Assessing the alignment of corporate ESG disclosures with the UN sustainable development goals: a BERT-based text analysis
| Date | 14 August 2024 |
| Pages | 19-40 |
| DOI | https://doi.org/10.1108/DTA-01-2024-0065 |
| Published date | 14 August 2024 |
| Author | Hyogon Kim,Eunmi Lee,Donghee Yoo |
Assessing the alignment
of corporate ESG disclosures with
the UN sustainable development
goals: a BERT-based text analysis
Hyogon Kim
Korea Land and Housing Corporation, Jinju, South Korea
Eunmi Lee
Department of Textile and Apparel Management, University of Missouri,
Columbia, Missouri, USA, and
Donghee Yoo
Department of Management Information Systems (Bus & Econ Res Inst.),
Gyeongsang National University, Jinju, South Korea
Abstract
Purpose –This study aims to provide measurable information that evaluates a company’s ESG performance
based on the conceptual connection between ESG, non-financialelements of a company and the UN Sustainable
Development Goals (SDGs) for resolving global issues.
Design/methodology/approach –A novel data processing method based on the BERT is presented and
applied to analyze the changes and characteristics of SDG-related ESG texts from companies’disclosures over
the past decade. Specifically, ESG-related sentences are extracted from 93,277 Form 10-K filings disclosed
between 2010 and 2022 and the similarity between these extractedsentences and SDGs statements is calculated
through sentence transformers. A classifier is created by fine-tuning FinBERT, a financial domain-specific
pre-trained language model, to classify the sentences into eight ESG classes.
Findings –The quantified results obtained from the classifier reveal several implications. First, it is observed
that the trend of SDG-related ESG sentences shows a slow and steady increase over the past decade. Second,
large-cap companies relatively have a greater amount of SDG-related ESG disclosures than small-cap
companies. Third, significant events such as the COVID-19 pandemic greatly impact the changes in disclosure
content.
Originality/value –This study presents a novel approach to textual analysis using neural network-based
language models such as BERT. The results of this study provide meaningful information and insights for
investors in socially responsible investment and sustainable investment and suggest that corporations need a
long-term plan regarding ESG disclosures.
Keywords ESG, SDGs, BERT, Corporate disclosure, Form 10-K
Paper type Research paper
1. Introduction
As environmental, social and governance (ESG) considerations have gained significant
global trends, the management and implementation of ESG have become a considerable
concern for both corporate entities and investors alike (Amel-Zadeh and Serafeim, 2018;
Kotsantonis et al., 2016). To enhance investment attractiveness, companies are increasingly
disclosing their ESG performance (Saad and Strauss, 2020); however, they are faced with the
obstacle of the absence of standardized ESG metrics (World Economic Forum, 2022). In
response to this challenge, some companies have turned to utilizing the Sustainable
Development Goals (SDGs) adopted by the United Nations (UN) in September 2015 as a guide
for ESG disclosure standards (Blasco et al., 2018). It is recognized that ESG and SDGs are
interconnected, and the pragmatic framework provided by SDGs serves as a possible starting
Data
Technologies and
Applications
19
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/2514-9288.htm
Received 18 January 2024
Revised 28 May 2024
Accepted 11 June 2024
Data Technologies and
Applications
Vol. 59 No. 1, 2025
pp. 19-40
© Emerald Publishing Limited
2514-9288
DOI 10.1108/DTA-01-2024-0065
point for companies to implement ESG (Berenberg, 2018). Consequently, an analysis of the
relationship between these two concepts in business activities can aid in explaining a
company’s ESG performance (Amel-Zadeh et al., 2021). Nonetheless, measuring a company’s
ESG alignment with the SDGs and determining the extent of the company’s contribution to
each SDG are not a straightforward task (Khaled et al., 2021). In light of these challenges, a
primary objective of this study is to quantitatively assess SDG-related ESG disclosures and
furnish information that can be utilized to evaluate corporate ESG performance, thereby
supporting investment decision-making.
Companies’disclosures, corporate social responsibility (CSR) reports and sustainability
reports are publicly available source that allows investors to verify a company’s ESG
activities and performance. In particular, Form 10-K, an annual report submitted by public
companies to the U.S. Securities and Exchange Commission (SEC), is considered a valuable
source of data for business analysis (Li, 2010;Yuthas et al., 2002). It contains a comprehensive
overview of a company’s annual management activities, including both qualitative data, such
as industry conditions, forecasts and investment plans, and quantitative data, such as
financial statements. Some U.S. public companies have been voluntarily disclosing
information about their non-financial management activities through Form 10-K even
before the increased interest in ESG and SDGs. Form 10-K has been used as the primary raw
data in various studies that analyzed non-financial activities of companies, such as climate
change risk (Doran and Quinn, 2008), gender equity (Nadeem, 2022) and accountability
(Enache and Hussainey, 2020).
Form 10-K composes a vast amount of unstructured text, including numeric data and
characters. Therefore, specific data processing procedures and techniques are required for its
analysis (Loughran and McDonald, 2016). Text mining is a process used to extract useful
information from unstructured text, such as features, patterns and relationships of interest
(Ignatow and Mihalcea, 2017). With the advancements in natural language processing (NLP)
and neural network technology, the technique of text mining is evolving. Recently, highly
advanced neural language models (NLM), such as Bidirectional Encoder Representations
from Transformers (BERT) and Generative Pre-trained Transformer (GPT), have been
developed. These models, which combine neural network technology and big data, have
demonstrated outstanding performance in text mining (Devlin et al., 2018;Floridi and
Chiriatti, 2020).
Despite changes in the corporate management, investment and technical environment,
few textual analysis studies simultaneously considered the ESG and SDG-related contexts
described in Form 10-K. This study aims to provide readers (e.g. consumers and investors)
with derived information that can be used in ESG performance evaluation by analyzing Form
10-K to quantify and visually describe the relationship between ESG and SDGs implied in the
text. In addition, this study also aims to present initial evidence for the automation potential
of processes that identify which companies are in alignment with the SDGs. To this end, this
study proposed a novel quantitative method that considered ESG keywords (Baier et al.,
2020) included in the sentence and the context of the sentence using BERT-based NLM.
Existing related studies that employed textual analytics mainly used lexicon and keywords-
based methods for measuring frequency, which have the problem of counting words used in
unrelated sentences even when keywords are well-defined for a particular domain (Loughran
and McDonald, 2015). The method proposed in this study could be a solution to this issue by
extracting and refining only meaningful sentences from Form 10-K, composed of large
amounts of text.
The detailed tasks of this study are as follows. First, the ESG-related sentences extracted
from Form 10-K were compared with 17 SDGs statements to be quantified. As a comparison
method, similarity analysis based on sentence embeddings using Sentence-Transformers
(Reimers and Gurevych, 2019) was performed. Second, the classifier was developed that
DTA
59,1
20
Get this document and AI-powered insights with a free trial of vLex and Vincent AI
Get Started for FreeStart Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting