ESG Risk AI’s ESG assessment methodology 1

Table of contents
1. Overview of ESG assessments……………………………………………………3
1.1. Scope of ESG assessments ................................................................................................................3
1.2. ESG assessment approach and models ............................................................................................4
1.2.1. Using survey vs disclosures for data and variations in rating scales ...............................................4
1.3. ESG Risk AI’s approach to ESG assessments .....................................................................................5
1.4. What provides predictive power to ESG assessment models ..........................................................5

2. ESG Risk AI’s approach to ESG rating ......................................... 6

2.1. Ratings as a summary of a comprehensive ESG assessment ...........................................................6
2.1.1. Understanding ESG Risk AI’s ESG data taxonomy ............................................................................7
2.2. Primary data sources for ESG Risk AI’s ESG ratings ..........................................................................9
2.3. ESG Risk AI’s ratings and quality process .......................................................................................10
2.3.1. ESG Risk AI’s approach to surveillance and rating updates ............................................................11

3. ESG Risk AI’s scoring methodology .......................................... 12

3.1. Components of ESG Risk AI’s ESG assessment ...............................................................................12
3.1.1. Indicator relevancy and materiality ...............................................................................................12
3.1.2. Accounting for indicator polarity ...................................................................................................15
3.1.3. Scoring each indicator ...................................................................................................................16
3.1.4. Scoring negative news ...................................................................................................................18
3.1.5. Aggregating scores of individual indicators to the overall score using weights ............................19
3.1.6. Assigning ratings based on the overall scores ...............................................................................20
3.1.7. Scoring a company’s ESG disclosures and transparency ......................................................... 21
 ESG Risk AI’s ESG assessment methodology 2

List of figures
Figure 1: ESG Risk AI’s framework for ESG assessments ...............................................................................6
Figure 2: Overview of ESG Risk AI's data taxonomy ......................................................................................7
Figure 3: Visual representation of our data taxonomy .................................................................................8
Figure 4: Five levels of ESG Risk AI’s data taxonomy………………………………………………………………………………. 9
Figure 5: ESG Risk AI’s ratings and quality process………………………………………………………………………………… 10
Figure 6: Example of materiality in financial services industry………………………………………………………………. 15
Figure 7: Understanding polarity of similar indicator……………………………………………………………………………. 16
Figure 8: Scoring the best and worst performers using percentiles………………………………………………………. 18
Figure 9: Impact of negative news on scoring………………………………………………………………………………………. 19
Figure 10: Process of assigning ratings…………………………………………………………………………………………………. 20

List of tables
Table 1: Data sources for our ESG assessments ............................................................................................9
Table 2: Example of score update and rating changes................................................................................11
Table 3: Industry agnostic and industry specific indicators in the Environment category .........................13
Table 4: Industry agnostic and industry specific indicator in the Social category ......................................14
Table 5: Weights based on polarity .............................................................................................................17
Table 6: Mapping scores to rating scale …………………………………………………………………………………………….....21
Table 7: Overall transparency score across hierarchy…………………………………………………………………………….22
 ESG Risk AI’s ESG assessment methodology 3

Investors increasingly acknowledge that using Environmental, Social and Governance (ESG) factors in
investment analysis is beneficial to portfolio selection as there is a strong correlation between ESG and
financial performance of companies. Companies with strong ESG risk management practices are more likely
to drive long-term sustainable performance and shareholder value. Hence, investors are keen to assess the
ESG performance of prospective investment opportunities as well as track the ESG performance of their
existing investments on an ongoing basis.

1.1. Scope of ESG assessments

To assess a company’s ESG risk management, investors need to understand the company’s strategy and
performance on ESG indicators. To provide meaningful insights, ESG assessments have to cover:

1. Details of all environmental risks, for example use of water, energy and natural resources, air emissions,
effluents discharged in water/land, how innovation is embedded in the company’s strategy and how the
strategy translates to superior ESG performance measured by achievement of numerical targets. The
evaluation must consider the materiality of risks, susceptibility of a company to specific ESG risks and
the company’s strategy to manage these risks. The efficacy of a company’s risk management framework
can be assessed by scrutinizing the results of the company’s environment management practices. Usually
the outcomes are evident through reductions in emissions, reduction in waste, better use of water, etc.
and if the results are aligned with the targets the company has set for itself, the ESG risk management
framework can be assumed as effective.
2. Similarly, for social compliance, the evaluation has to examine how a company manages its relationships
with employees, suppliers, customers and communities. For example, does the company take employee
health and safety, career development and labor rights into consideration while developing its policies,
plan location and investment outlays? Is the company engaged in community support and development?
Does it require its supply chain to follow ESG principles? Like in case of environmental assessment, the
social assessment also has to evaluate the materiality of risks, susceptibility to a risk and the company’s
management framework.
 ESG Risk AI’s ESG assessment methodology 4

3. For governance, the evaluation has to cover the board independence, diversity, leadership, executive
pay, audits, internal controls, and shareholder rights. For example, the choice of its board members,
independence, diversity and experience, shareholder rights measured by their ability to vote on
important issues, etc. Again, the evaluation has to cover both materiality of risks, susceptibility to risks
and the company’s risk management framework.

1.2.ESG assessment approach and models

ESG assessments are complex. While reliable evaluation approaches exist, the challenge lies in the lack of
standardized / comparable information that can be used for evaluation as well as a standardized approach of
ESG assessments across providers. Each company publishes information in proprietary formats and each
rating agency uses its own set of criteria to measure ESG performance, increasing the complexity for
investors.

The absence of standardization is partially ameliorated by issuers that align their disclosures to frameworks
and common themes proposed by SEBI and GRI. Today, many companies provide information through various
mandated and voluntary disclosures, with the former covering financial reports and other regulatory filings,
and the latter being investor presentations, social responsibility reports and other ad hoc disclosures.

1.2.1. Using survey vs disclosures for data and variations in rating scales
ESG assessments are always based on information disclosed by issuers, either through specific questionnaire
administered by the assessment agencies and/or analysis of publicly available disclosures available in
sustainability/CSR reports, integrated reports, annual reports and websites.

Investors are looking for standardized, accurate and comparable data and metrics to support investment
decisions, many companies report ESG information inconsistently and in a manner that investors find difficult
to use. To overcome the limitation of inconsistent disclosures, some agencies providing ESG assessments
request companies for information and evaluate companies primarily based on information collected through
questionnaires while supplementing their analysis using publicly available information (CSR reports, annual
reports, news).

Survey-based assessments have the advantage of direct/targeted questions, collection of up-to-date

information and specific answers not requiring interpretation. Drawbacks are the respondent bias,
incomplete responses, dependence on voluntary participation and varying response time frames.

Assessments based on public disclosure undoubtedly have the advantage of transparency since the
information reported is publicly available. The main drawbacks are that the disclosure formats are
inconsistent, information is subject to interpretation and might not fully capture the company’s initiatives.
To overcome these limitations, initiatives like GRI (Global Reporting Initiative) are providing companies with
a global common language to communicate their ESG initiatives and impacts. SEBI in India provides a similar
framework through the Business Responsibility Reporting (BRR) guideline.

Each ESG rating agency uses its own ESG assessment methodology. Some rating agencies provide ESG scores
on a scale of 100, others provide ratings ranging from AAA to C. Updates to the scores or ratings vary with
companies being monitored on an ongoing basis (dynamically) or on weekly or annual basis.
 ESG Risk AI’s ESG assessment methodology 5

1.3. ESG Risk AI’s approach to ESG assessments

ESG Risk Assessments and Insights (AI) provides ESG assessments. Our model and report are designed to help
investors quickly understand issuer’s ESG risk exposure and risk management framework, enabling investors
to directly integrate ESG factors in their portfolio construction and management. While all data used for ESG
assessments provided by ESG Risk AI is collected and analyzed from publicly available sources, ESG Risk AI
also provides companies the option to review data used and make corrections to the data if needed.

ESG Risk AI’s ESG score provides a summary of the company’s ESG strategy, programs/initiatives, results and
negative news across 19 themes including energy, emissions, water, environmental management, ESG
reporting, human rights, community, supply chain, shareholders’ rights, among others. The ESG scores are
based on a wide range of ~739 data points and 525 indicators that have been selected and assigned weights
based on their materiality and relevance to specific industries.

Indicators are weighted, normalized and scored based on the company’s key issue specific performance.
Scores are aggregated using materiality and polarity to derive the ESG score, following which the scores are
reviewed by the analysts to assign the ESG rating on a AAA-C scale. This document details the approach to
ESG Risk AI’s ESG assessment.

1.4. What provides predictive power to ESG assessment models

Classical ESG analysis looks at relevant ESG factors and past performance on these factors. However, analysis
of past performance has limited predictive power as the performance may not be consistent. Hence modern
analysis also covers risks in the present as well as foreseeable future and evaluates not only past performance
but also the ESG risk identification and mitigation strategies, processes and the overall ESG risk management
framework of the company.

Thus, combining past performance with the company’s ESG risk management strategy and process allows
assessments to predict the ability of the company to foresee and manage ESG risks as and when they occur,
thereby giving assessments adequate predictive power.

Companies that report on their risk identification and management of environmental, social and governance
risks, can be assessed with a fair degree of accuracy and the contrary is also true with the predictive power
of assessments declining due to inadequate disclosure of information. Since risks cannot be adequately
assessed for companies providing low disclosures, assessment agencies typically treat the absence of
disclosures in a specific area as absence of a risk management framework to address that specific risk and
conservatively reduce the company’s ratings.
 ESG Risk AI’s ESG assessment methodology 6

It is now well established that environmental, social and governance (ESG) issues are financially material and
contribute substantially to a company’s performance. Investors are relying increasingly on assessment of ESG
factors as important inputs for risk management and business outlook, which in turn influences financial
performance.

2.1. Ratings as a summary of a comprehensive ESG assessment

ESG Risk AI’s ESG Ratings have been developed to help investors understand a company’s ESG performance
and link it to the investor’s portfolio risk. Our ESG ratings are a summary of financially material ESG factors.
The rating report provides the performance on all these factors. Our ratings and the rating report can be used
for portfolio construction and management as well as for ESG performance comparisons and benchmarking.

Figure 1: ESG Risk AI’s framework for ESG assessments

 ESG Risk AI’s ESG assessment methodology 7

2.1.1. Understanding ESG Risk AI’s ESG data taxonomy

A comprehensive ESG assessment requires identifying all material ESG risks and evaluating the company’s
risk management practices to proactively address these risks. Since every company has exposure to a wide
variety of risks and each risk impacts a company to varying degrees, the evaluation of exposure and scoring
of the risk management process has to be structured in a hierarchy where individual data points pertaining
to the risk exposure and management can be aggregated to evaluate the performance. ESG Risk AI aggregates
data in three levels, viz.: The Key Issue, Theme and Category level, each of which is the next level of
aggregation for hierarchical risk evaluation.

Figure 2: Overview of ESG Risk AI's data taxonomy

 ESG Risk AI’s ESG assessment methodology 8

Figure 3: Visual representation of our data taxonomy

As evident from above chart, ESG Risk AI’s ESG ratings are based on three categories, 19 themes with – 9
in Environment, 4 in Social and 6 in Governance.
 ESG Risk AI’s ESG assessment methodology 9

The performance on these 19 themes are assessed by measuring the strategy, performance and results on
35 Key Issues, ~525 indicators and ~739 data points as shown in Table 1 below.

Figure 4: Five levels of ESG Risk AI’s data taxonomy

2.2. Primary data sources for ESG Risk AI’s ESG ratings
ESG Risk AI bases its ESG assessments on company disclosures and publicly available information. Information
sources such as the company’s website, annual reports, CSR/sustainability reports, 10-K/Q (in case the
company has US investors), notice for meetings, vote results as well as local and global NGO and news
websites are being used to assess the company’s performance on ESG issues.

Information Source Frequency of Update

Annual Report Annual
CSR, BRR report Annual
ESGM/AGM notice, Press releases, vote results Annual
Company website Annual
NGO/Government websites Annual
News Daily
Table 1: Data sources for our ESG assessments
 ESG Risk AI’s ESG assessment methodology 10

2.3. ESG Risk AI’s ratings and quality process

All data collected and analyzed for the assessment of a company’s ESG performance are from publicly
available sources. Once the data is collected, there is an in-depth quality assurance process at each stage.

After quality assurance, the data is used by the scoring model to calculate the initial ESG scores. These
initial ESG scores along with the peer comparison are used by the analyst to review the company’s
performance on ESG parameters for assigning ESG ratings. The ratings assigned by the analyst are then
reviewed for quality and process compliance before it is sent to clients.

Figure 5: ESG Risk AI’s ratings and quality process

 ESG Risk AI’s ESG assessment methodology 11

2.3.1. ESG Risk AI’s approach to surveillance and rating updates

Although most of the data for ESG assessments is sourced from annual disclosures, some of the data sources
provide event-based updates. To reflect changes to ratings from data that are dependent on ad-hoc
information, ESG Risk AI will continuously monitor negative news and corporate events.

With every update on negative news or corporate event, we update the ESG scores for the company.
However, ESG Risk AI’s analysts review and update the ratings of a company only in case of a rating change.
For example, consider a company that has a score of 880 has a negative news which causes its score to fall to
824. The drop in the score results in the rating of the company to change from AAA to AA. However, if the
score only would have changed from 880 to 877, then the ESG rating would remain the same. The latter
illustrates that how a change in the ESG score can keep the rating unchanged. In this case, ESG Risk AI will not
update the report until new disclosures are published by the company.

ACTION
Company A Update score and change rating
Company B Only update score
Table 2: Example of score update and rating changes
 ESG Risk AI’s ESG assessment methodology 12

The ESG Risk AI’s ESG assessment framework evaluates the company’s performance across ~740 individual
parameters, that aggregate to key issues and themes which are then combined to evaluate the performance
on E, S and G. The score on each key issue is a combination of the company’s risk management framework to
deal with a specific ESG risk and materiality of the risk to the industry in which the company operates.

Hence each indicator is assigned weights based on the risk materiality and the significance of the indicator in
the company’s risk management framework. The scores of the indicators are totalled and normalized so that
the aggregate score for each industry totals 1,000.

At times despite a robust risk management framework and robust program implementation, there may be
certain events that expose gaps in the ESG management framework. These events usually come to light
through negative news or controversies and are also factored in our assessments.

The aggregated scores are then analyzed to assign a rating.

This section details the scoring methodology.

3.1. Components of ESG Risk AI’s ESG assessment

The ESG Risk AI’s ESG assessment is comprehensive and includes the following steps:

1. Assigning relevancy and materiality to the indicators

2. Accounting for polarity
3. Scoring each indicator
4. Aggregating scores of individual indicators to the overall score using weights
5. Assigning ratings
6. Scoring a company on transparency

Each step is detailed in the sections that follow.

 ESG Risk AI’s ESG assessment methodology 13

3.1.1. Indicator relevancy and materiality

The materiality and relevance of environmental and social indicators vary across industries. Since not all
indicators are relevant to every industry, ESG Risk AI has identified the industry applicability of each indicator
and, as evident in the tables below, 90% of environmental and 93% of social indicators are sector specific.

Indicators that are Indicator that are Total Percentage of industry

Key issues
industry agnostic industry specific Indicator specific indicator

Air pollutant
0 9 9 100%
emissions
Biodiversity impact 0 7 7 100%
Energy efficiency 3 11 14 79%
Environmental
0 13 13 100%
management
ESG reporting
2 1 3 33%
transparency
GHG emissions 3 11 14 79%
Green product/service 0 20 20 100%
Green supply chain 0 6 6 100%
Material efficiency 0 5 5 100%
Waste use 0 9 9 100%
Water efficiency 3 6 9 67%
Water pollution 0 6 6 100%

Environmental 11 104 115 90%

Table 3: Industry agnostic and industry specific indicators in the Environment category
 ESG Risk AI’s ESG assessment methodology 14

Indicator that are Indicator that are Total % of industry

Key Issues
industry agnostic industry specific indicator specific indicator
Community support &
4 2 6 33%
development
Data privacy & security 0 9 9 100%
Employee development 0 6 6 100%
Employee safety 0 16 16 100%
Employment quality 0 8 8 100%
Equal opportunity 0 12 12 100%
Human rights 3 4 7 57%
Product quality 0 9 9 100%
Product responsibility 1 28 29 97%
Product safety 0 11 11 100%
Responsible
0 7 7 100%
procurement

Social 8 112 120 93%

Table 4: Industry agnostic and industry specific indicator in the Social category

Not all indicators are equally material to all industries. Hence for assessments, weights corresponding to the
indicator’s materiality in a specific industry are assigned. They range from very high materiality to marginal
material. For example, GHG emission reduction is considered as very high materiality for mining companies
while GHG emissions are only marginally material for real estate and financial services companies. Diversity
and inclusion strategy on the other hand is not highly material for mining companies but is of very high
materiality for financial services and real estate companies.

The ESG Risk AI materiality and relevancy framework ensures a company’s score is not negatively impacted
if the company does not disclose their risk management framework on issues that are not considered material
to that specific industry. Vice versa, the company’s score is adversely impacted if it does not report on issues
that are material.
 ESG Risk AI’s ESG assessment methodology 15

Figure 6: Example of materiality in financial services industry

3.1.2. Accounting for indicator polarity

Indicators in each category are assigned a polarity to denote if their high performance indicator represents
good or poor risk management. For example in the environmental category, answering “yes” to nuclear
production or “yes” to animal testing in the social category have a negative polarity. However, answering
“yes” to policies or initiatives and targets to reduce nuclear waste and animal testing have a positive polarity.

Quantitative indicators that have a high value reported for energy use or water use in the environmental
category or employee fatalities in the social category have a negative polarity, while a high percentage of
recycled water or average training hours would a positive polarity.

For governance, all indicators under strategy and compliance have a positive polarity, while negative news
are always assigned a negative polarity.

The above principles are used to determine and assign polarity for each ESG indicator assessed by ESG Risk
AI.
 ESG Risk AI’s ESG assessment methodology 16

Figure 7: Understanding polarity of similar indicator

3.1.3. Scoring each indicator

Each indicator is scored based on two parameters, viz. its function in the risk management framework
indicator and the risk it represents.

A company’s risk management framework is evaluated using its strategy/compliance/targets,

programs/initiatives, as well as the results of the same measured against targets set by the companies
themselves.

To evaluate specific aspects of the risk management framework, ESG Risk AI assesses a company’s:

1. Strategy – by evaluating:
a. Policies and governance frameworks that the company has instituted to address crucial ESG risks
and issues.
b. Targets to measure performance (results) against the objectives set by the company.
c. Company’s compliance with specific sections of the Company’s Act 2013 or SEBI guidelines
2. Performance – by evaluating programs or initiatives that the company has put in place to address
key issues.
3. Results – by evaluating quantitative information reported by the company on specific ESG issues
such as energy use, total CO2 emissions, women employees, fatalities, training costs, etc. Results
are usually measured against the targets set by the company.
To ensure comparability, quantitative data points will be normalized consistently across all industries in the
universe. Some common normalization factors used in the model are percentage, ratio to people / revenue
or conversion to standard units.
 ESG Risk AI’s ESG assessment methodology 17

For example, in the environment category, percentage of water recycled will be considered as an indicator in
the scoring model. This data point is already normalized and can be easily compared across companies.

In the social category, total injury rate is a data point which is normalized to million hours worked. In case the
company reports the number of injuries instead of the rate, the latter can be calculated by assuming 1 worker
works 8 hours per day, 2,000 hours per year.

Indicators are then weighted based on their industry materiality and relevancy, e.g.: if not relevant or material
to an industry, the indicator will be assigned a weight of 0, if considered low relevancy, weight assigned is 1
and 4 if very high. More details regarding indicator materiality and relevancy are provided in the sections
Indicator Materiality and Relevancy.

For illustration, a simplified version of our scoring approach is shown in the table below:

Multiplier for Positive Polarity Multiplier for Negative Polarity

Function
Yes No/NA Yes No/NA
Strategy/Target/Compliance 1 0 0 1
Performance 2 0 0 2
Results 3 0 0 3
Table 5: Weights based on polarity

Results are usually quantitative and cannot be scored in isolation to industry benchmarks. Hence ESG Risk AI
consistently uses a comparison based approach for scoring reasults, where the comparison of performance
is with other peers in the specific industry. In such cases the weight is assigned using a percentile approach.

Of the ~525 indicator used in our assessment model, ~83 indicators relate to the results of company
performance on various key issues and are reported as numerical values. Each of these numerical data points
need to be weighted and scored in comparison to peers. While all these data points are first normalized by
either revenue or headcount to make it comparable, the comparison itself is more complex for the following
reasons:

1. The comparison will always be with a sample set of peers as the entire population of peers will not be
listed and all listed companies will not be covered in the initial years.
2. Comparative scoring will require availability of all data simultaneously but given that companies
disclose data at varying points in time, the comparisons cannot be done on actual peer data.

Given the above constraints, using an absolute percentile-based scoring for numerical data points will require
either bunching of disclosures or bunching of publications. With the weekly refresh, adopting the absolute
percentile approach will delay publications, requiring a stochastic approach.

After considering all the possible options, ESG Risk AI, has chosen the Z-Score based determination of
percentile using the area under the normal distribution curve.
 ESG Risk AI’s ESG assessment methodology 18

Results
3.5

2.5

1.5

0.5

0
Best Worst

Positive Negative

Figure 8: Scoring the best and worst performers using percentiles

3.1.4. Scoring negative news

Negative news/controversies are unfailing indicators of the gaps in an issuer’s risk management framework.
The inability to foresee and manage a risk is starkly evident when the company faces controversies arising
from its inability to address adverse events when they are encountered in the normal course of business.

A company’s involvement in controversial events (negative news) that have an impact on the environment
or society are also considered in the assessment of the ESG scoring. Data points under the “Negative News”
function, are pre-assigned a weight ranging from 0-4 based on their impact on the company, with 0 signifying
no impact and 4 signifying very high impact.

Negative news/controversies have varying levels of impact and the issuers themselves have varying
approaches to manage adverse events. ESG Risk AI’s model evaluates the fragility of the risk management
framework based on the magnitude of the controversy’s impact. The approach on how controversies impact
the scores (deduction of scores) is explained below:
 ESG Risk AI’s ESG assessment methodology 19

Business Impact of Controversy

Very High

High

Medium

Low

0 20 40 60 80 100 120

Deduction of Score from Governance KIs of Business Ethics and Board Structure and Functioning
Deduction of Score in Specific Key Issue (KI)

Figure 9: Impact of negative news on scoring

In case the controversy has a very high impact, the category scores will be deducted by 20% in addition to
the KI deduction. The deduction is made on the last published score.

3.1.5. Aggregating scores of individual indicators to the overall score using weights
To calculate the overall ESG score, the total weight of each indicator is calculated as indicator weight= Risk
materiality * Functions weight * indicator value.

Using the above framework, ESG Risk AI aggregates the scores of individual indicator to calculate the key
issue, theme, category and overall scores.
 ESG Risk AI’s ESG assessment methodology 20

3.1.6. Assigning ratings based on the overall scores

Figure 10: Process of assigning ratings

After the total scores are calculated, the analyst assigns the company’s rating, following the below steps:

1. Review and change the industry classification if required. As explained in section 3.11, both relevancy
and materiality are industry specific and assigning the correct industry code is crucial to evaluate the
appropriate risks. ESG Risk AI’s analysts, based on the company’s reviews and exposure to different
industry segment, confirm or change the industry classification.
2. Review and change the materiality of indicator if needed. Ever so often, companies are exposed to
specific risks due to operational reasons (for example over-dependence on hazardous materials for
manufacturing) or business reasons (for example trade with countries that are ranked poorly on
corruption indexes). In such cases materiality of certain risks may need to be increased. Hence, the
analysts review the business construct and change the risk materiality where relevant.
3. Select different peers if needed to make the comparison more meaningful. This is done to make the
analysis more representative of a company’s ESG risk given the business exposure. For example if a
company is processing tobacco as well as manufacturing confectionary, the peer selection may have to
cover multiple industries.
 ESG Risk AI’s ESG assessment methodology 21

Scores What the rating signifies

Rating Scale
High End Low End
ESG-RISK AAA Above 871 An ESG leader who is successfully managing all ESG risks
ESG-RISK AA 870 721 An ESG leader reliably managing all material ESG risks
An ESG leader with a largely positive track record of managing material
ESG-RISK A 720 571 risks
A company with a good track record of risk management, but no
ESG-RISK BBB 570 421 evidence of a robust framework
A company with a mixed track record of risk management and no
ESG-RISK BB 420 271 evidence of a robust framework
A company with poor track record of risk management and absence of
ESG-RISK B 270 121
a risk management framework
ESG-RISK C Below 250 A company that is drastically impacted by ESG risks
Table 6: Mapping scores to rating scale

Finally, based on the above, the analysts assign the ratings and write a summary explaining the category
specific and the overall risks and strengths of the company’s ESG risk assessment.

To assign the ratings, the analysts use the scores to ratings mapping table as a guidance. The analysts have
the flexibility to change the ratings by a notch, based on their analysis and the table only serves as a guidepost.

3.1.7. Scoring a company’s ESG disclosures and transparency

Based on the company’s disclosure of indicators, ESG Risk AI will compute and publish two transparency
scores, one will score the level of overall disclosures and the second will score the BRR disclosures, relevant
largely in the Indian context.

Overall transparency score: The overall transparency score is calculated as:

Number of indicators where performance can be ascertained through disclosures / Total material indicator

Transparency scores are also calculated at different levels as shown below:

Transparency Score Description

Number of indicators where performance can be ascertained through
Key issue level disclosures/ Total material indicator under key issue
Number of indicators where performance can be ascertained through
Theme level disclosures/ Total indicator under theme
Number of indicators where performance can be ascertained through
Category level disclosures/ Total indicator under category
Table 7: Overall transparency score across hierarchy
 ESG Risk AI’s ESG assessment methodology 22

BRR transparency score: BRR transparency score is based on indicators that correspond with BRR disclosures
and is calculated as Number of indicators corresponding to BRR disclosures material to the industry where
company has performed or complied / Total material indicator corresponding to BRR disclosures.

Methodology maintenance and update: In regular intervals, ESG Risk AI reviews the materiality of each
indicator assigned to each industry as well as their weights. The revision is a forward-looking process to
identify emerging issues and reduce or eliminate issues that are receding in prominence. As part of the
review, ESG Risk AI updates its clients about proposed changes and seeks their feedback.

Intermediate reviews will be performed on a discretionary basis.