Environmental, social and governance reporting in annual reports: A textual analysis

• Author: Florian Kiesel,Marc Berninger,Philipp Baier
• DOI: http://doi.org/10.1111/fmii.12132
• Date: 01 August 2020
• Published date: 01 August 2020

DOI: 10.1111/fmii.12132

ORIGINAL ARTICLE

Environmental, social and governance reporting in

annual reports: A textual analysis

Philipp Baier1Marc Berninger1Florian Kiesel2

1TechnicalUniversity of Darmstadt,

Darmstadt 64289, Germany

2Grenoble Ecole de Management, Grenoble

38000, France

Correspondence

MarcBerninger, Department of Business

Administration,Economics and Law, Technical

Universityof Darmstadt, 64289 Darmstadt,

Germany.

Email:berninger@bwl.tu-darmstadt.de

Declarationsof interest: This research did

notreceive any specific grant from funding

agenciesin the public, commercial, or not-for-

profitsectors

Abstract

Considering environmental, social, and governance (ESG)

factors becomes increasingly important for companies and

investors. However, “ESG” is not clearly defined so far and,

therefore, it is difficult to measure the ESG activity of com-

panies. We analyze the extent and changes in 10-K reports

and proxy statements on ESG, using a textual analysis and

creating an ESG dictionary. The results show an average of

4.0 % ESG words on total words in the reports. The ESG word

list with 482 items can be used to quantitatively examinethe

extent of ESG reporting, which will be helpful especially for

SRI investors. Our classification of 40 subcategories allows

a highly granular analysis of different ESG related aspects.

Moreover, indications for a relation between changes in

reporting and real events, especially negative media pres-

ence, are detected. Regulatory bodies have to be aware of

the use of such words and how they are used.

KEYWORDS

Content analysis, Environment, ESG, Governance, Investorcommu-

nication, Social, Sustainability measurement, Textualanalysis

JEL CLASSIFICATION

C80, G30, K32, Q01, Q58

1INTRODUCTION

It was recently reported that its new ethic guidelines cost the Norwegian sovereign wealth fund €1.3 billion over the

lastdecade (Handelsblatt, 2017). The fund does not invest in companies producing weapons or infringing human rights.

This is anopen access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits

use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or

adaptations are made.

© 2020 The Authors. Financial Markets, Institutions & Instruments published by New York University Salomon Center and Wiley

PeriodicalsLLC

Financial Markets,Inst. & Inst. 2020;29:93–118. wileyonlinelibrary.com/journal/fmii 93

94 BAIER ET AL.

The major fund manager announced to stay with this strategy because it shares the opinion of the Norwegian public.

Corresponding investmentstrategies are summarized under the term socially responsible investment (SRI). Moreover,

thetotal assets under managementin the US using socially responsible screens when creating a stock portfolio grew to

$8.7 trillion in 2016. This is an increase of 33 % since 2014. In total, one of five dollars under professional management

is invested respecting SRI criteria (Forumfor Sustainable and Responsible Investment, 2016).

Typically, when facing a capital allocation decision, criteria for choosing the best investment are simply of a finan-

cial nature. In contrast,SRI is a process that identifies investments with high contribution to environmental, social, and

governance(ESG) values (see e.g., Auer & Schuhmacher, 2016; Renneboog, TerHorst, & Zhang, 2008). It integrates sus-

tainability criteria into investment decisions. Thereby,it is also possible to create social benefit and not only financial

profit (Bollen, 2007). The main reason for the increasing popularity is the growing pressure concerning sustainability

exerted by regulators and society.This effect was reinforced by the recent global financial crisis which lowered con-

fidence in financial markets and caused new regulations and policies (Sahut & Pasquini-Descomps, 2015). Since the

turmoil of the financial crisis, shareholders want to know how their money is invested and whether it corresponds to

the company’s and their own interests (Hopkins, 2016).

Thisraises the question of how to define ESG criteria. So far, there is no standardized definition, but firms show their

commitments to SRI by voluntary signings to responsible investment organizations. These organizations create own

definitions and guidelines. For example,the US SIF is an organization, which aims to advance sustainable and respon-

sible investing across all asset classes. Their aim is to generate a more sustainable and equitable society by assessing

ESG impacts of investment decisions (Forum for Sustainable and Responsible Investment, 2016). These three major

topics are widely used in SRI literature (e.g., Auer & Schuhmacher,2016; Dimson, Karaka ¸s, & Li, 2015; Lokuwaduge &

Heenetigala, 2016).

The increased attention of SRI among investors begs the question of how companies deal with ESG topics. One

could assume that especially companies who neglected sustainability issues in the past adapt their behavior, since

especially public listed companies depend on investors to buy shares of stock. FollowingESG criteria as a firm is often

referredto as corporate social responsibility (CSR). The current problem that results from SRI research is the measure-

ment of CSR activity, since it is based on qualitativeinformation. Most SRI research relies on SRI ratings provided by

third party organizations, such as ratingagencies. However, not all companies are rated which might lead to a selection

bias.

This paper contributes to the literature in different ways. The first aim of this paper is to measure directly CSR

announced by the firm. Financial reporting, or more precisely annual reports, are identified to be the most reliable

disclosure to quantify a firm’s contribution to CSR. Simultaneously,it is observed that methods of textual analysis are

increasingly applied in accounting and finance literature on various research questions (e.g.,Antweiler & Frank, 2004;

Tetlock,2007; Li, 2008). These methods are used to obtain quantitative from narrative data. Studies analyzing annual

reports concerning ESG issues already apply methods of textual analysis (e.g., Giles & Murphy,2016; Lokuwaduge &

Heenetigala, 2016; Loughran, McDonald, & Yun, 2009; Wilmshurst & Frost,2000). Some of them require text read-

ing in their entirety and to manually classify the sentences. The extent of ESG reporting is judged using checklists or

decision rules (e.g., Giles & Murphy,2016; Tilling & Tilt, 2010). Due to manual reading, they only allow to analyze small

sample sizes and include the problem of personal opinions. Other authors develop measures based on the GRI frame-

work. Those studies mostly try to include further disclosures or information besides annual reports (e.g., Lokuwaduge

& Heenetigala, 2016; Verbeeten,Gamerschlag, & Möller, 2016).

The method of using a word list tries to draw conclusions from the frequency of certain words in documents

(Loughran & McDonald, 2016). A different extentof self-contribution in the creation of word lists is determined in lit-

erature. It depends on the availabilityof an appropriate word list. Analyzing annual reports utilizing a word list is done

by many researchers in the field of accounting and finance (e.g., Loughran& McDonald, 2011, Tetlock, 2007). Regard-

ing ESG issues, a few researchers use word count methods. Loughran et al. (2009) search for ethics-related terms

applying only some keywords.Wilmshurst and Frost (2000) count words on environmental issues without providing a

word list. Verbeeten et al. (2016) developa list of 32 keywords based on Global Reporting Initiative (GRI) framework