem of mutually co-dependent data providers can be encouraged, then it can be
anticipated that the dynamics of collaboration and competition will mutually reinforce each

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
27
other to create stronger incentives for data interoperability. In the words of Deng et al. (2019)
describing the vision for an Internet of Impact: “When data sets from different sources can be
accessed, processed, and integrated without losing meaning, data becomes interoperable,
which in turn unlocks massive network effects”.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
28
5. A Different Technological Mindset Could Unlock New Opportunities for Agile
Data Providers
“When the Facts Change, I Change My Mind. What Do You Do?” 20
Previously, this paper suggested that more meaningful analysis of non-financial issues
could be facilitated by a different data science construct. This is referred to as the 1n∞
model, which is summarised below. Underlying this model is the working hypothesis that a
paradigm shift is required for how non-financial information is sourced, processed, and
consumed21. This conjecture is presented on the basis of the industry-wide systemic
contradictions evidenced by Berg et al. (2020) in their Aggregate Confusion Project, and
identified separately by GPIF (2017) and Wilton (2019).
The following diagram shows eloquently how the current paradigm of indirect data
collection leads to data siloes that are fragmented, duplicated, inconsistent, and potentially
obsolete if the data change over time – ultimately, resulting in data are “rendered redundant”
(Burgess et al. 2018).
20 This quote is typically attributed to the economist John Maynard Keynes, although according to
Quote Investigator (2011) the quote originates from another economist Paul Samuelson, who was
awarded the 1970 Nobel Prize in economics.
21 These highlighted terms have been consciously used to correspond with the three categories within
the Berg et al. (2020) framework, namely linking sourced to measurement, processed to scope, and
consumed to weights.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
29
Figure 4: Fragmentation can lead to redundant data
Note. From “From Billions to Trillions” by Cameron Burgess, Astrid Scholz, Arthur Wood,
& Audrey Selian p. 22. https://sphaera.world/wp-
content/uploads/2018/03/Billions%20to%20Trillions.pdf
The debilitating nature of this situation can be demonstrated with the following real-
world scenario. One of the five largest beer companies in the world
22
recently started
investing in a new data system to manage their total carbon dioxide (CO
2
) footprint for each
of their different products. This includes their own direct emissions, as well as their indirect
22 The name of company has been anonymised because the information was obtained by the author
during a private meeting in 2019 with the Head of Sustainability at the company.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
30
emissions from the energy they consume, and the CO2 embodied across their entire supply
chain. Considering the laborious nature of collecting all these data, it was only realistic for
the company to try reduce the ultimate reporting delay from three years to one year
retrospectively. It is not possible for management teams to make meaningful decisions about
non-financial issues using such non-current data, especially in a just-in-time production
context where much decision-making is already done using real-time data. If these challenges
exist within companies that have full control over their internal operations, and strong
influence over their supply chains, then the challenges are amplified for independent
assessors or practitioners in the financial sector.
A substantial challenge with accurately reporting CO2 emissions is the problem of
double counting (Hoepner & Yu, 2016). This is especially compounded when standards like
the Greenhouse Gas Protocol (GHG)23 require companies to report their direct emissions as
well as all their indirect emissions from the energy they consume and the emissions embodied
within their supply chain. Hoepner & Yu (2016) argue that “it becomes a substantial
challenge for anyone aiming to analyse an aggregation of organizations, either in the supply
chain or in terms of an investment portfolio”. Hoepner & Yu (2016) conclude that “the way
carbon emissions are counted (at the moment) is overly complicated” and instead the focus
23 https://ghgprotocol.org/

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
31
should be on those emissions that “are actually physically produced and emitted to the
atmosphere” by an organisation.
Changing the prism though which a problem is viewed can often unlock breakthrough
solutions. According to Haefeker (2019), the underlying cause of these problems is that the
analogy for digital data is wrong. People often think of digital data stores as if they are
electronic versions of paper filing cabinets. They assume that to have their own access to the
data, they need to make a copy of the original data and transfer the copy into their own data
store. This is then repeated until they accumulate all the data they need in one place.
Paradoxically, this approach leads to the fragmentation and inefficiency identified by Burgess
et al. (2018) because the act of copying data breaks the link to the original source. This is
arguably a major cause of the problems regarding non-financial data that were identified by
Berg et al. (2020), GPIF (2017), and Wilton (2019), as well as the errors of double counting
identified by Hoepner & Yu (2016).
Consider again the beer company scenario described previously, the original sources
of the raw data required in calculating the CO2 footprint are by their nature distributed across
a diverse supply chain. Now expand the complexity of this challenge to the full range of non-
financial issues that need to be monitored and managed by any organisation, such as
environmental, societal, governance, and economic factors. Haefeker (2019) argues that new

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
32
web technologies such as Linked Data24 protocols should be used to work with such
information that is distributed at source. Instead of having to repetitively copy troves of data
from multiple sources, especially those that change over time, Linked Data protocols allow
web-based data systems to link once to a required source, and then read the data directly in
real-time whenever it is required (subject to access controls and assurances of availability).
This is similar to how, for example, web-based navigation software is able to recommend
travel options based upon the actually current times of buses, trains, airplanes, boats as well
as real-time traffic congestion, road closures, and so forth.
This works well when the sources of the data are known, and can be linked to directly.
However, in the context of non-financial information about companies and their products,
goods, or services, it is not always known where all relevant sources of non-financial data are
located. It is a documented problem that “the lack of reliable query solutions for [distributed]
live public data … prevents their use in real-world applications” (Verborgh et al., 2014).
In 2019 a technical experiment was undertaken by a technology collective (Instans,
2020) to prototype a “federated search engine for distributed data”. They were successful in
doing “search queries across multiple distributed data pods in real-time”.
24 Linked Data is a term coined by Tim Berners-Lee (2006) and refers to a set of best practices for
publishing structured data on the Semantic Web, see https://www.w3.org/DesignIssues/LinkedData

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
33
The technical challenges of working with distributed data appear to be solvable and,
according to Instans (2020), should enable “third party applications to work with distributed
data sets as if they are an integrated database”. However, a final challenge remains, which is
to align these technical solutions with viable business models that will enable a next
generation of data providers to evolve and scale in an economically sustainable way.
The foundational layer of the 1n∞ model, presented previously in this paper, calls for
unbundled raw data with an emphasis on objectivity, accuracy, verifiability, and granularity.
This implies that the next generation of impact data providers will potentially need to develop
their business models around very small amounts of data, with correspondingly small
accounting units and payments. The connection between a distributed web of small amounts
of data and the need for a viable commercial model has already been anticipated. In the
1960s, Ted Nelson (2015) coined the term “micropayment” to refer to the idea that “each
portion [of data] is sold from where it sits in its original content” and that each “portion, no
matter how small, is paid for according to size (the number of characters) and the price per
character. … There need be no minimum download”.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
34
6. Conclusion
This paper covers an array of issues that allow a future to be imagined in which
meaningful analysis of non-financial information is easy, accessible, and real-time.
The paper started with a commentary on global best practice in non-financial
reporting, which identified operational challenges as well as theoretical arguments for how to
improve the management of non-financial information. A high-level comparative review was
undertaken of the Aim-Approach-Action model (3A model) developed by the China Alliance
of Social Value Investment (CASVI), which identified challenges and opportunities that all
sustainability rating organisations or data providers are confronted with. Reflecting upon the
global situation as well as the specific case study, it was found that the roots of confusion lie
in a known tension between the facticity (factual accuracy) and validity (interpretative value
judgement) of non-financial data. The 1n∞ model (pronounced “innate” model25) was
proposed as a potential data science approach for managing the resulting dichotomy,
suggesting it could serve as a guide for developing practical solutions. By framing the
underlying challenge as a data science problem, several known approaches or suitable
technologies were identified that appear to offer practical and implementable solutions.
25 “Innate” also refers to an inherent attribute or essential character of something.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
35
The 1n∞ model was evolved throughout this paper and is presented primarily as a
meta framework for conceptualising the whole industry of non-financial data. It is also
posited that the 1n∞ model could enable innovation and disruption within the industry by
informing the design of more competitive market offerings and business models. The key
elements are summarised below into three discrete tiers, with complexity increasing from top
to bottom in terms of philosophical composition, data model design, and likely interpretive
variance.
Tier Expected
interpretative
variance
Primary
functional
attributes
Market
offering
Means of
authentication
1. Sourcing
of raw data
[1] Single
incontrovertible
interpretation
Measurement
and data capture
Unbundled/
disaggregated
raw data
Can be verified
as either correct
or incorrect
2. Processing
of sourced
data
[n] Several
interpretations
based upon
standardised
rules
Frameworks
reflecting
specific scope
and mandate
requirements
Organised data
for reporting in
consistent and
comparative
formats
Can be
independently
audited for
accuracy and
compliance
3. Consuming
of processed
data
[∞] An infinite
number of
theoretically
possible
interpretations
based upon
individual
values,
priorities,
ethics, and
moral context
Weights, filters,
and algorithms
that can
represent
multiple
concurrent
perspectives
Customisable
analytical
solutions for
each unique
decision-
making context
Can be
modelled to
compare
predictions and
simulations
with actually
resulting
behaviours and
effects
Figure 5: Summarised 1n∞ model (pronounced “innate” model)
Note. Source: Author

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
36
The 1n∞ model is not suggested as a literal structure for the non-financial data
ecosystem, but rather to help organisations identify the role, or roles, they could play within
it. This includes companies in the real economy who are the primary sources of impact
creation, positive or negative, and therefore are best placed to also be the sources of raw data
about their impacts, ideally obtained directly from their management information systems
(MIS) in real-time without the need for retrospective questionnaires or analysist trawling
through non machine-readable reports. It also includes the ESG rating providers who
currently do the tedious work of collecting data to feed into their analytical processes, as well
as regulators and standard setting organisations, and the technology providers that can help
facilitate easier and better information exchange.
Ultimately the 1n∞ model reveals the potential for data providers to differentiate
themselves within a growing and diversifying market. Some may specialise according to a
specific tier within the 1n∞ model whilst others may specialise according to their domain
knowledge about certain types of non-financial issues. This increased diversification of data
providers and intermediaries, combined with interoperability protocols to facilitate easier data
exchange, would allow organisations to expand their access to data in a way that is more cost
effective than sourcing data directly themselves. Alternatively, some data providers may
choose to provide integrated solutions that span across all three tiers (similar to the current
market). Ultimately, the data market would be more diverse and better serve the consumers of
data. It would do this by offering greater choice between data that either represent factual
accuracy or interpretative value judgements, without forcing consumers to accept data that
conflate the two or embody conflicting world views. The modular nature of the 1n∞ model
means that each layer is additive, which will allow greater reuse of data, as well as more

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
37
relevant data that can be organised and consumed more efficiently. Perhaps
counterintuitively, this is likely to also result in less fragmentation, less duplication, and less
redundancy.
It is recognised that the implementation of the 1n∞ model may cause disruption
within the existing industry of non-financial information providers. The paper therefore
considers some of the challenges that data providers may be confronted with, and eludes to
how forward thinking data providers could take advantage of the associated opportunities.
Proactive regulators or legislators may also find some of the arguments useful in deciding
whether certain types of non-financial data should be treated as a public good, and they could
require companies to make these data available as part of their statutory reporting obligations.
Finally, it is posited that agile data providers, who adopt the 1n∞ model, will be able provide
next-generation solutions that could be more relevant, customisable, and cost effective than
current offerings, and which would support end-users in undertaking more meaningful
analysis of impact and sustainability issues.

38
References
Berg, F. Koelbel, JF. Rigabon, R. (17 May 2020). Aggregate Confusion: The Divergence of
ESG Ratings. MIT Sloan School of Management. Retrieved from
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3438533
Berners-Lee, T. (27 Jul 2006). Linked Data. W3C. Retrieved from
https://www.w3.org/DesignIssues/LinkedData
Burgess, C., Scholz, A., Wood, A. & Selian, A. (Mar 2018). From Billions to Trillions.
Retrieved from https://sphaera.world/wp-
content/uploads/2018/03/Billions%20to%20Trillions.pdf
Deng, K., Lewkowitz, M., Ruff, K. & Ryan, A. (April 2019). Pathways Towards Data
Interoperability for the Sustainable Development Goals in Canada. MaRS.
https://www.marsdd.com/wp-content/uploads/2019/05/Pathways-Towards-Data-
Interoperability-for-the-Sustainable-Development-Goals-in-Canada.pdf
Elkington, J. (25 Jun 2018). 25 Years Ago I Coined the Phrase “Triple Bottom Line.” Here’s
Why It’s Time to Rethink It. Harvard Business Review. Retrieved from
https://hbr.org/2018/06/25-years-ago-i-coined-the-phrase-triple-bottom-line-heres-
why-im-giving-up-on-it
European Commission. (18 Jun 2020). Sustainable Finance: Commission welcomes the
adoption by the European Parliament of the Taxonomy Regulation. European
Commission. Retrieved from
https://ec.europa.eu/commission/presscorner/detail/en/ip_20_1112
Foubert, A-L. (9 Mar 2020). ESG Data Market: No Stopping Its Rise Now. Opimas.
Retrieved from http://www.opimas.com/research/547/detail/

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
39
Global Impact Investing Network (GIIN). (3 April 2019). What is impact investing?. GIIN.
Retrieved from https://thegiin.org/impact-investing/need-to-know/#what-is-impact-
investing
GPIF (Japanese Government Pension Investment Fund). (3 Jul 2017). GPIF Selected ESG
Indices. Retrieved from
https://www.gpif.go.jp/en/investment/pdf/ESG_indices_selected.pdf
Gray, R. Dey, C. Owen, D. Evans, R. & Zadek, S. (Feb 1997). Struggling with the praxis of
social accounting – stakeholders, accountability, audits and procedures. Accounting,
Auditing & Accountability Journal, Vol. 10 No. 3. p 325-364.
Green, M. (22 Apr 2015). Why we shouldn’t judge a country by its GDP. TED Conferences.
Retrieved from https://ideas.ted.com/why-we-shouldnt-judge-a-country-by-its-gdp/
Haefeker, W. (23 Sep 2019). Creating a common European data space for environmental and
climate-related data. Comments made at a workshop organised by DG CNECT of the
European Commission, Brussels.
Hoepner, AGF. & Yu, PS. (13 Jan 2016). Science Based Targets Without Science Based
Disclosure? Towards a Complete Carbon Data Science. CSEAR 2017 paper.
Retrieved from https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2715101
Iacurci, G. (14 Jan 2020). Money moving into environmental funds shatters previous record.
CNBC. Retrieved from https://www.cnbc.com/2020/01/14/esg-funds-see-record-
inflows-in-2019.html
Instans. (Accessed Jul 2020). Instans. Retrieved from https://instans.net
International Business Council. Deloitte. EY. KPMG. & PwC. (Jan 2020). Toward Common
Metrics and Consistent Reporting of Sustainable Value Creation. World Economic

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
40
Forum. Retrieved from https://www.weforum.org/whitepapers/toward-common-
metrics-and-consistent-reporting-of-sustainable-value-creation
Khemani, RS. & Shapiro, DM. (1993). Glossary of Industrial Organisation Economics and
Competition Law. OECD, Directorate for Financial, Fiscal and Enterprise Affairs.
Retrieved from http://www.oecd.org/regreform/sectors/2376087.pdf
Ma, W. (2019). 2018, Discovering “SV 99” in China, A-share Listed Company Sustainable
Development Value Assessment. CASVI. Retrieved from https://www.casvi.org/en/h-
col-103.html
Mirchandani, B. (29 May 2020). What To Make Of The SEC’s Warnings And
Recommendations For ESG Investing. Forbes. Retrieved from https://www-forbes-
com.cdn.ampproject.org/c/s/www.forbes.com/sites/bhaktimirchandani/2020/05/29/wh
at-to-make-of-the-secs-warnings-on-esg-ratings-and-recommendations-for-esg-
disclosures/amp/
Nelson, TH. (2 Sep 2015). A Thought for Your Pennies: Micropayment and the Liberation of
Content. Project Xanadu. Retrieved from http://transcopyright.org/hcoinRemarks-
D28.html
Nicholls, A. (26 Mar 2018). A General Theory of Social Impact Accounting: Materiality,
Uncertainty and Empowerment. Journal of Social Entrepreneurship, 9:2, p. 132-153.
Retrieved from https://doi.org/10.1080/19420676.2018.1452785
Quote Investigator. (22 Jul 2011). When the Facts Change, I Change My Mind. What Do
You Do, Sir?. Quote Investigator. Retrieved from
https://quoteinvestigator.com/2011/07/22/keynes-change-mind/
Richter, KH. & Schramade, W. (5 Jul 2019). Towards an Internet of Impact. Unpublished
manuscript.

Richter, KH. (8 January 2021). Pain spots and opportunities regarding
environment, social, and
governance (ESG) data
.
41
Rigobon, R. (Accessed Jul 2020). Aggregate Confusion Project. Retrieved from
http://web.mit.edu/rigobon/www/aggregate-confusion-project.html
Ruff, K. & Olsen, S. (2018). The Need for Analysts in Social Impact Measurement: How
Evaluators Can Help. American Journal of Evaluation. Retrieved from
https://www.researchgate.net/publication/326385129_The_Need_for_Analysts_in_So
cial_Impact_Measurement_How_Evaluators_Can_Help
Positive Impact. (Mar 2020). Impact Mapping (Excel sheet: Impact Areas). United Nations
Environment Programme Finance Initiative. Retrieved from
https://www.unepfi.org/wordpress/wp-content/uploads/2020/07/PII-Impact-
Mappings-March-2020.xlsx
Verborgh R. Hartig, O., De Meester, B., Haesendonck, G., De Vocht, L., Vander Sande, M.,
Cyganiak, R., Colpaert, P., Mannens, E., & Van de Walle, R. (2014). Querying
Datasets on the Web with High Availability. In: Mika P. et al. (eds) The Semantic
Web – ISWC 2014. ISWC 2014. Lecture Notes in Computer Science, vol 8796.
Springer. Retrieved from https://link.springer.com/chapter/10.1007/978-3-319-11964-
9_12
Wilton, DE. (Sep 2019). Pricing impact: extending impact investing to price externalities and
lower the cost of capital to impactful investors. Zheng Partners LLC. Retrieved from
https://zhengpartners.co
Wilton, DE. (May 2020). Measuring Sustainability: Data + Flexible Analytical Tools versus
Ratings. Zheng Partners LLC. Retrieved from https://zhengpartners.co

Appendix 1 – CASVI-3A Model
Sub-model for Social Value 99 Assessment (2020)
Tier One (3)
Tier Two (9) Tier Three (28) Tier Four (59)
Indicator
Weight
Indicator
Weight
Indicator
Weight
Indicator
Weight
1. AIM
Driving Force
10%
1.1 Value Driven 4% 1.1.1 Core Values 2% 1.1.1.1 Mission, Vision, Purpose 2%
1.1.2 Business Ethics 2% 1.1.2.1 Values, Operation Principle 2%
1.2 Strategic Driven 3% 1.2.1 Strategic Objectives 1% 1.2.1.1 Sustainable Development Strategic Objectives 1%
1.2.2 Strategic Planning 2% 1.2.2.1 Medium and Long-term Strategic Planning 2%
1.3 Business Driven 3% 1.3.1 Business Positioning 2% 1.3.1.1 Main Business Scope 2%
1.3.2 Target Customers & Users 1% 1.3.2.1 Structure of Customers & Users 1%
2. APPROACH
Innovation Force
30%
2.1 Technical
Innovation 8%
2.1.1 Research and Development 4%
2.1.1.1 R&D Investment 1%
2.1.1.2 R&D Output 1%
2.1.1.3 R&D Efficiency 1%
2.1.1.4 R&D Quality 1%
2.1.2 Products & Services 4% 2.1.2.1 Product & Service Disruptive Innovation 2%
2.1.2.2 Social Value Driven Products/Services Innovation 2%
2.2 Model
Innovation 6%
2.2.1 Business Model 2% 2.2.1.1 Profit Model 1%
2.2.1.2 Operation Model 1%
2.2.2 Industry Impact 4% 2.2.2.1 Setting the Industry Standard 2%
2.2.2.2 Industrial Transformation and Upgrading 2%
2.3 Management
Innovation 16%
2.3.1 Corporate Governance 6%
2.3.1.1 Governance of Directors, Supervisors, and Senior Executives 2%
2.3.1.2 Investor Relations Management 2%
2.3.1.3 Stakeholder Identification and Engagement 2%
2.3.2 Information Disclosure 4% 2.3.2.1 Financial Information Disclosure 2%
2.3.2.2 Non-financial Information Disclosure 2%
2.3.3 Risk Control 4% 2.3.3.1 Internal Control System 2%
2.3.3.2 Risk Control System 2%
2.3.4 Incentive Mechanism 2% 2.3.4.1 Incentive to Award Enterprise Innovation 1%
2.3.4.2 Employee Stock Option Plan 1%
3. ACTION
Transformation Force
60%
3.1. Economic
Contribution 30%
3.1.1. Profitability 5% 3.1.1.1. Rate of Return on Equity 3%
3.1.1.2. Operating Profit Ratio 2%
3.1.2. Operation Efficiency 5% 3.1.2.1. Total Assets Turnover 3%
3.1.2.2. Receivables Turnover Ratio 2%
3.1.3. Solvency 7%
3.1.3.1. Current ratio 2%
3.1.3.2. Debt to Assets Ratio 2%
3.1.3.3. Net asset 3%
3.1.4. Growth Capability 5% 3.1.4.1. Compound Annual Growth Rate of Revenue in Past 3 Years 3%
3.1.4.2. Compound Annual Growth Rate of Net Assets in Past 3 Years 2%
3.1.5. Financial Contribution 8%
3.1.5.1. Total Tax Paid 3%
3.1.5.2. Dividend Yield Ratio 2%
3.1.5.3. Market Capitalization 3%
3.2. Social
Contribution 15%
3.2.1. Value to Customer & Users 3% 3.2.1.1. Quality Management System 2%
3.2.1.2. Customer Satisfaction 1%
3.2.2. Employee Rights and Interests 3%
3.2.2.1. Fair Employment Policy and Outcome 1%
3.2.2.2. Employee Rights and Interests Protection 1%
3.2.2.3. Employee Career Development 1%
3.2.3. Business Partner 3% 3.2.3.1. Compliance Operations 2%
3.2.3.2. Supply Chain Management Measures and Outcome 1%
3.2.4. Safe Operations 3% 3.2.4.1. Security Management System 2%
3.2.4.2. Safe Operations and Outcome 1%
3.2.5. Contribution for Public Good 3% 3.2.5.1. Donations for Public Good 2%
3.2.5.2. Community Capacity Building 1%
3.3. Environmental
Contribution 15%
3.3.1. Environmental Management 7%
3.3.1.1. Environmental Management System 2%
3.3.1.2. Input towards Environmental Protection 2%
3.3.1.3. Environmental Violations and Penalties 2%
3.3.1.4. Environmental Friendly Purchasing Policy, Measures and Outcome 1%
3.3.2. Utilization of Natural Resources 4%
3.3.2.1. Integrated Energy Consumption Management and Outcome 1%
3.3.2.2. Water Resources Management and Outcomes 1%
3.3.2.3. Material Consumption Management and Outcomes 1%
3.3.2.4. Environmental Friendly Management and Outcomes 1%
3.3.3. Pollution Prevention and Control 2% 3.3.3.1. Measures and Outcomes Three Reductions (Wastewater, Emissions & Solid Waste) 2%
3.3.4. Ecology and Climate 2% 3.3.4.1. Measures and Outcomes for Ecological Conservation 1%
3.3.4.2. Measures and Outcomes for Climate Change Mitigation 1%
Source: China Alliance of Social Value Investment (CASVI)
https://www.casvi.org

Appendix 2 – CASVI Exclusion Criteria
Sub-model for Social Value 99 Screening (2020 Revised Version)
No. Tier-1 Indicators Tier-2 Indicators Tier-3 Indicators
1 Prohibited & Restricted Industry
Prohibited Industry
Obsoleted Industry identified in the Catalogue for Guidance of Industrial Structure
Adjustment and regulated by National Development and Reform Commission;
Companies with major business in tobacco manufacturing and sales;
Companies with major business in wild animal trading;
Companies with major business in lottery；
Companies with major business in adult entertainment.
Companies with major business in controversial weapons manufacturing and sales
(controversial weapons include cluster weapons, landmines, biological or chemical weapons,
depleted uranium weapons, white phosphorus weapons or nuclear weapons)
Restricted Industry Restricted industry that has been identified in the Catalogue for Guidance of Industrial
Structure Adjustment and regulated by National Development and Reform Commission
2 False Information Disclosure
False Financial Information Disclosure Severe false financial information disclosure has been verified.
False non-financial Information Disclosure Severe non-financial information disclosure has been verified.
3 Negative Impact to the Economy
Violation of Laws and Regulations
Listed company and its holding subsidiaries (whose revenue accounts for more than 30% of
the consolidated party) has strongly violated laws and regulations in the aspect of economic
responsibility （e.g. tax dodging, tax evasion and fraud)
Audit Report Auditing agency has published non-standard unqualified audit report.
4 Negative Impact to the Society
Violation of Laws and Regulations
List company and its holding subsidiaries (whose revenue accounts for more than 30% of the
consolidated party) have strongly violated the laws and regulations on social responsibility
(e.g. labor right protection, production safety, employee's health and community relations).
Social Influence
Listed company and its holding subsidiaries (whose revenue accounts for more than 30% of
the consolidated party) have caused negative social impact and not responded actively in
terms of social responsibility (e.g. labor right protection, production safety, employee's health
and community relations).
5 Negative Impact to the
Environment
Violation of Laws and Regulations
List company and its holding subsidiaries (whose revenue accounts for more than 30% of the
consolidated party) have strongly violated laws and regulations in the environmental aspects
(e.g. pollutant emission and doing harm to environmental protection.)
Environmental Impact
Listed companies and its holding subsidiaries (whose revenue accounts for more than 30% of
the consolidated party) have caused severe negative environmental impact (e.g. pollutant
emission & doing harm to environment protection) and not tackled the issues actively.
6 Special Treatments
ST and *ST Listed companies that have been suspended.
Violation of Global Agreement & Regulations Strongly violate relevant international conventions, principles and standards that PRC has
signed and joined.
Source: China Alliance of Social Value Investment (CASVI) https://www.casvi.org