ssrn-4336082

Michael J.
Brennan Irish Finance

Working Paper Series
This paper can be downloaded without charge from
Financial Economic Network Electronic Paper Collection:
http://ssrn.com/abstract=4336082
Electronic copy available at: https://ssrn.com/abstract=4336082

ESG Disclosure as Advertisement of Corporate Bond Issuances
Authors in alphabetical order, corresponding author marked with Asterix:
Andreas G. F. Hoepner
University College Dublin, Dublin, Republic of Ireland
European Commission Platform on Sustainable Finance
[email protected]
Frank Schiemann
University of Bamberg, Bamberg, Germany
[email protected]
Fabiola I. Schneider
University College Dublin, Dublin, Republic of Ireland
European Commission Platform on Sustainable Finance
[email protected]
Raphael Tietmeyer*
University of Hamburg, Hamburg, Germany
[email protected]
JEL Classifications: F340, M410, Q5, M370
Keywords: ESG reporting, signaling theory, refinancing risk, corporate bonds, voluntary
disclosure
Abstract This paper investigates whether firms strategically increase their

environmental, social, and governance (ESG) disclosure levels beyond what is expected as a
signal to investors to obtain better access to finance. Analyzing 3,122 firm-year observations
in the US corporate bond market from 2009 to 2017, we find this signal across all three ESG
dimensions for firms facing high refinancing risk. Moreover, we find that firms that issue a
bond have generally higher ESG disclosure levels than firms that do not. We further confirm
our hypothesis by empirically proving that financial benefits in terms of lower bond spreads
are realized. We demonstrate that the signal is particularly prevalent within firms
characterized by high earnings forecast dispersion and error. Our findings highlight the
important role of bond markets in incentivizing voluntary disclosure.

1. Introduction
"I used to think that if there was reincarnation, I wanted to come back as the president or the
pope or as a .400 baseball hitter. But now I would like to come back as the bond market. You
can intimidate everybody."
James Carville, political adviser of Bill Clinton1
Voluntary disclosure of non-financial information is common among large firms, even

though it incurs costs (Ott et al., 2017). It serves a broad stakeholder base and, in the context
of sustainability, is needed to evaluate progress against set targets originating from for
example the 2015 Paris Agreement. Moreover, prior research suggests that environmental,
social, and governance (ESG) data is also valued by investors (Amel-Zadeh & Serafeim,
2018). Yet, most existing research on sustainability focuses on equity markets (Gigante &
Manglaviti, 2022). Primary market transactions - where fresh capital is raised in an exchange
between investor and company - are more common in debt than equity markets, making it
worthwhile shining a light specifically on debt.
Given the recognized demand for and benefits of voluntary ESG disclosure,
companies may exploit this by increasing reporting before a public debt issue. Relying on
Signaling Theory (Spence, 1973), this paper investigates whether firms utilize ESG disclosure
strategically to obtain preferred financing conditions when raising bond capital on primary
markets. Distinguishing between the normal, expected levels of ESG disclosure and the
deviation from them, we demonstrate that firms under high refinancing pressure rely on
additional, abnormally high ESG disclosure when issuing a bond. We frame this as a form of
advertisement for the corporate bond issue to ensure success when facing refinancing risk. We
find this signal across all three dimensions of ESG and show that especially firms
characterized by high earnings forecast dispersion and error rely on it. We also show that
bond-issuing firms have higher normal levels of ESG disclosure compared to non-issuing
firms. In an additional analysis of bond spreads, we confirm the positive effect on the cost of
debt that previous literature notes for ESG disclosure. Thus, our paper concludes that firms
successfully utilize increasing ESG disclosure as a strategy to access preferred financing
conditions. Lastly, we add to the ongoing greenwashing debate by showing that increased
ESG disclosure does not correspond to improved ESG performance.
https://web.archive.org/web/20110805042208/http://www.bloomberg.com/apps/news?pid=newsarchive&sid=a6eMpGVUDe
eE&refer=home

Our study advances research on voluntary non-financial disclosure along several lines.
The first concerns the asset class. We demonstrate the powerful role in incentivizing
voluntary disclosure of ESG information that the bond market plays. Generally, research on
access to finance and ESG is tilted towards equities (Gigante & Manglaviti, 2022; Raimo et
al., 2021), while this paper focuses on public debt.
From an empirical point of view, previous research on equity, such as that by Dhaliwal
et al. (2011) and El Ghoul et al. (2011), uses binary variables to represent (the initiation of)
ESG disclosure. A binary variable can only address the mere decision to issue a sustainability
report or to disclose certain indicators, and it cannot capture differences in ESG disclosure
levels or their changes, as Gao et al. (2016) remark. Our paper on the other hand explicitly
quantifies different levels of disclosure. Importantly, we do not only consider one specific
indicator, such as greenhouse gas emissions, but rather ESG reporting as a whole. This
approach is distinct from the popular method of using third-party ESG ratings. Rather than
relying on the (often untransparent) judgment of data providers we directly use the level of
disclosure without applying any judgement of it.
Furthermore, studies typically directly address the (perceived) benefits of ESG

disclosure for firms and investors. Cheng et al. (2014) highlight the importance of capital
markets for evaluating ESG disclosure in the capital allocation process. Choi and Wang
(2009) show that investors value increased disclosure as increased transparency. We aim to
address the strategic use of ESG disclosure by firms and thereby connect a set of analyses on
both the determinants and the consequences of ESG disclosure. This contributes to a better
understanding of firms’ motivations to engage in voluntary ESG reporting to pursue particular
ends.
This paper begins by introducing literature on ESG and voluntary reporting in this
context, and by highlighting the role of debt markets. Next, the theoretical foundation for the
analysis is outlined. Our analysis is based on Signaling Theory (Spence, 1973) which we
apply to capital raising and voluntary ESG reporting. We develop three hypotheses.
Subsequently, our data and methodology are explained. This is followed by the results of our
analysis and a section dedicated to additional analyses and robustness tests. We investigate
how earnings forecast dispersion and error affect the signaling, if ESG disclosure improves
financial performance in the form of bond spreads, analyze the three dimensions of ESG
separately, and lastly explore whether ESG performance changes in line with disclosure. A
discussion of the implications and limitations is given. The final section concludes.
3

2. Literature and Background
2.1. Voluntary ESG disclosure
ESG has recently gained momentum in academia, policymaking, and practice

(Christensen et al., 2019). In the corporate context, this revolves around firms’ management
of the risks and opportunities related to environmental, social, and governance factors. In this
paper, we refer to “ESG disclosure” and “ESG performance”. However, the literature often
uses synonymous terms, such as “corporate social responsibility” (CSR) or “sustainability”
performance and disclosure. We intentionally use the term ESG in this paper because it relates
best to the methodology that we use for our main variable: the Bloomberg ESG Disclosure
Score (RX317). Disclosure is distinct from performance. Disclosure and the level of
disclosure concern what kind of and how much information is provided. Performance, on the
other hand, judges the disclosed information. Eliwa et al. (2021) note that the market often
fails to distinguish between the two and demonstrate that both ESG disclosure and ESG
performance lead to a lower cost of lending in Europe.
Voluntary ESG disclosure requires financial resources for the preparation of the
disclosure but can also lead to proprietary costs after the disclosure of information (Ott et al.,
2017). This raises the question of what incentivizes firms’ disclosure despite these costs.
Leading reasons from existing literature are that firms aim to gain legitimacy (e.g. Arrigo et
al., 2022; Cho & Patten, 2009; Ochi, 2018) or to profit from value-increasing effects (e.g.
Lambert et al., 2007). For example, Liesen et al. (2016) find that corporate emission
disclosure does not represent a burden on corporate financial resources but instead leads to
outperformance in risk-adjusted share price performance.
Various occurrences are found to be determinants of ESG disclosure, mostly

connected to external stakeholders and societal pressure. Firms respond to environmental
incidents such as oil spills (Heflin & Wallace, 2017; Patten, 1992) or nuclear disasters
(Bonetti, 2023) with higher ESG disclosure. Likely the changing attention to sustainability
issues acts as a driver towards more disclosure and affects the consequences of it. This brings
in regional differences, as awareness and sensitivity in European financial markets are more
pronounced. This is highlighted when comparing Hoepner et al. (2016) and Eliwa et al.
(2021). The former find no significant impact of corporate ESG performance on global
interest rates charged by banks for a period predating the 2015 Paris Agreement. Yet for

Europe, the latter conclude that both, ESG disclosure and ESG performance, benefitted
lending for a sample going beyond the Paris Agreement.
For firms from the United Kingdom, Brammer and Pavelin (2006) demonstrate that
the likelihood of voluntary environmental disclosure is positively correlated with size and
dispersed ownership2, but negatively with indebtedness yet note significant differences among
sectors. Firms in industries with higher pollution, or “sin” industries, might experience greater
public scrutiny and therefore show higher ESG reporting levels (Grougiou et al., 2016).
The effect on financial performance likely acts as an incentive for disclosure. Chen
and Xie (2022) demonstrate that for nonfinancial firms ESG disclosure is beneficial for the
bottom line. Their findings are more pronounced for firms characterized by having ESG
investors, longer inception, high media attention, and high agency costs. Existing literature
indicates that both voluntary disclosure in general (He et al, 2018; Karamanou and Nishiotis,
2009), as well as voluntary ESG disclosure in particular, can benefit capital market
performance. Dhaliwal et al. (2011), El Ghoul et al. (2011), and Healy et al. (1999) explore
the link between voluntary ESG disclosure and the cost of equity capital. Gong et al. (2021)
demonstrate that high-quality ESG reporting moderates the effect of corporate violations on
the cost of debt. Moreover, ESG disclosure can help to build insurance-like protection for
companies (Godfrey et al., 2009; Shiu & Yang, 2017).
In fact, ESG disclosure is actively sought by investors. This demand is highlighted by

the stellar growth of the United Nations Principles for Responsible Investment (UN PRI). In
2023, they were signed by over 5,000 supporters with combined assets under management of
over 140 trillion USD3. The third principle specifically commits to seeking ESG disclosure.
Even though ESG reporting has been largely voluntary, it is a widespread practice among
large companies: in 2022, 98 per cent of the S&P 500 firms published sustainability reports,
an increase of 23 percentage points since 2014 (Governance & Accountability Institute,
2024). However, the regulatory environment is changing rapidly. In the European Union, the
Corporate Sustainability Reporting Directive (CSRD) extends the scope of the Non-Financial
Reporting Directive (NFRD) to all large companies and all companies listed on regulated
markets (European Commission, 2021). In the United States (US), the Securities and
Exchange Commission (SEC) recently adopted the mandatory inclusion of specific
disclosures of climate-related risks, including greenhouse gas (GHG) emissions, in firms’
2 Likewise Patten (1992) demonstrates that change in voluntary disclosure is related to firms size and ownership.
3 https://www.unpri.org/about-us/about-the-pri

periodic reports (SEC, 2024). Moreover, the International Financial Reporting Standards
(IFRS) Foundation created the International Sustainability Standards Board (ISSB), which
also released standards on general sustainability-related disclosures and climate-related
disclosures (IFRS, 2023).
2.2. Debt markets
It is worthwhile distinguishing between asset classes in the context of voluntary ESG

disclosure and capital market effects. Several studies investigate ESG reporting’s influence on
equity (e.g. Dhaliwal et al., 2011, 2012, 2014), noting positive stock price reactions
(Dutordoir et al., 2018). Most equity transactions occur on secondary markets as there is no
principal to be repaid or refinanced.
On the other hand, debt markets follow refinancing cycles, in which firms must
regularly access primary markets to raise capital to refinance maturing principals. They are
considered a critical lever for ESG impact (Hoepner & Schneider, 2022). Nevertheless, even
though debt markets are a much larger source of external finance than equity markets
(Henderson et al., 2006),4 relevant literature is still scarce (Gigante & Manglaviti, 2022;
Chiesa et al., 2021; Raimo et al., 2021, Cojoianu et al., 2022). The existing studies paint a
similar picture to that for equities: there are lenders with a preference for ESG disclosure
(Hamrouni et al., 2019). Sangiogi and Schopohl (2021) analyze motivations for investors to
purchase green bonds and note the importance of transparency. Thus, the reduction in agency
conflict attained through ESG disclosure can also be found in debt financing (Li et al., 2022)5.
Research also indicates lower cost of debt related to ESG disclosure. For the S&P 1200,
Raimo et al. (2021) find a negative relationship between ESG disclosure and the cost of debt
financing. Likewise, Eliwa et al. (2021) analyze European firms and find lower cost of debt
measured using accounting variables.
Specifically for bonds, one of the first papers was published by Menz (2010). At the
time, no significant relationship was found. This changed, as for example Chiesa et al. (2021)
show that environmental performance lowers the cost of bond financing during the period
between 2016 and 2018. This extends work by Gao et al. (2016) who found that ESG
disclosure leads to lower yields to maturity in bond issuances for the Dutch market. Allaya et
4 In 2022, the global equity market and the global bond market are valued at around USD 125 trillion (see
https://www.sifma.org/wp-content/uploads/2022/07/CM-Fact-Book-2022-SIFMA.pdf), while the global bank credit market
adds another USD 85 trillion to the overall debt market (see https://stats.bis.org/statx/srs/table/f2.5).
5 See also related literature on ethics: Barigozzi & Tedeschi (2015) and Kim et al. (2014).

al. (2022) show that for French firms, debt maturity increases with the level of voluntary
disclosure. For the Japanese bond market, Okimoto and Takaoka (2024) show that higher
ESG scores are related to lower credit spreads. This confirms work by Aman and Nguyen
(2013) which lists disclosure quality as a key factor for credit ratings of Japanese firms. For
the Chinese corporate bond market, Gong et al. (2018) demonstrate that higher ESG
disclosure quality is linked to lower costs. Moreover, Ferriani (2023) focuses on bond
issuance during the COVID-19 pandemic and confirms an ESG premium using ESG ratings.
Also using ESG scores, Agnese and Giacomini (2023) find lower cost at issuance for bank
bonds, specifically linking this to ESG reporting and transparency practices.
Kleimeier and Viehs (2023) attribute the reduction in the cost of debt to environmental
risk reduction rather than investor preference, in line with Schneider (2011) showing that
corporate environmental performance will be reflected in its bond pricing due to influence on
insolvency risks. With regard to access to capital, Choi and Gam (2022) show how
reputational risk in relation to environmental issues affects banks’ credit provision in regions,
which are sensitive to the climate transition.
This study specifically looks at the event of bond issuances. These are primary market
transactions, occurring regularly and with reasonable transparency. During the refinancing
cycles, the cost of debt is determined repeatedly and is readily observable. While for public
equity transparency is also a given, initial public offerings (IPOs) occur much less frequently
and are generally not used for refinancing. Thus, the signaling environment (Connelly et al.,
2011) is a very different one as the risk implications differ; the signaler is in a different
position. Failing to secure the refinancing of a bond or incurring high refinancing cost may
result in serious financial concerns for the issuing firm. This has been recognized by Hoepner
& Schneider (2022) who assess investor impact mechanisms and postulate the largest investor
power during the time preceding a debt issuance. Optimal timing for refinancing has been
described as challenging (Kraus, 2009). Hale and Santos (2008) list the firm’s reputation as a
key factor, which relates to voluntary ESG disclosure as a signal to ensure success when
facing refinancing pressure. Lang and Lundholm (1992) predict that the level of disclosure
grows in line with sensitivity to external perceptions and we argue that during the year before
issuing a bond a firm facing refinancing risk is particularly sensitive to outside perceptions.

3. Theory and Hypothesis Development
A whole set of theories has evolved to explain voluntary disclosure as the topic has
been researched for decades6. This includes Agency Theory, Legitimacy Theory, and
Signaling Theory.
Agency Theory (Jensen & Meckling, 1976) revolves around issues arising when
delegating tasks from principals to agents, under consideration of conflicting interests
between them. When investing in a company, investors delegate their decision-making
authority to its management (Healy & Palepu, 2001). In the voluntary disclosure context, this
translates to managers as the agents - who, as insiders of the firm, know more about it -
revealing information to external funders as the principals to counter information asymmetry.
Agency costs can motivate management to voluntarily disclose more information (Wong,
1988; Leftwich et al., 1981). If the investor and the firm have similar levels of interest in
sustainability, ESG disclosure can reduce agency conflicts. Research shows that high
disclosure quality ratings from financial analysts are linked to lower cost of issuing debt
(Sengupta, 1998) and that specifically ESG disclosure reduces information asymmetry (Hung
et al., 2013).
Improved capital market performance can be explained as a consequence of the

increased transparency due to the disclosure of additional information. Increased disclosure
saves investors time and resources for the data collection necessary for their investigations.
Disclosure can directly reduce the cost of capital by lowering the degree of covariance
between a firm’s cash flow and that of the market (Lambert et al., 2007), and because it leads
to broader analyst coverage (Schiemann & Tietmeyer, 2022; Gao et al., 2016, Aerts et al.,
2008). Especially in the sustainability context, highlighting good performance can build trust
with investors (Lins et al., 2017) and helps develop social capital (Amiraslani et al., 2022).
Banks factor the cost of complying with future regulation into their credit assessment,
therefore being proactive in disclosure can lead to lower cost of debt (Khlif et al., 2015).
Cormier and Magnan (2013) show that environmental disclosures are useful in enhancing
analyst forecasts. Additionally, they explicitly speak about the tension between information
needs of external funders and sustaining legitimacy with other stakeholders that managers
face.
6 See for example Verrecchia (1983) and Verrecchia (1990), Land and Lundholm (1993) or Dye (1985) and Dye (1986). Note
that these are mainly concerned with voluntary disclosure to financial market participants and are not specific to ESG factors.

In the context of voluntary disclosure, Legitimacy Theory (Suchmann, 1995) explains
firms’ motivation for disclosures, which economically can cost more than they return in terms
of direct effect on financial performance. In the words of Clarkson et al. (2008, p. 303):
“Socio-political theories explain patterns in the data (“legitimization”) that cannot be
explained by economics disclosure theories”. The argument here is that voluntary disclosure
is a response to pressure from society, including for example customers, employees, or
regulators. Environmentally underperforming firms encounter heightened political and social
pressures, putting their legitimacy in jeopardy. They may bolster voluntary disclosures to
reshape stakeholder perceptions about their actual performance (Clarkson et al., 2008). High
quality environmental disclosures can contribute to environmental legitimacy (Aerts &
Cormier, 2009). Indeed, Cho and Patten (2007, p.639) explicitly call environmental
disclosures “tools of legitimacy”. Specifically looking at carbon-intensive firms, Liu et al.
(2023) portray voluntary disclosure as a communicative legitimacy strategy to protect firm
value. The social license to operate ultimately is granted by society and voluntary disclosure
might be essential to obtain it.
For this paper, we focus on a single stakeholder, namely bond investors. We build our
hypothesis around the signaling angle of voluntary corporate ESG disclosure in advance of a
bond issuance.
3.1. Signaling Theory
Signaling Theory (Spence, 1973) builds on the idea that one party (the sender)
conveys some information about themselves through a signal to another party (the receiver).
In the context of this paper, this translates into firms using voluntary ESG disclosure as a
signal to bond investors. Voluntary ESG disclosure meets the signaling criteria of being
observable and costly (Connelly et al., 2011). We take previous approaches one step further
and investigate whether firms use the signal of increased ESG disclosure strategically for their
benefit. In other words, we investigate whether ESG disclosure is an “active” or “passive”
signal (Lys et al., 2015).
Corporate bonds are particularly suited in this context as they are often publicly
traded. Therefore, advertising them in advance of an issuance seems intuitive. Other debt
instruments such as loans are usually the result of private bi- or multilateral agreements where
lenders can seek information as part of the lending process. Here increasing public voluntary
disclosure is not necessary to attract additional demand or awareness. This is supported by

Tan et al. (2020)’s finding that firms with better ESG disclosure rely more on public debt than
bank debt.
Li et al. (2024, p. 3) specifically link ESG and Signaling Theory, and state that
“managers can use the firm's high ESG scores to attract long-term debt at lower rates”. This
echoes the work by Zerbini (2017), who portrays CSR initiatives as a pathway for a firm to
signal the ethical nature of a business to outsiders, which allows it to overcome adverse
selection issues. The author explicitly expresses that “firms can use these initiatives
strategically” to be reflected in market responses (Zerbini, 2017, p. 1), suggesting an
instrumental view of them. Likewise, Li et al. (1997) model disclosure of environmental
liabilities with the objective of strategically maximizing shareholder value, echoing Barth et
al.’s (1997) finding that capital markets are a driver of disclosure of environmental liabilities.
3.2. Hypotheses Development
Firms have a clear incentive to increase their ESG disclosure before raising capital to
improve their access to the market and to achieve lower (re-)financing costs. Indeed, for
equity markets, Dhaliwal et al. (2011, p. 59) demonstrate that initiation of ESG disclosure
makes subsequent equity raising more likely, as firms “exploit the benefit of a lower cost of
capital associated with the initiation of CSR disclosure”. Likewise, Dutordoir et al. (2018)
demonstrate that high ESG scores are linked to less negative stock price reactions following
subsequent equity offerings (SEOs). For debt, Ge & Liu (2015) find that better ESG scores
are associated with better credit ratings and lower yield spreads in new corporate bond issues
for the years between 1992 and 2009.
Thus, it is worthwhile exploring whether firms utilize ESG disclosure strategically in the
context of public debt issuance. This is based on the assumption that the reporting and
financing departments within a company actively exchange strategic ideas, or, in the absence
of an active exchange, the reporting department is simultaneously aware of the positive
effects of ESG disclosures and the refinancing needs of the company. In such cases, firms
anticipate their next (re)financing activities and increase their ESG disclosure to exploit
capital market benefits. Generally, existing research has long demonstrated a positive
association between voluntary disclosure and the tendency to access capital markets (Frankel
et al., 1995).
In this context, ESG disclosure can be framed as a form of advertising, in which investors
represent consumers in the primary bond market: increased transparency of ESG information
10

is used as an active signal to (potential) buyers like advertisement is used in product
marketing to consumers. ESG has long been associated with advertisement through
environmental disclosure for impression management (Arena et al., 2015) or through raising
investor awareness of the firm (Bhattacharya et al., 2017). The marketing literature shows that
it may send a positive signal about the firm’s quality and type (Kitzmueller & Shimshack,
2012). Lys et al. (2015) specifically acknowledge that a firm can use ESG disclosure as an
active signal to distinguish itself from its competitors. The authors suggest that ESG reporting
can be used as a communication tool to indicate positive future financial performance, based
on their finding that the anticipation of future financial performance leads firms to undertake
ESG initiatives. Therefore, the authors argue that ESG disclosure is a channel to convey
private information to outsiders in contrast to being used for investment or charity reasons.
Likewise, Richardson and Welker (2001) state that ESG disclosure can provide investors with
information regarding future cash flows. Thus, ESG reporting as an active signal to investors
may be used to advertise financial strength ahead of a primary market issue.
Previous research (Dhaliwal, 2011; Dhaliwal et al., 2012) commonly relies on a binary
disclosure variable, indicating whether firms have issued a sustainability report. Our study
goes beyond the existing research by taking a non-binary approach to measuring ESG
disclosure rather than a dummy coding initiation or a specific key performance indicator. Our
variables are based on ESG disclosure levels, which can provide further insights into the
normal level of disclosure and deviations from it, as well as the granular changes in them.
Following Lys et al. (2015), we assume that any firm has an expected ESG disclosure
level based on firm-level characteristics and determinants. Certain financial and operational
characteristics, such as sales, income, cash flow, leverage, market-to-book ratio, research and
development, advertising, and litigation, are prerequisites for a certain level of ESG
disclosure. Industry and institutional setting play a role. We hypothesize that raising external
capital can be understood as an additional determinant connected to increases in ESG
disclosure, framing increasing ESG disclosure before bond issuances as a strategic choice of
firms. Thus, our first hypothesis is the following:
H1: In anticipation of issuing a bond, firms show higher levels of expected ESG disclosure.
ESG disclosure may serve as a form of marketing to investors; thus, it is conceivable that
firms exceed the expected levels of disclosure. This additional disclosure on top of the normal
levels, the deviation from the expected, could be framed as an advertisement directed at
11

external funders. A firm may deploy extra resources to obtain more preferable financing
conditions. Hence, we hypothesize that:
H2: In anticipation of issuing a bond, firms deviate from the expected ESG disclosure levels.
Moreover, a firm’s external capital needs are likely to influence ESG disclosure in
anticipation of capital raising. Firms under higher pressure to raise external capital may be
more inclined to increase their ESG disclosure. Firms must regularly refinance bonds through
capital raising on primary markets when their principal becomes due. Most firms do not have
the cash holdings to pay off the maturing debt but instead roll over the bond. From a firm
perspective, refinancing is connected to several risks. Market conditions outside the firm’s
control might lead to refinancing at a considerably higher interest rate (Froot et al., 1993).
Lenders could also underestimate the continuation value of the firm, which would prevent
refinancing from taking place (Diamond, 1991). The consequences of not being able to
refinance on acceptable terms may entail inefficient liquidation (Diamond, 1991), pressure to
sell important firm assets at low prices (Brunnermeier & Yogo, 2009), and additional costs
stemming from underinvestment problems (Almeida et al., 2009). Overall, high refinancing
risk adds pressure to the bond issuance. Consequently, we hypothesize that firms aim to
mitigate the adverse effects of refinancing risk through the active signal of abnormally high
ESG disclosure levels. Such additional ESG disclosures serve as an advertisement for a bond
issuance at times when a successful issuance is vital to a firm’s survival. Therefore, the
motivation for such an additional increase should be exceptionally high when firms
simultaneously plan to issue a bond and are confronted with high refinancing risks. The final
related hypothesis is stated below:
H3: In anticipation of issuing a bond, firms with high refinancing risk deviate from their
expected levels of ESG disclosure.
To summarize, we formulate one hypothesis revolving around the rise of expected levels
of ESG disclosure in expectation of issuing a bond and two Signaling hypotheses related to
exceeding these expected levels.
4. Material and Methods
4.1 Research design
Empirically, we study the relationship between refinancing risk, the bond refinancing
decision, and ESG disclosure. As part of our research design, we develop the distinction
12

between the expected, or normal, level of ESG disclosure and the deviation from this normal
level of disclosure, as introduced by Lys et al. (2015). We transpose their assumption that,
based on a firm’s current ESG disclosure, operating characteristics, and industry, there is an
optimal ESG disclosure level. The deviation from this optimal, normal, level is likely to be a
signal to stakeholders, including (prospective) investors.
Therefore, we split the ESG disclosure score into two components. We call the first
component Normal ESG, which is the disclosure level that can be explained by industry,
performance, and firm characteristics. We label the unrelated component ESG Deviation.
Following Lys et al. (2015) and Yu and Luu (2021), the firm-level determinants of ESG
disclosure are sales, income, cash flow (CF), leverage, market-to-book ratio (MTB), size,
research and development (RnD), advertising, and litigation. We regress them on the ESG
Disclosure Score to estimate the two components of ESG disclosure:
𝐸𝑆𝐺𝑖,𝑡 = 𝛽0 + 𝛽1 𝑆𝑎𝑙𝑒𝑠𝑖,𝑡 + 𝛽2 𝐼𝑛𝑐𝑜𝑚𝑒𝑖,𝑡 + 𝛽3 𝐶𝐹𝑖,𝑡 + 𝛽4 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 + 𝛽5 𝑀𝑇𝐵𝑖,𝑡 +

𝛽6 𝐴𝑑𝑣𝑒𝑟𝑡𝑖𝑠𝑖𝑛𝑔𝑖,𝑡 + 𝛽7 𝐿𝑖𝑡𝑖𝑔𝑎𝑡𝑖𝑜𝑛𝑖,𝑡 + 𝛽8 𝑅𝑛𝐷𝑖,𝑡 + 𝛽9 𝑆𝑖𝑧𝑒𝑖,𝑡 + 𝐼𝑛𝑑𝐹𝐸 + 𝜀𝑖,𝑡
(1)
The fitted value from the model is our Normal ESG variable, while the residual from
the equation is our ESG Deviation variable. We subsequently use these two measures as
dependent variables.
For our main analysis, we use Bond Decision as the first independent variable. This
binary variable captures whether the firm decides to issue at least one bond in the following
year or not. This assumes a one-year lag between the decision to issue a bond (implicit or
explicit) and issuing a bond.
Our second independent variable is refinancing risk. We follow Harford et al. (2014)
and define refinancing risk as the fraction of a firm’s long-term debt due in the next three
years. Third, we include the refinancing risk as an independent variable to interact with the
bond refinancing decision.
We control for several firm-level and bond-level characteristics: Debt level, financial
constraints, the amount issued, and size. We use proxies for firms’ debt levels following
Kayhan and Titman (2004). We include a control variable to account for financial constraints
based on the KZ Index following Cheng et al. (2014), Kaplan and Zingales (1997), and
Lamont et al. (2001), see Table A1 in the appendix. Even though capital-constrained firms
have a higher incentive to increase their ESG disclosure, they have less means to do so. We
13

include the natural logarithm of total assets as we need to control for size in the relationship
(Cheng et al., 2014), as a proxy for a firm’s general information environment (Atiase, 1985)
and various correlated factors (Hope, 2003). We control for the amount issued in the
respective year to account for the fact that firms raising significantly larger amounts
experience a higher incentive to achieve low cost of capital. All the continuous variables are
winzorised at the 95 per cent level. We control for financial performance measured as Return
on Assets (ROA) to account for the fact that high performing firms may have more means for
disclosure. Looking at cleanup costs within the natural resources sector, Li & McConomy
(1999) make the point that financially well performing firms may be in a better position to
absorb potential consequences from disclosing negative information. We apply industry and
year fixed effects and use double clustering of standard errors at the firm-year level.
We use a quasi-experimental design. This design allows us to compare a treatment

group, consisting of firms that decide to raise capital in year t, with the control group,
consisting of firms that do not raise capital in the same year. In other words, the event of bond
issue represents the treatment.
Ultimately, we establish two models. For hypothesis H1, we use Normal ESG as the
dependent variable.
𝑁𝑜𝑟𝑚𝑎𝑙 𝐸𝑆𝐺𝑖,𝑡 = 𝛽0 + 𝛽1 𝐵𝑜𝑛𝑑 𝐷𝑒𝑐𝑖𝑠𝑖𝑜𝑛𝑖,𝑡 + 𝛽2 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 + 𝛽3 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 ∗

𝐵𝑜𝑛𝑑 𝐷𝑒𝑐𝑖𝑠𝑖𝑜𝑛𝑖,𝑡 + 𝛽4 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝐶𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑡𝑠 + 𝛽5 𝐷𝑒𝑏𝑡 𝐿𝑒𝑣𝑒𝑙𝑖,𝑡 + 𝛽6 𝐴𝑚𝑜𝑢𝑛𝑡 𝐼𝑠𝑠𝑢𝑒𝑑𝑖,𝑡 + 𝛽7 𝑆𝑖𝑧𝑒𝑖,𝑡 +
𝛽8 𝑅𝑂𝐴𝑖,𝑡 + 𝐼𝑛𝑑𝐹𝐸 + 𝜀𝑖,𝑡 (2)
To test our Signaling hypotheses (H2 and H3), we use ESG Deviation as our
dependent variable.
𝐸𝑆𝐺 𝐷𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛𝑖,𝑡 = 𝛽0 + 𝛽1 𝐵𝑜𝑛𝑑 𝐷𝑒𝑐𝑖𝑠𝑖𝑜𝑛𝑖,𝑡 + 𝛽2 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 + 𝛽3 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 ∗

𝐵𝑜𝑛𝑑 𝐷𝑒𝑐𝑖𝑠𝑖𝑜𝑛𝑖,𝑡 + 𝛽4 𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝐶𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑡𝑠 + 𝛽5 𝐷𝑒𝑏𝑡 𝐿𝑒𝑣𝑒𝑙 + 𝛽6 𝐴𝑚𝑜𝑢𝑛𝑡 𝐼𝑠𝑠𝑢𝑒𝑑𝑖,𝑡 + 𝛽7 𝑆𝑖𝑧𝑒𝑖,𝑡 +
𝛽8 𝑅𝑂𝐴𝑖,𝑡 + 𝐼𝑛𝑑𝐹𝐸 + 𝜀𝑖,𝑡 (3)
4.2 Data
We obtain ESG disclosure data and data on bonds from Bloomberg. The field “RX317
– ESG Disclosure Score” provides a number between 0 and 100. This score captures the
amount of data a company discloses across all three pillars of ESG without judging the
disclosure quality.
14

Our initial sample consists of all the US firms that have ESG disclosure scores in any
of the years from 2009 to 2017. By focusing on one country, disclosure requirements from
regulators are more aligned than for a global sample. Compared to for example the European
market (Alessi et al., 2023), ESG disclosure regulations are less demanding in the US which
leaves more space for voluntary disclosure. The sample period is chosen to end before the
Anti-ESG movement in the US gained substantial traction. 2018 is the first year when
investors were offered two dedicated Anti-ESG funds7 and marks the year that the CalPERS’
president was ousted by an anti-ESG candidate8.
We exclude all bonds from financial firms (SIC 6000–6999) due to differences in
terms of capital structure and regulation. All the remaining industries are represented in our
sample. We merge this data set with financial and operational data from Compustat. Our
sample contains 3,122 firm-year observations, which amount to approximately 320
observations per year. In total, we study 507 Firms and 1,068 Bonds. Table A1 in the
appendix provides detailed descriptions of our variables.
We draw on data from the Morgan Stanley Capital International Intangible Value
Assessment (MSCI IVA) and the Thomson Reuters Refinitiv database for our additional
analyses.
5. Results
5.1 Descriptive statistics
Table 1 shows the descriptive statistics for the ESG disclosure values, bond decision,
refinancing risk, and control variables. The average Bloomberg ESG disclosure score is
29.74, with a maximum of 78.11. Moreover, the highest refinancing risk in our sample is 97
per cent, while the mean is 10.7 per cent. In our total sample, around a third of firms issued at
least one bond in a year.
7 https://www.morningstar.co.uk/uk/news/236061/anti-esg-might-be-over-before-it-even-got-going.aspx
8
https://www.responsible-investor.com/calpers-president-priya-mathur-ousted/
15

Table 1. Summary statistics after winsorizing the top and bottom 1%.
N Mean Median Std Dev. p5 p25 p75 P99 Max.

ESG Disclosure Variables
Bloomberg Score 3122 29.764 24.380 16.062 11.2 16.12 42.32 68.18 78.110
Normal ESG 3122 29.091 28.745 10.041 13.798 21.635 35.899 53.185 61.718
ESG Deviation 3122 .673 -0.944 12.249 -16.901 -8.198 8.58 33.046 43.431
Normal ENV 3122 18.336 17.956 12.196 -.216 9.111 26.683 47.575 59.125
ENV Deviation 3122 11.428 10.532 12.442 -7.233 2.747 19.33 44.365 57.474
Normal SOC 3122 22.802 21.968 10.34 7.637 15.163 29.612 47.542 57.396
SOC Deviation 3122 6.962 5.322 12.463 -11.051 -1.665 14.911 39.514 49.418
Normal GOV 3122 58.248 57.958 5.448 49.894 54.441 61.822 71.584 93.740
GOV Deviation 3122 -28.484 -32.361 13.499 -44.954 -39.385 -18.188 5.006 14.767
Independent variables
Bond Decision 3122 .337 0.000 .473 0 0 1 1 1
Refinancing Risk 3122 .107 0.060 .15 0 .003 .141 .739 .977
Main Model
Financial Constraints 3122 -3.262 -3.267 .105 -3.406 -3.35 -3.206 -2.972 -2.707
Debt level 3122 .498 0.000 .5 0 0 1 1 1
Amount Issued 3122 4.186 0.000 11.034 0 0 4 55 150
ROA 3122 5.475 5.654 9.03 -4.861 2.744 9.125 23.158 39.648
Estimation Model
Sales 3122 .904 0.724 .659 .232 .426 1.179 3.402 3.85
Income 3122 .07 0.074 .143 -.113 .032 .126 .344 .573
Cash Flow 3122 .11 0.102 .058 .034 .072 .14 .271 .364
Leverage 3122 .491 0.471 .177 .242 .371 .582 1.084 1.262
MTB 3122 2.207 1.907 1.101 1.081 1.473 2.587 6.348 9.312
Size 3122 9.171 9.058 1.187 7.364 8.316 10.004 12.039 12.732
RnD 3122 .028 0.000 .059 0 0 .025 .279 .392
Advertising 3122 .01 0.000 .023 0 0 .006 .115 .146
Litigation 3122 .002 0.000 .007 0 0 0 .035 .163
16

Table 2 provides a breakdown by industry. The table shows the firm-year distribution
of the sample across 1-digit SIC codes. Financial firms (SIC Code 6) have been excluded due
to differences in terms of capital structure and regulation (Ge & Liu, 2015).
Table 2. Breakdown by industry.
1-digit
Description Firm-Years Percent of Total
SIC
1 Minerals and Construction 321 10,21%
2–3 Manufacturing 1478 47,01%
4 Transportation, Communications, and Utilities 529 16,83%
5 Wholesale Trade and Retail Trade 399 12,69%
7-9 Service Industries and Public 417 13,26%
Total 3,122 100%
We calculate the Pearson correlations between the variables. See Table A2 in the
appendix for the exact figures.
5.2 Regression models
The results of our regression models are shown in Table 3. The results for testing
Hypothesis 1 are presented in column (2), with Normal ESG as the dependent variable.
Column (3) contains the results for ESG Deviation which tests for our two Signaling hypotheses.
Column (1) shows the results of using the ESG Disclosure Score as reported by Bloomberg.
17

Table 3. Main regression model: All the continuous variables are winsorised at the 1 per cent and 99 per cent
levels to mitigate the influence of outliers. Estimated coefficients are followed by standard errors in parentheses.
Significance levels at 10 per cent, 5 per cent, and 1 per cent, one-tailed, are indicated by *, **, and ***,
respectively.
Variable ESG Disclosure Score Normal ESG ESG Deviation

(1) (2) (3)
Bond Decision 0.291 0.963*** -0.672
(0.724) (0.319) (0.614)
Refinancing Risk 0.693 1.064** -0.371
(1.544) (0.552) (1.322)
Bond Decision 4.518 -2.082 6.600**
x Refinancing Risk (4.901) (2.195) (3.952)
Financial constraints -14.830*** -3.140* -11.690***
(2.931) (1.758) (2.289)
Debt level 3.440*** 0.382* 3.058***
(0.711) (0.294) (0.606)
Amount issued -0.048* -0.017 -0.0315
(0.026) (0.014) (0.025)
ROA 0.016 0.088** -0.072***
(0.057) (0.039) (0.024)
Size 4.346*** 6.387*** -2.041***
(0.344) (0.201) (0.291)
Constant -52.850*** -42.510*** -10.340
(10.660) (10.780) (9.176)
Observations 3,122 3,122 3,122
R-squared 0.494 0.770 0.238
Industry FE YES YES YES
Year FE YES YES YES
Cluster FIRM-YEAR FIRM-YEAR FIRM-YEAR
Our results show that the decision to issue a bond is positively related to higher
expected ESG levels (Normal ESG, 0.963, p-value < 0.01), confirming our first hypothesis.
Refinancing risk, however, is not significantly related to Normal ESG, which means that a
higher refinancing risk alone does not lead to a significant increase in normal ESG disclosure
levels.
The interaction term Bond Decision × Refinancing Risk is insignificant, meaning that
when firms decide to issue a bond there is no significant association between higher
18

refinancing risk and normal ESG disclosure levels. We follow Fieberg et al. (2021) and
present the results of our interaction plot in Figure 1.
Figure 1. Interaction plot for Normal ESG. This figure presents the marginal effects of refinancing risk on
normal ESG disclosure levels, in interaction with the decision to issue bond. The solid grey line represents the
estimated marginal effect of refinancing risk on normal ESG disclosure levels if firms decide not to issue a bond.
The solid black line represents the estimated marginal effect of refinancing risk on normal ESG disclosure levels
if firms decide to issue a bond. The long-dash and short-dash lines represent confidence intervals (upper and
lower boundary) of the interaction term with p values of 10 and 5 per cent.
The results of our third regression model show no significant relation between the
decision to issue a bond or the refinancing risk with deviation from the normal, expected level
of ESG disclosure. Consequently, we can reject H2.
However, the results support H3. We note a positive association of Bond Decision ×
Refinancing Risk (6.600, p-value < 0.05) on abnormally high ESG disclosure levels.
Specifically, if a firm decides to issue a bond in the following year, a change of one standard
deviation (ca. 4 per cent) in refinancing risk is associated with an over 6.5 per cent increase in
the ESG Deviation variable. Bond Decision × Refinancing Risk, which can be considered a
proxy for the pressure and action to refinance, is positively related to exceeding expected ESG
disclosure levels. This implies that, if a firm issues a bond while facing high refinancing
pressure, it abnormally increases its ESG disclosure in our sample beyond the expected levels.
19

Our control variables are in line with expectations. Financially constrained firms
disclose less, potentially because they do not have the means to do so. Higher debt levels are
linked to increased ESG deviation, supporting the notion that firms who more desperately
need successful refinancing go the extra mile. The change in the algebraic sign of the Size
and ROA variables is noteworthy. The highly significant positive coefficient for normal ESG
disclosure levels turns negative for ESG disclosure deviation. Normal ESG disclosure levels
increase with firm size and profitability, but larger and more profitable firms rely less on
abnormally high ESG reporting as a signal. A conceivable explanation is that these firms do
not need additional disclosure beyond the expected as an advertisement for their bond issue as
they are already more established in the market. A larger size and high profitability may also
decrease the vulnerability to refinancing risk. The interaction plot of our regression can be
found in Figure 2.
Figure 2. Interaction plot for ESG Deviation. This figure presents the marginal effects of refinancing risk on
ESG disclosure deviation, in interaction with the decision to issue a bond. The solid grey line represents the
estimated marginal effect of refinancing risk on ESG disclosure deviation if firms decide not to issue a bond. The
solid black line represents the estimated marginal effect of refinancing risk on ESG disclosure deviation if firms
decide to issue a bond. The long-dash and short-dash lines represent confidence intervals (upper and lower
boundary) of the interaction term with p values of 10 and 5 per cent.
20

Taken together, we can conclude that firms that decide to issue a bond experience a
stronger positive relation between refinancing risk and normal ESG disclosure levels than
firms that do not decide to issue a bond. They also experience a stronger positive relation
between refinancing risk and ESG disclosure deviation than firms that do not issue bonds.
Therefore, we posit that firms use ESG disclosure as an active signal when they need to
refinance and decide to refinance. We can confirm H1 (firms increase their expected
disclosure levels in anticipation of a bond refinancing) and H3 (firms facing high refinancing
pressure deviate from the expected levels and send a signal through abnormally high ESG
disclosure when deciding to issue a bond).
6. Additional Analyses
6.1 Earnings forecast dispersion and error
It seems plausible that (potential) bond investors are particularly interested in

additional information about an issuer when there is more uncertainty regarding a firm. In the
context of this paper, this means that the signal of ESG deviation is more valuable in this case.
We carry out a further analysis where we split our issuer universe into half by earnings
forecast dispersion and error. We define earnings forecast dispersion as the coefficient of
variation of all earnings estimates. It is measured as the standard deviation of all estimates
that make up the consensus as a percentage of the absolute value of the mean value of all
estimates for a company. We measure earnings forecast error following Dhaliwal et al. (2012)
as all forecasts made in the year for target earnings minus the real earnings per share, scaled
by the stock price at the beginning of the year.
Results are shown in Tables 4 and 5. The subset of firms with higher earnings forecast
dispersion (Table 4) and higher earnings forecast error (Table 5) relied more heavily on
exceeding ESG disclosure expectations, as evident in both higher significance but also a much
coefficient. This is in line with the signaling hypothesis.
21

Table 4. The association between refinancing risk, bond decision, and the computed ESG disclosure
deviation with a sample split for forecast dispersion. All the continuous variables are winsorised at the 1 per
cent and 99 per cent levels to mitigate the influence of outliers. Standard errors are clustered by firm and year to
account for heteroscedasticity. Estimated coefficients are followed by standard errors in parentheses.
respectively.
Variable ESG Deviation ESG Deviation

(low forecast dispersion) (high forecast dispersion)
(1) (2)
Bond Decision -0.917 -2.504**
(1.054) (1.334)
Refinancing Risk -0.375 -3.977
(2.179) (3.035)
Bond Decision 9.424* 16.302**
x Refinancing Risk (6.275) (8.641)
Financial constraints -9.957** -8.788**
(4.178) (5.054)
Debt level 3.217*** 4.229***
(1.031) (1.183)
Amount issued 0.008 0.046
(0.041) (0.061)
ROA -0.057 -0.041
(0.057) (0.052)
Size -1.719*** -2.751***
(0.548) (0.566)
Constant -15.178 -2.371
(11.824) (15.463)
Observations 1,126 856
R-squared 0.055 0.074
Industry FE YES YES
Year FE YES YES
Cluster FIRM-YEAR FIRM-YEAR
22

Table 5. The association between refinancing risk, bond decision, and the computed ESG deviation with a
sample split for forecast error. All the continuous variables are winsorised at the 1 per cent and 99 per cent
levels to mitigate the influence of outliers. Standard errors are clustered by firm and year to account for
heteroscedasticity. Estimated coefficients are followed by standard errors in parentheses. Significance levels at
10 per cent, 5 per cent, and 1 per cent, one-tailed, are indicated by *, **, and ***, respectively.
Variable ESG Deviation ESG Deviation

(low forecast error) (high forecast error)
(1) (2)
Bond Decision -1.235 -2.326**
(1.117) (1.237)
Refinancing Risk -2.059 -1.461
(2.519) (2.498)
Bond Decision 9.459* 18.687**
x Refinancing Risk (6.677) (7.987)
Financial constraints -11.445** -13.547***
(4.435) (4.457)
Debt level 2.167** 5.841***
(1.044) (1.153)
Amount issued 0.025 0.038
(0.044) (0.052)
ROA -0.080 -0.016
(0.063) (0.047)
Size -1.804*** -3.397***
(0.521) (0.584)
Constant -20.221 -12.095
(13.309) (12.889)
Observations 1,021 961
R-squared 0.044 0.080
Industry FE YES YES
Year FE YES YES
Cluster FIRM-YEAR FIRM-YEAR
6.2 Financial benefits of ESG disclosure
Another aspect of the Signaling literature is feedback and market perception (Bardos
et al., 2020), describing the receiver’s reaction to the signal. In the context of this paper, the
capital market performance of the bond issuance represents the feedback from investors.
Previous research highlights the financial benefits associated with ESG disclosure,
mainly focusing on the impact of ESG on equities (Gigante & Manglaviti, 2022; Raimo et al.,
2021, Chiesa et al., 2021). The existing research indicates lower cost of debt (Okimoto &
Takaoka (2024), Raimo et al. 2021, Eliwa et al. 2021; Cheng et al., 2014). Specifically for
corporate bank bonds, Agnese and Giacomini (2023) emphasize the importance of ESG
reporting for lower cost at issuance.
23

We add to this literature by demonstrating that both normal ESG disclosure levels and
deviation from them are negatively associated with Bond Spreadt+1 in our sample of bonds
issued by non-financial firms in the US. This implies that higher ESG disclosure levels lower
the cost of capital but also that an additional increase in disclosure further lowers the
borrowing cost. This is in line with Tan et al.’s (2020) finding that firms with higher levels of
ES disclosure obtain lower bond yield spreads in the US between 2005 and 2013. See Table 6
for the full results.
Table 6. Regression models for the association between normal ESG disclosure levels and deviation from
them on bond spreads. All the continuous variables are winsorised at the 1 per cent and 99 per cent levels to
mitigate the influence of outliers. Estimated coefficients are followed by standard errors in parentheses.
respectively.
Variables Bond Spreadt+1
(1)
Normal ESG -1.796**
(0.934)
ESG Deviation -0.485*
(0.319)
Financial Constraints 32.957
(70.539)
Debt Level 15.564
(12.193)
Amount Issued 0.613**
(0.286)
ROA -0.033***
(0.007)
Size -20.140**
(8.081)
Constant 443.876*
(261.089)
Observations 819
R-squared 0.200
Industry FE YES
Year FE YES
Cluster FIRM-YEAR
6.3 ESG performance
Whether increased ESG disclosure is credible in translating into improved ESG

performance is contested (Arena et al., 2015; Callery & Perkins, 2021; Dando & Swift, 2003).
Hence, we test whether increased ESG disclosure correlates with increased ESG performance
in this additional analysis.
24

Specifically in the context of Signaling Theory, Lopez-Santamaria et al. (2021) study
Colombian firms and show that sustainability disclosure is used for camouflage signals. For
the Chinese market, Xing et al. (2021) find a negative effect for soft environmental disclosure
on obtaining corporate loans and portray greenwashing as key factor for deterring firms from
access to finance. Recent developments around greenwashing in the disclosure as
advertisement context are noteworthy.9 When communications around ESG do not match the
underlying actions of the corporation, the signal may be interpreted as false (Connelly et al.,
2011) and penalized by investors (Harjoto et al., 2024). Hence, merely increasing disclosure
while not matching it with increased performance may not be sufficient to reap capital market
benefits. Credibility is a key aspect of Signaling Theory (Hahn et al., 2019) and has been
questioned in the context of ESG reporting (Arena et al., 2015; Callery & Perkins, 2021;
Dando & Swift, 2003).
Naturally, increasing disclosure is easier than increasing performance. Thus, it is

worthwhile confirming whether additional disclosure is merely a strategic advertisement tool
or equally can be interpreted as a credible commitment to sustainable action. We run an
additional analysis, using the Morgan Stanley International Intangible Value Assessment
(MSCI IVA) Performance Score10 as a proxy for ESG performance. However, it must be
noted that the MSCI IVA score constitutes a comparison against sector peers, using a best-in-
class ratings system. It also focuses on financial materiality. Thus, it may not meet the
expectations of all stakeholders in terms of assessing sustainability performance as is the case
with many ESG ratings11.
Matching our data set with the MSCI IVA results in a subsample of 434 firms. We test
the representativeness of this subsample and find that the results of our main analysis remain
robust. We establish a model to predict the expected, normal ESG performance (Normal ESG
Performance) and the deviation from it (ESG Performance Deviation) based on predictions
using the firm-level financial and operating variables and the MSCI IVA Performance Score,
adopting the same approach that we use in our main model for Normal ESG and ESG
Deviation based on Lys et al. (2015). We find no evidence suggesting that ESG performance
changes in relation to the decision to issue a bond, neither expected levels nor deviation from
them (Table 7).
9 https://www.europarl.europa.eu/thinktank/en/document/EPRS_BRI(2023)753958/
10 https://www.msci.com/documents/10199/25a39052-0b0e-4a10-bef8-e78dbc854168
11 https://www.bloomberg.com/graphics/2021-what-is-esg-investing-msci-ratings-focus-on-corporate-bottom-line/
25

Table 7. Regression models for ESG performance. All the continuous variables are winsorised at the 1 per
cent and 99 per cent levels to mitigate the influence of outliers. Estimated coefficients are followed by standard
errors in parentheses. Significance levels at 10 per cent, 5 per cent, and 1 per cent, one-tailed, are indicated by *,
**, and ***, respectively.
Variable ESG Performance Normal ESG ESG Performance
Performance Deviation
(1) (2) (3)
Bond Decision 0.354 0.311 0.043
(0.834) (0.403) (0.660)
Refinancing Risk 0.786 1.702*** -0.916
(1.500) (0.579) (1.260)
1.822 -2.960* 4.782
Bond Decision
Financial -4.417* -1.312 -3.105

Constraints
(2.976) (1.569) (3.024)
Debt Level 1.639** 0.636** 1.002*
(0.774) (0.290) (0.730)
Amount Issued 0.002 -0.017 0.019
(0.030) (0.016) (0.024)
ROA 0.031 0.050* -0.019
(0.036) (0.030) (0.026)
Size -0.565 0.133 -0.697*
(0.447) (0.205) (0.389)
Constant 39.686*** 48.780*** -9.094
(8.748) (4.732) (8.247)
R-squared 0.120 0.399 0.026
Year FE YES YES YES
26

6.4 E, S, and G individually
Previous research shows that the environmental and social pillars yield better access to
capital (Tan et al., 2020; Cheng et al., 2014). Hence, firms are incentivized to use the
respective E and S disclosures as a signal. Tables 8 and 9 show the results for the
differentiated E, S, and G regressions for both normal ESG disclosure levels and deviation
from them. The results are broadly in line with the main analysis, confirming the signal as
well as heightened normal disclosure levels across all three dimensions of ESG. We find
support for hypothesis H1 (Table 8) as Bond Decision is associated with higher Normal Env
(0.908, p-value < 0.05), Normal Soc (1.125, p-value < 0.01), and Normal Gov (0.798, p-value
< 0.01). The results likewise support H3 (Signaling Hypothesis, Table 9): the interaction term
Bond Decision × Refinancing Risk is associated with increased Env Deviation (7.065, p-value
< 0.05) and increased Soc Deviation (6.468, p-value < 0.1) and Gov Deviation (5.763, p-value
< 0.1).
27

Table 8. Regression models for the association between refinancing risk, bond decision, and the normal
environmental, social, and governance disclosure individually. All the continuous variables are winsorised at
the 1 per cent and 99 per cent levels to mitigate the influence of outliers. Estimated coefficients are followed by
standard errors in parentheses. Significance levels at 10 per cent, 5 per cent, and 1 per cent, one-tailed, are
indicated by *, **, and ***, respectively.
Variable Normal Env Normal Soc Normal Gov
(1) (2) (3)
Bond Decision 0.908** 1.125*** 0.798***
(0.399) (0.369) (0.179)
Refinancing Risk 1.267** 1.069** 0.636*
(0.747) (0.589) (0.443)
Bond Decision -2.547 -1.950 -1.245
Financial Constraints -3.923** -3.148** -1.415*
(2.215) (1.751) (0.954)
Debt Level 0.872** -0.092 -0.028
(0.347) (0.350) (0.176)
Amount Issued -0.025* -0.023* -0.000
(0.019) (0.014) (0.007)
ROA 0.111** 0.072** 0.044**
(0.051) (0.031) (0.018)
Size 7.628*** 6.318*** 3.364***
(0.217) (0.269) (0.122)
Constant -65.714*** -46.099*** 22.258***
(6.621) (5.771) (2.909)
R-squared 0.768 0.760 0.705
Year FE YES YES YES
28

Table 9. Regression models for the deviation from normal environmental, social, and governance
disclosure levels individually. All the continuous variables are winsorised at the 1 per cent and 99 per cent
levels to mitigate the influence of outliers. Estimated coefficients are followed by standard errors in parentheses.
respectively.
Variables Env Deviation Soc Deviation Gov Deviation
(1) (2) (3)
Bond Decision -0.616 -0.833* -0.507

(0.593) (0.633) (0.689)
Refinancing Risk -0.574 -0.376 0.057
(1.225) (1.349) (1.587)
Bond Decision 7.065** 6.468* 5.763*
Financial Constraints -10.902*** -11.677*** -13.411***
(2.324) (2.343) (2.480)
Debt Level 2.568*** 3.531*** 3.468***
(0.583) (0.633) (0.674)
Amount Issued -0.023 -0.025 -0.048**
(0.025) (0.026) (0.026)
ROA -0.095*** -0.056* -0.028
(0.017) (0.034) (0.043)
Size -3.282*** -1.972*** 0.982***
(0.302) (0.279) (0.317)
Constant 5.355 -14.259** -82.617***
(6.951) (6.931) (7.985)
R-squared 0.255 0.256 0.318
Firm FE YES YES YES
29

6.5 Robustness tests
We perform a series of robustness tests12. First, we ensure that our results are not
driven by our research design choices and exclude the industry fixed effects and firm-year
clustered standard errors to reduce concerns about an overfitted model. Furthermore, we
exhibit the analysis by clustering the standard errors only by firm and year instead of two-way
clustering. Next, we perform our regression analysis with only the independent variables (see
Naughton et al., 2018) and exclude the control variables. Moreover, we exclude each control
variable separately to investigate their influence in our regression analyses. The results remain
robust.
A limitation of our sample is that firms which issue bonds but do not have a
Bloomberg Disclosure Score are excluded. We cross check how many bonds we miss due to
no Bloomberg Disclosure Score and identify 219 bonds from 114 ultimate parents. However,
we note that these bonds are to a large extent issued between 2009 and 2012 (140 bonds,
64%). To test whether sample selection bias affected our results, we run a subset consisting
only of bonds issued after 2012 and find that results remain robust.
Furthermore, we investigate to what extent the dependent variables are affected by the
timing of the dependent variables. Therefore, we rerun our analysis with forwarded dependent
variables for the first, second and third year. The results remain robust for our main variables
of interest, ESG Deviationt-1 and Normal ESGt-1.
In addition, we want to ensure that our results are driven by the inherent differences
between firms that have high refinancing risk and firms that have low refinancing risk.
Therefore, we follow the steps suggested by Hainmueller (2012) and rerun our analysis with
an entropy-balanced sample. As a pre-processing step, we perform a median split based on the
Refinancing Risk variable for the weighting. Subsequently, we calibrate the unit sets based on
the control variables of our main analysis. The procedure addresses the possibility that firms
that face high refinancing risk satisfy a large set of similar covariate balance conditions
compared to firms that have low refinancing risk. The results show qualitatively similar
results to our base analyses.
In addition, we examine whether firms that are financially constrained also have fewer
resources to invest in their voluntary ESG disclosure at the time when they prepare for a bond
12 Result tables for robustness tests are available upon request.
30

issue. Hence, we rerun our main analysis with an interaction variable calculated as Bond
Decision x Financial Constraints to carve out whether we find a negative effect on our
dependent variables. The results show that for all variables, we find a negative effect, which
indicates that financially constrained firms indeed have less resources for ESG disclosure.
Moreover, we follow Flammer et al. (2021) and control for the firm’s age, using the
number of years containing data in Compustat as a proxy to account for the possibility that
older firms are more capable of managing refinancing risk13. Results remain robust.
It is furthermore conceivable that an external event influences our results. A

particularly important event is the adoption of the Paris Agreement in December 2015, which
is a legally binding international treaty on climate change. Given the momentum for
sustainability that this spurred, it could amplify the effect of an ESG disclosure signal.
Therefore, we rerun our analysis and use the binary dummy variable Paris Agreement as a
control variable and find that results remain robust.
Moreover, ESG disclosure might be influenced by a firm being incorporated in a “blue

state” (i.e., one where the Democrats won the presidential election during the sample year).
Democrats have a stronger preference for pro-CSR policies than Republicans (Rubin, 2008),
which may be relevant to our findings. Therefore, we use the 2016 presidential elections as a
reference point and control for the firm’s state location. The results remain robust.
7. Discussion
7.1. Implications
Our findings demonstrate that firms strategically utilize increased voluntary ESG
disclosure as a form of advertisement for a bond issue on the primary market, adding to the
growing body of research on debt markets. This highlights the powerful role of public debt
markets in incentivizing voluntary ESG reporting.
Primary markets are where fresh capital is being raised, in transactions between
investors and companies. This is different to secondary markets, where transactions occur
between investors, but firms and their cash flows are not directly impacted. Research on
primary equity markets shows that better ESG performance leads to less underpricing of IPOs
during the pandemic (Zhang & Neupane, 2024).
13We also considered to test following Flammer (2021) by rerunning our analysis with Green Bonds from the Bloomberg
database. However, after merging to our sample, the sample size was too low for a meaningful analysis.
31

Yet, the implications of primary market issuances are different for debt than for
equity. Debt usually comes with fixed maturities when the principal has to be repaid. This not
only makes primary market transactions more common, but the risk is different in case of
refinancing compared to raising capital in an IPO or SEO. Not being able to refinance
maturing debt bears the risk of default.
This matches our finding that the active signal of exceeding expectations in terms of
voluntary ESG disclosure can only be found when a firm is facing heightened refinancing
pressure. The refinancing risk drives the respective company to go the extra mile in terms of
additional disclosure beyond what is expected to ensure the issuance of the bond is successful.
Given that debt undergoes cycles of refinancing, our findings may imply a positive
effect on transparency. If firms increase their ESG disclosure in advance of a bond issuance,
investors first appreciate this higher level of transparency but then expect this higher level of
transparency in subsequent years. This can create a virtuous cycle, whereby firms need to
increase their ESG disclosure constantly – not only to distinguish themselves from other, less
transparent firms but also to meet and exceed investors’ demands for ESG information
continuously. Moreover, the competitive use of reporting to stand out from competitors may
lead to a race to the top for reporting.
7.2. Limitations and future research
While this study allows us to assess the cost of debt and thus access to capital in terms
of preferred financing conditions, our data set does not include bond issuances that failed.
While it is uncommon for a firm to pull a bond after board approval when investment banks
are already trying to sell the instrument to investors on the market, anecdotal evidence points
towards an increase in withdrawn and postponed deals.14 In a buyer’s market, in which high
yields are not enough to attract investors, it is conceivable that ESG reporting as a form of
advertisement may be a decisive factor. Therefore, it would be most interesting to investigate
how ESG disclosure influences the likelihood of success for a bond issuance.
8. Conclusion
This paper sets out to analyze voluntary corporate ESG reporting in the context of
primary market public debt issuance. Principally, we are interested in whether ESG reporting
is strategically utilized like advertisement by bond-issuing firms to achieve preferable
14 https://www.ft.com/content/4a145840-7cb2-4f42-b83a-bd1a69e71eef
32

(re)financing conditions. We factor in how urgently a firm needs the capital it is trying to raise
by compiling a refinancing pressure variable.
Our results demonstrate that firms indeed use ESG disclosure to gain better access to
capital. Analyzing 3,122 firm-year observations in the US corporate bond market for
nonfinancial firms from 2009 to 2017, we provide evidence that firms use abnormally high
ESG disclosure as an active signal when deciding to issuing a bond under refinancing
pressure. Moreover, bond-issuing firms have elevated normal levels of ESG disclosure
compared with non-issuing firms.
Having established the relationship between voluntary ESG reporting and corporate
bond financing, we provide further insights by conducting additional analyses. We distinguish
between the different dimensions of ESG and confirm this signal across all three ESG
dimensions for firms facing high refinancing pressure that issue a bond. We demonstrate that
the signal is especially pronounced for firms characterized by high earnings forecast
dispersion and error. Moreover, we show that investors reward increased ESG disclosure
through lower bond spreads and thus confirm financial benefits of voluntary non-financial
reporting. The deviation from the normal levels of ESG disclosure is associated with even
lower bond spreads, meaning additional disclosures further lower the cost of debt. Our
findings imply that there might be a virtuous cycle in debt financing: to meet and exceed
investors’ expectations, ESG reporting must increase every time a new bond is issued.
Furthermore, the competitive nature of advertisement may lead to a race to the top in ESG
reporting. However, we find no evidence that alongside increased ESG disclosure there is an
increase in ESG performance following the decision to issue a bond. This is interesting in the
context of greenwashing, which is currently a hot topic for financial markets.
33

References
Aerts, W., & Cormier, D. (2009). Media legitimacy and corporate environmental
communication. Acc. Org. Soc., 34(1), 1-27. https://doi.org/10.1016/j.aos.2008.02.005
Aerts, W., Cormier, D., & Magnan, M. (2008). Corporate environmental disclosure, financial
markets and the media: An international perspective. Eco. Econ., 64(3), 643-659.
http://dx.doi.org/10.1016/j.ecolecon.2007.04.012
Agnese, P. & Giacomini, E. (2023). Bank's funding costs: Do ESG factors really matter?.
Finance Research Letters, 51, 103473. https://doi.org/10.1016/j.frl.2022.103437
Alessi, L., Cojoianu, T., French, D., & Hoepner, A. (2023). Special issues in support of
evidence-based policy making in sustainable banking & finance. Int. Rev. Financial Anal.,
102818. https://doi.org/10.1016/j.irfa.2023.102818
Allaya, M., Derouiche, I., Muessig, A. (2022). Voluntary disclosure, ownership structure, and
corporate debt maturity: A study of French listed firms. Int. Rev. Financial Anal., 81,
101300. https://doi.org/10.1016/j.irfa.2018.12.008
Almeida, H., Campello, M., Laranjeira, B., & Weisbenner, S. (2009). Corporate debt maturity
and the real effects of the 2007 credit crisis. NBER Working Paper Series Working Paper
14990. https://doi.org/10.3386/w14990
Aman, H., & Nguyen, P. (2013). Does good governance matter to debtholders? Evidence
from the credit ratings of Japanese firms. Res. in Intl. Bus. & Fin, 29 (C), 14-34.
10.1016/j.ribaf.2013.02.002
Amel-Zadeh, A., & Serafeim, G. (2018). Why and how investors use ESG information:
Evidence from a global survey. Fin. Anal. J., 74(3), 87-103.
Amiraslani, H., Lins, K., Servaes, H. & Tamazyo, A. (2022). Trust, social capital, and the
bond market benefits of ESG performance. Rev. Acc. Stud. https://doi.org/10.1007/s11142-
021-09646-0
Arena, C., Bozzolan, S., & Michelon, G. (2015). Environmental reporting: Transparency to
stakeholders or stakeholder manipulation? An analysis of disclosure tone and the role of
the board of directors. Corp. Soc. Responsib. Environ. Manag., 22(6), 346–361.
https://doi.org/10.1002/csr.1350.
Arrigo, E., Di Vaio, A., Hassan, R., & Palladino, R. (2022). Followership behavior and
corporate social responsibility disclosure: Analysis and implications for sustainability
research. J. Clean. Prod., 360. https://doi.org/10.1016/j.jclepro.2022.132151
Atiase, R. K. (1985). Predisclosure information, firm capitalization, and security price
behavior around earnings announcements. J. Acc. Res., 23(1), 21.
Bardos, K. S., Ertugrul, M., & Gao, L. S. (2020). Corporate social responsibility, product
market perception, and firm value. J. Corp. Finance, 62, 101588.
Barigozzi, F., & Tedeschi, P. (2015). Credit markets with ethical banks and motivated
borrowers. Rev. Finance, 19(3), 1281–1313. https://doi.org/10.1093/rof/rfu030
Barth, M., McNichols, M., & Wilson, P. (1997). Factors Influencing Firms’ Disclosures about
Environmental Liabilities. Rev. Acc. Studies, 2(1), 35-64.
https://doi.org/10.1023/A:1018321610509
34

Bhattacharya, C., Du, S., Sen, S., & Yu, K. (2017). The business case for sustainability
reporting: Evidence from stock market reactions. J. Public Policy Mark., 36(2).
https://doi.org/10.1509/jppm.16.112
Bonetti, P., Cho, C. H., & Michelon, G. (2023). Environmental disclosure and the cost of
capital: Evidence from the Fukushima nuclear disaster. Eur. Acc. Rev.
https://doi.org/10.1080/09638180.2023.2203410
Brammer, S. & Pavelin, S. (2006). Voluntary Environmental Disclosures by Large UK
Companies. J. Bus. Finance Acc., 33(7-8), 1168-1188. https://doi.org/10.1111/j.1468-
5957.2006.00598.x
Brunnermeier, M. K., & Yogo, M. (2009). A note on liquidity risk management. Am. Econ.
Rev., 99(2), 578–583. https://doi.org/10.1257/aer.99.2.578
Callery, P. J., & Perkins, J. (2021). Detecting false accounts in intermediated voluntary
disclosure. Acad. Manag. Discov., 7(1). https://doi.org/10.5465/amd.2018.0229
Chen, Z., & Xie, G. (2022). ESG disclosure and financial performance: Moderating role of
ESG investors. Int. Rev. Financial Anal., 83, 102291.
https://doi.org/10.1016/j.irfa.2022.102291
Cheng, B., Ioannou, I., & Serafeim, G. (2014). Corporate social responsibility and access to
finance. Strateg. Manag. J., 35(1), 1–23. https://doi.org/10.1002/smj.2131
Chiesa, M., McEwen, B., & Barua, S. (2021). Does a company’s environmental performance
influence its price of debt capital? Evidence from the bond market. J. Impact ESG Invest.,
1(3), 75–99. https://doi.org/10.3905/jesg.2021.1.015
Cho, C. H., & Patten, D. M. (2007). The role of environmental disclosures as tools of
legitimacy: A research note. Acc. Org. Soc., 32(7–8), 639–647.
Choi, D. & Gam, Y. (2022). Environmental reputation and bank liquidity: Evidence from
climate transition. J. Bus. Finance Acc., Early View. https://doi.org/10.1111/jbfa.12669
Choi, J., & Wang, H. (2009). Stakeholder relations and the persistence of corporate financial
performance. Strateg. Manag. J., 30(8), 895–907. https://doi.org/10.1002/smj.759
Christensen, H. B., Hail, L., & Leuz, C. (2019). Economic analysis of widespread adoption of
CSR and sustainability reporting standards: Structured overview of CSR literature. NBER
Working Paper Series, Working Paper 26169.
https://www.nber.org/system/files/working_papers/w26169/revisions/w26169.rev0.pdf
Clarkson, P., Li, Y., Richardson, G., & Vasvari, F. (2008). Revisiting the relation between
environmental performance and environmental disclosure: An empirical analysis. Acc.
Organ. Soc., 33(4-5), 303-327. https://doi.org/10.1016/j.aos.2007.05.003
Cojoianu, T., Hoepner, A., & Lin, Y. (2022). Private market impact investing firms:
Ownership structure and investment style. Int. Rev. Financial Anal., 84, 102374.
1https://doi.org/10.1016/j.irfa.2022.102374
Connelly, B., Certo, T., Ireland, R., & Reutzel, C. (2011). Signaling theory: A review and
assessment. J. Manag., 37(1), 39–67.
Dando, N., & Swift, T. (2003). Transparency and assurance minding the credibility gap. J.
Bus. Ethics, 44, 195–200. https://doi.org/10.1023/A:1023351816790
35

Dhaliwal, D. S., Li, O. Z., Tsang, A., & Yang, Y. G. (2011). Voluntary nonfinancial
disclosure and the cost of equity capital: The initiation of corporate social responsibility
reporting. Acc. Rev., 86(1), 59–100. https://doi.org/10.2308/accr.00000005
Dhaliwal, D. S., Li, O. Z., Tsang, A., & Yang, Y. G. (2014). Corporate social responsibility
disclosure and the cost of equity capital: The roles of stakeholder orientation and financial
transparency. J. Acc. Public Policy, 33(4), 328–355.
https://doi.org/10.1016/j.jaccpubpol.2014.04.006
Dhaliwal, D. S., Radhakrishnan, S., Tsang, A., & Yang, Y. G. (2012). Nonfinancial disclosure
and analyst forecast accuracy: International evidence on corporate social responsibility
disclosure. Acc. Rev., 87(3), 723-759.
Diamond, D. W. (1991). Debt maturity structure and liquidity risk. Q. J. Econ., 106(3), 709–
737. https://doi.org/10.2307/2937924
Dutordoir, M., Strong, N. C., & Sun, P. (2018). Corporate social responsibility and seasoned
equity offerings. J. Corp. Finance, 50, 158–179.
https://doi.org/10.1016/j.jcorpfin.2018.03.005
Dye, R. (1985). Disclosure of non-proprietary information. J. Acc. Res., 23(1), 123-145.
https://doi.org/10.2307/2490910
El Ghoul, S., Guedhami, O., Kwok, C. C., & Mishra, D. R. (2011). Does corporate social
responsibility affect the cost of capital? J. Bank. Finance, 35(9), 2388–2406.
https://doi.org/10.1016/j.jbankfin.2011.02.007
Eliwa, Y., Aboud, A., & Saleh, A. (2021). ESG practices and the cost of debt: Evidence from
EU countries. Crit. Perspect. Acc., 79. https://doi.org/10.1016/j.cpa.2019.102097
European Commission. (2021). Proposal for a Directive of the European Parliament and of
the Council amending Directive 2013/34/EU, Directive 2004/109/EC, Directive
2006/43/EC and Regulation (EU) No 537/2014, as regards corporate sustainability
reporting. https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX:52021PC0189
Ferriani, F. (2023). Issuing bonds during the Covid-19 pandemic: Was there an ESG
premium?. Int. Rev. Financial Anal., 88, 102653.
Fieberg, C., Lopatta, K., Tammen, T., & Tideman, S. A. (2021). Political affinity and
investors’ response to the acquisition premium in cross‐border M&A transactions – A
moderation analysis. Strateg. Manag. J., 42(13), 2477–2492.
https://doi.org/10.1002/smj.3325
Flammer, C. (2021). Corporate green bonds. J. Financial Econ., 142(2), 499–516.
https://doi.org/10.1016/j.jfineco.2021.01.010
Flammer, C., Toffel, M. W., & Viswanathan, K. (2021). Shareholder activism and firms’
voluntary disclosure of climate change risks. Strateg. Manag. J., 42(10), 1850–1879.
Frankel, R., McNichols, M., & Wilson, P. (1995). Discretionary disclosure and external
financing. Acc. Rev., 70, 135–150. https://www.jstor.org/stable/248392
36

Froot, K. A., Scharfstein, D. S., & Stein, J. C. (1993). Risk management: Coordinating
corporate investment and financing policies. J. Finance, 48(5), 1629–1658.
https://doi.org/10.1111/j.1540-6261.1993.tb05123.x
Gao, F., Dong, Y., Ni, C., & Fu, R. (2016). Determinants and economic consequences of non-
financial disclosure quality. Eur. Acc. Rev., 25(2), 287–317.
https://doi.org/10.1080/09638180.2015.1013049
Ge, W., & Liu, M. (2015). Corporate social responsibility and the cost of corporate bonds. J.
Acc. Public Policy, 34(6), 597–624. https://doi.org/10.1016/j.jaccpubpol.2015.05.008
Gigante, G. & Manglaviti, D. (2022). The ESG effect on the cost of debt financing: A harp
RD analysis. Int. Rev. Financial Anal., 84, 102382.
Godfrey, P., Merrill, C., & Hansen, J. (2009). The relationship between corporate social
responsibility and shareholder value: An empirical test of the risk management hypothesis.
Strateg. Manag. J., 30(4), 425–445. https://doi.org/10.1002/smj.750
Gong, G., Xin, H., Wu, S., Tian, H., & Li, W. (2021). Punishment by securities regulators,
corporate social responsibility and the cost of debt. J. Bus. Ethics, 171, 337–356.
https://doi.org/10.1007/s10551-020-04438-z
Gong, G., Xu, S., Gong, X. (2018). On the Value of Corporate Social Responsibility
Disclosure: An Empirical Investigation of Corporate Bond Issues in China. J. Bus. Ethics,
150, 227-258. 10.1007/s10551-016-3193-8
Governance & Accountability Institute. (2024). Sustainability reporting trends.
https://www.ga-institute.com/research/ga-research-directory/sustainability-reporting-
trends.html
Grougiou, V., Dedoulis, E., & Leventis, S. (2016). Corporate social responsibility reporting
and organizational stigma: The case of “sin” industries. J. Bus. Res., 69(2), 905–914.
Hahn, R., Reimsbach, D., Kotzian, P., Feder, M., & Weißenberger, B. (2019). Legitimation
strategies as valuable signals in nonfinancial reporting? Effects on investor decision-
making. Bus. Soc., 60(4). https://doi.org/10.1177/0007650319872495
Hainmueller, J. (2012). Entropy balancing for causal effects: A multivariate reweighting
method to produce balanced samples in observational studies. Political Anal., 20(1), 25–
46. https://doi.org/10.1093/pan/mpr025
Hale, G., & Santos, J. A. (2008). The decision to first enter the public bond market: The role
of firm reputation, funding choices, and bank relationships. J. Bank. Finance, 32(9), 1928–
1940.
Hamrouni, A., Uyar, A., & Boussaada, R. (2019). Are corporate social responsibility
disclosures relevant for lenders? Empirical evidence from France. Manag. Decis.
Harford, J., Klasa, S., & Maxwell, W. F. (2014). Refinancing risk and cash holdings. J.
Finance, 69(3), 975–1012. https://doi.org/10.1111/jofi.12133
Harjoto, M., Hoepner, A., & Schneider, F. (2024). Scope 3 greenhouse gas disclosure:
Evidence from oil and gas producers. Working paper. https://dx.doi.org/10.2139/ssrn.4100089
37

He, J., Plumlee, M., Wen, H. (2018). Voluntary disclosure, mandatory disclosure and the cost
of capital. J. Bus. Finance Acc., 46(3-4), 307-355. https://doi.org/10.1111/jbfa.12368
Healy, P., Hutton, A., & Palepu, K. (1999). A Stock Performance and Intermediation
Changes Surrounding Sustained Increases in Disclosure. Cont. Acc. Res., 16, 485–520.
https://doi.org/10.1111/j.1911-3846.1999.tb00592.x
Healy, P., & Palepu, K. (2001). Information asymmetry, corporate disclosure, and the capital
markets: A review of the empirical disclosure literature. J. Acc. Econ., 31(1-3), 405-440.
https://doi.org/10.1016/S0165-4101(01)00018-0
Heflin, F., & Wallace, D. (2017). The BP oil spill: Shareholder wealth effects and
environmental disclosures. J. Bus. Finance Acc., 44(3–4), 337–374.
Henderson, B., Jegadeesh, N., & Weisbach, M. (2006). World markets for raising new capital.
J. Financial Econ., 82, 63–101. https://doi.org/10.1016/j.jfineco.2005.08.004
Hoepner, A., Oikonomou, I., Scholtens, B. & Schröder, M. (2016). The Effects of Corporate
and Country Sustainability Characteristics on The Cost of Debt: An International
Investigation. J. Bus. Finance Acc., 43(1-2), 158-190. https://doi.org/10.1111/jbfa.12183
Hoepner, A., & Schneider, F. (2022). Exit vs voice vs denial of (re)entry: Assessing investor
impact mechanisms on corporate climate transition. Working paper.
https://dx.doi.org/10.2139/ssrn.4193465
Hope, O.‑K. (2003). Disclosure practices, enforcement of accounting standards, and analysts’
forecast accuracy: An international study. J. Acc. Res., 41(2), 235–272.
https://doi.org/10.1111/1475-679X.00102
Hung, M., Shi, J., & Wang, Y. (2013). The effect of mandatory CSR disclosure on
information asymmetry: Evidence from a quasi-natural experiment in China. Asian
Finance Association (AsFA) 2013 Conference. https://dx.doi.org/10.2139/ssrn.2206877
IFRS. (2023) General Sustainability-related Disclosures. https://www.ifrs.org/news-and-
events/news/2023/06/issb-issues-ifrs-s1-ifrs-s2/
Jensen, M. & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency costs
and ownership structure. J. Financial Econ., 3(4), 305-360. https://doi.org/10.1016/0304-
405X(76)90026-X
Kaplan, S., & Zingales, L. (1997). Do investment-cash flow sensitivities provide useful
measures of financing constraints? Q. J. Econ., 112(1), 169–215.
Karamanou, I. & Nishiotis, G. (2009). Disclosure and the Cost of Capital: Evidence from the
Market's Reaction to Firm Voluntary Adoption of IAS. J. Bus. Finance Acc., 36(7-8), 793-
821. https://doi.org/10.1111/j.1468-5957.2009.02154.x
Kayhan, A., & Titman, S. (2005). Firms’ histories and their capital structures. J. Financial
Econ., 83(1), 1–32. https://dx.doi.org/10.3386/w10526
Khlif, H., Guidara, A.; Souissi, M. (2015). Corporate social and environmental disclosure and
corporate performance - Evidence from South Africa and Morocco. J. Acc. in Emerging
Economies, 5(1), 51-69. https://doi.org/10.1108/JAEE-06-2012-0024
38

Kim, M., Surroca, J., & Tribó, J. A. (2014). Impact of ethical behavior on syndicated loan
rates. J. Bank. Finance, 38, 122–144. https://doi.org/10.1016/j.jbankfin.2013.10.006
Kitzmueller, M., & Shimshack, J. (2012). Economic perspectives on corporate social
responsibility. J. Econ. Lit., 50(1), 51–84. https://doi.org/10.1257/jel.50.1.51
Kleimeier, S., & Viehs, M. (2023). Carbon disclosure, emission levels, and the cost of debt.
Working Paper. https://doi.org/10.2139/ssrn.2719665
Kraus, A. (1973). The bond refunding decision in an efficient market. J. Financial Quantit.
Anal., 8(5), 793–806.
Lambert, R., Leuz, C., & Verrecchia, R. (2007). Accounting information, disclosure, and the
cost of capital. J. Acc. Res., 45(2), 385–420. https://doi.org/10.1111/j.1475-
679X.2007.00238.x
Lamont, O., Polk, C., & Saaá-Requejo, J. (2001). Financial constraints and stock returns. Rev.
Financial Stud., 13(2), 529–554. https://doi.org/10.1093/rfs/14.2.529
Lang, M. & Lundholm, R. (1992). An Empirical Assessment of Voluntary Disclosure Theory.
Standford Business School Working Paper No 1188.
Lang, M. & Lundholm, R. (1993). Cross-Sectional Determinants of Analyst Ratings of
Corporate Disclosures. J. Acc. Res., 31(2), 246-271. https://doi.org/10.2307/2491273
Leftwich, R., Watts, R., & Zimmerman, J. (1981). Voluntary corporate disclosure: The case
of interim reporting. J. Acc. Res., 19, 50-77. https://doi.org/10.2307/2490984
Li, Y., Chen, R., & Xiang, E. (2022). Corporate social responsibility, green financial system
guidelines, and cost of debt financing: Evidence from pollution‐intensive industries in
China. Corp. Soc. Responsib. Environ. Manag. 29(3), 593-608.
https://doi.org/10.1002/csr.2222
Li, Y., & McConomy, B. (1999). An empirical examination of factors affecting the timing of
environmental accounting standard adoption and the impact on corporate valuation. J.
Acc., Aud. & Fin., 14(3), 279–313.
Li, W., Padmanabhan, P., & Huang, C. (2024). ESG and debt structure: Is the nature of this
relationship nonlinear? Int. Rev. Financial Anal., 91, 103027.
Liesen, A., Figge, F., Hoepner, A. & Patten, D. (2016). Climate Change and Asset Prices: Are
Corporate Carbon Disclosure and Performance Priced Appropriately?. J. Bus. Finance
Acc., 44(1-2), 35-62) https://doi.org/10.1111/jbfa.12217
Lins, K. V., Servaes, H., & Tamayo, A. N. (2017). Social capital, trust, and firm performance:
The value of corporate social responsibility during the financial crisis. J. Finance, 72(4),
1785–1824. https://doi.org/10.1111/jofi.12505
Liu, Y., & Jiraporn, P. (2010). The effect of CEO power on bond ratings and yields. J. emp.
Finance, 17(4), 744-762. https://doi.org/10.1002/csr.2222
Liu, Y., Zhou, X., Yang, J., Hoepner, A., & Kakabadse, N. (2023). Carbon emissions, carbon
disclosure and organizational performance. Int. Rev. Financial Anal., 90, 102846.
39

Lopez-Santamaria, M., Amaya, N., Hinestroza, M., & Cuero, Y. (2021). Sustainability
disclosure practices as seen through the lens of the signaling theory: A study of companies
listed on the Colombian Stock Exchange. J. Clean. Prod., 317, 128416.
https://doi.org/10.1016/j.jclepro.2021.128416
Lys, T., Naughton, J. P., & Wang, C. (2015). Signaling through corporate accountability
reporting. J. Acc. Econ., 60(1), 56–72. https://doi.org/10.1016/j.jacceco.2015.03.001
Menz, K. (2010). Corporate Social Responsibility: Is it Rewarded by the Corporate Bond
Market? A Critical Note. J. Bus. Ethics, 96, 117-134. https://doi.org/10.1007/s10551-010-
0452-y
Naughton, J. P., Wang, C., & Yeung, I. (2019). Investor sentiment for corporate social
performance. Acc. Rev., 94(4), 401-420.
Ochi, N. (2018). Reporting of real option value related to ESG: Including complementary
systems for disclosure incentives. Int. J. Financial Res., 9(4), 19–34.
https://doi.org/10.5430/ijfr.v9n4p19
Okimoto, T., & Takaoka, S. (2024). Sustainability and credit spreads in Japan. Int. Rev.
Financial Anal., 91, 103052. https://doi.org/10.1016/j.irfa.2023.103052
Ott, C., Schiemann, F., & Günther, T. (2017). Disentangling the determinants of the response
and the publication decisions: The case of the carbon disclosure project. J. Acc. Public
Policy, 36(1), 14–33.
Patten, D. M. (1992). Intra-industry environmental disclosures in response to the Alaskan oil
spill: A note on legitimacy theory. Acc. Org. Soc., 17(5), 471–475.
Raimo, N., Caragnano, A., Zito, M., Vitolla, F., & Mariani, M. (2021). Extending the benefits
of ESG disclosure: The effect on the cost of debt financing. Corp. Soc. Responsib.
Environ. Manag., 28(4), 1412–1421. https://doi.org/10.1002/csr.2134
Richardson, A., & Welker, M. (2001). Social disclosure, financial disclosure and the cost of
equity capital. Acc. Org. Soc., 26(7–8), 597–616. https://doi.org/10.1016/S0361-
3682(01)00025-3
Rubin, A. (2008). Political views and corporate decision making: The case of corporate social
responsibility. Financial Rev., 43(3), 337–360. https://doi.org/10.1111/j.1540-
6288.2008.00197.x
Sangiorgi, I., & Schopohl, L. (2021). Why do institutional investors buy green bonds:
Evidence from a survey of European asset managers. Int. Rev. Financial Anal., 75, 101738.
Schiemann, F., & Tietmeyer, R. (2022). ESG Controversies, ESG Disclosure and Analyst
Forecast Accuracy. Int. Rev. Financial Anal., 84, 102373.
Schneider, T. E. (2011). Is environmental performance a determinant of bond pricing?
Evidence from the U.S. pulp and paper and chemical industries. Contemp. Acc. Rev., 28(5),
1537–1561. https://doi.org/10.1111/j.1911-3846.2010.01064.x
SEC. (2024). SEC Adopts Rules to Enhance and Standardize Climate-Related Disclosures for
Investors. https://www.sec.gov/news/press-release/2024-31
40

Sengupta, P. (1998). Corporate disclosure quality and the cost of debt, Acc. Rev., 73(4), 459-
475. https://www.jstor.org/stable/248186
Shiu, Y., & Yang, S. (2017). Does engagement in corporate social responsibility provide
strategic insurance-like effects?. Strateg. Manag. J., 38(2), 455–470.
Spence, M. (1973). Job market signaling. Q. J. Econ., 87(3), 355.
Suchman, M. (1995). Managing Legitimacy: Strategic and Institutional Approaches. Acad.
Manag. Rev., 20(3), 571-610. https://doi.org/10.2307/258788
Tan, W., Tsang, A., Wang, W., & Zhang, W. (2020). Corporate social responsibility (CSR)
disclosure and the choice between bank debt and public debt. Acc. Horiz., 34(1), 151–173.
https://doi.org/10.2308/acch-52631
Verrecchia, R. (1983). Discretionary disclosure. J. Acc. Econ., 5, 179–194.
https://doi.org/10.1016/0165-4101(83)90011-3
Verrecchia, R. (1990). Information quality and discretionary disclosure. J. Acc. Econ., 12(4),
365-380. https://doi.org/10.1016/0165-4101(90)90021-U
Wong, J. (1988). Economic incentives for the voluntary disclosure of current cost financial
statements. J. Acc. Econ., 10(2), 151–167. https://doi.org/10.1016/0165-4101(88)90018-3
Xing, C., Zhang, Y., & Tripe, D. (2021). Green credit policy and corporate access to bank
loans in China: The role of environmental disclosure and green innovation. Int. Rev.
Financial Anal, 77, 101731. https://doi.org/10.1016/j.irfa.2021.101838
Yu, E., & Luu, B. (2021). International variations in ESG disclosure – Do cross-listed
companies care more?. Int. Rev. Financial Anal., 75, 101731.
Zerbini, F. (2017). CSR Initiatives as Market Signals: A Review and Research Agenda. J.
Bus. Ethics., 146, 1–23. https://doi.org/10.1007/s10551-015-2922-8
Zhang, Z., & Neupane, S. (2024). Global IPO underpricing during the Covid-19 pandemic:
The impact of firm fundamentals, financial intermediaries, and global factors. Int. Rev.
Financial Anal., 91, 102954. https://doi.org/10.1016/j.irfa.2023.102954
41

Appendix
Table A1. Variable definitions and sources

Variable Name Definition Source
Dependent variable
Bloomberg ESG ESG The Bloomberg ESG Disclosure Score (RX317) is a proprietary Bloomberg score based on the extent of a Bloomberg
Disclosure company’s environmental, social, and governance (ESG) disclosure. The score ranges from 0 for
Score companies that do not disclose any of the ESG data included in the score to 100 for those that disclose
every data point. Companies that are not covered by Bloomberg for ESG data will have no score and will
show N/A. A consistent list of topics, data fields, and field weights applies across sectors and regions.
While the topics and data fields included in the score have been selected based primarily on industry
agnostic frameworks, certain topics may not apply to all industries. The environmental (E), social (S), and
governance (G) pillars are equally weighted within the overall ESG Disclosure Score, each topic within a
pillar is equally weighted, and topic weights are allocated across fields related to the issue, with
quantitative fields weighted more heavily than binary fields. This score measures the amount of ESG data
that a company reports publicly and does not measure the company’s performance on any data point.
Bloomberg ENV The Bloomberg ENV Disclosure Score. Bloomberg
ENV Disclosure
Score
Bloomberg GOV The Bloomberg GOV Disclosure Score. Bloomberg
GOV Disclosure
Score
Bloomberg SOC SOC The Bloomberg SOC Disclosure Score. Bloomberg
Disclosure
Score
Normal ESG NORMAL_ESG Normal total ESG disclosure, defined as the predicted value from a regression of total ESG disclosure on Constructed
disclosure various economic determinants described (below) and industry and year fixed effects.
ESG disclosure ESG_DEVIATION Deviation from normal total ENV disclosure, defined as the difference between ENV and Constructed
deviation NORMAL_ENV.
Normal ESG NORMAL_ENV Normal total ESG disclosure, defined as the predicted value from a regression of total ESG disclosure on
ESG disclosure ENV_DEVIATION Deviation from normal total ENV disclosure, defined as the difference between ENV and
42

Normal SOC NORMAL_SOC Normal total SOC disclosure, defined as the predicted value from a regression of total SOC disclosure on Constructed
SOC disclosure SOC_DEVIATION Constructed
Deviation from normal total SOC disclosure, defined as the difference between SOC and NORMAL_SOC.
deviation
Normal GOV NORMAL_GOV Normal total GOV disclosure, defined as the predicted value from a regression of total GOV disclosure on Constructed
GOV disclosure GOV_DEVIATION Deviation from normal total GOV disclosure, defined as the difference between GOV and Constructed
deviation NORMAL_GOV.
MSCI IVA ESG ESG PERF The MSCI IVA ESG Performance Score. MSCI IVA
Performance Score
Normal ESG NORMAL_ESG_PERF Normal ESG performance, defined as the predicted value from a regression of total ESG performance on Constructed
performance various economic determinants described (below) and industry and year fixed effects.
ESG performance ESG_PERF_DEVIATION Deviation from normal total ENV disclosure, defined as the difference between ENV and Constructed
Bond spreads BOND_SPREADS The bond spread of the respective bond. Bloomberg
Explanatory independent variables
Bond decision BOND_DECISION Indicator variable showing whether the firm decided to issue a bond (0) or not (1) in the next year. This is
Bloomberg
estimated with the lagged bond issue.
Refinancing risk REF_RISK Long-term debt due in the next three years minus cash divided by total long-term debt. Compustat
Financially FIN_CONSTRAIN Measure of whether a firm is capital constrained based on the KZ Index. The capital constrained variable is
constrained constructed as a linear combination of five accounting ratios: (1) cash flow to total capital; (2) market-to-book Compustat
ratio; (3) debt to total capital; (4) dividends to total capital; and (5) cash holdings to capital.
Level of debt DEBT_LEVEL Indicator variable for the debt level based on the following lagged variables: (1) market-to-book assets; (2)
property, plant, and equipment/book assets; (3) research and development expenses/sales; (4) a dummy variable
indicating whether a firm reports no research and development expenses; (5) selling expenses/sales, (6) the
Compustat
natural logarithm of sales; and (7) Fama–French 48 industry and year dummy variables. We use the fitted values
from this model (2014) and dichotomise the variable (0 = lower debt level than average, 1 = higher debt level
than average).
Amount issued AMT_ISSUED Total amount issued via bonds in the respective year. Bloomberg
43

Return on Assets ROA Earnings before extraordinary items divided by assets Compustat
Size SIZE Natural logarithm of firm size, computed as the natural logarithm of total assets. Compustat
Firm-level economic determinants of ESG disclosure
Sales SALES Net sales divided by total assets, measured at the end of fiscal year t. Compustat
Income INCOME Income before extraordinary items divided by net sales in fiscal year t. Compustat
Cash flow CF Cash flow from operations (calculated using the indirect method) divided by total assets, measured at the end of Compustat
fiscal year t.
Leverage LEVERAGE Sum of long-term debt and debt in current liabilities divided by total assets, measured at the end of fiscal year t. Compustat
Market-to-book MTB Sum of market value of equity, long-term debt, debt in current liabilities, liquidation value of preferred stock Compustat
ratio and deferred taxes, and investment credit divided by total assets, measured at the end of fiscal year t.
Research and RnD Research and development expense scaled by net sales in fiscal year t. Compustat
development
Advertising ADVERTISING Advertising expense scaled by net sales in fiscal year t. Compustat
Litigation LITIGATION Litigation expense scaled by net sales in fiscal year t. Compustat
44

Table A2. Pearson Correlation Table
Bloomberg Normal ESG ESG Bond Decision Refinancing Financial Debt Level Amount ROA Size
Score Deviation Risk Constraints Issued
Bloomberg Score 1
Normal ESG 0.647*** 1
ESG Deviation 0.781*** 0.0291 1
Bond Decision 0.258*** 0.354*** 0.0488** 1
Refinancing Risk -0.0275 -0.0487** 0.00391 -0.0889*** 1
Financial Constraints -0.412*** -0.529*** -0.107*** -0.248*** 0.0665*** 1
Debt Level 0.458*** 0.595*** 0.113*** 0.288*** -0.0486** -0.436*** 1
Amount Issued 0.206*** 0.332*** -0.00286 0.533*** -0.0476** -0.153*** 0.267*** 1
ROA 0.00770 0.0802*** -0.0557** 0.0183 -0.0180 -0.0460* 0.00809 0.00883 1
Size 0.521*** 0.802*** 0.0261 0.382*** -0.0653*** -0.571*** 0.721*** 0.419*** 0.0180 1
Notes: Significance levels at 10 per cent, 5 per cent, and 1 per cent, one-tailed, are indicated by *, **, and ***, respectively.
45

ssrn-4336082

Uploaded by

Copyright:

Available Formats

ssrn-4336082

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ssrn-4336082

Uploaded by

Copyright:

Available Formats

Michael J.

Brennan Irish Finance

Electronic copy available at: https://ssrn.com/abstract=4336082

JEL Classifications: F340, M410, Q5, M370

Abstract This paper investigates whether firms strategically increase their

Electronic copy available at: https://ssrn.com/abstract=4336082

Voluntary disclosure of non-financial information is common among large firms, even

Electronic copy available at: https://ssrn.com/abstract=4336082

Furthermore, studies typically directly address the (perceived) benefits of ESG

Electronic copy available at: https://ssrn.com/abstract=4336082

2.1. Voluntary ESG disclosure

ESG has recently gained momentum in academia, policymaking, and practice

Various occurrences are found to be determinants of ESG disclosure, mostly

Electronic copy available at: https://ssrn.com/abstract=4336082

In fact, ESG disclosure is actively sought by investors. This demand is highlighted by

Electronic copy available at: https://ssrn.com/abstract=4336082

2.2. Debt markets

It is worthwhile distinguishing between asset classes in the context of voluntary ESG

Electronic copy available at: https://ssrn.com/abstract=4336082

Electronic copy available at: https://ssrn.com/abstract=4336082

Improved capital market performance can be explained as a consequence of the

Electronic copy available at: https://ssrn.com/abstract=4336082

3.1. Signaling Theory

Electronic copy available at: https://ssrn.com/abstract=4336082

3.2. Hypotheses Development

Electronic copy available at: https://ssrn.com/abstract=4336082

Electronic copy available at: https://ssrn.com/abstract=4336082

4. Material and Methods

4.1 Research design

Electronic copy available at: https://ssrn.com/abstract=4336082

𝐸𝑆𝐺𝑖,𝑡 = 𝛽0 + 𝛽1 𝑆𝑎𝑙𝑒𝑠𝑖,𝑡 + 𝛽2 𝐼𝑛𝑐𝑜𝑚𝑒𝑖,𝑡 + 𝛽3 𝐶𝐹𝑖,𝑡 + 𝛽4 𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒𝑖,𝑡 + 𝛽5 𝑀𝑇𝐵𝑖,𝑡 +

Electronic copy available at: https://ssrn.com/abstract=4336082

We use a quasi-experimental design. This design allows us to compare a treatment

𝑁𝑜𝑟𝑚𝑎𝑙 𝐸𝑆𝐺𝑖,𝑡 = 𝛽0 + 𝛽1 𝐵𝑜𝑛𝑑 𝐷𝑒𝑐𝑖𝑠𝑖𝑜𝑛𝑖,𝑡 + 𝛽2 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 + 𝛽3 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 ∗

𝐸𝑆𝐺 𝐷𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛𝑖,𝑡 = 𝛽0 + 𝛽1 𝐵𝑜𝑛𝑑 𝐷𝑒𝑐𝑖𝑠𝑖𝑜𝑛𝑖,𝑡 + 𝛽2 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 + 𝛽3 𝑅𝑒𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑛𝑔 𝑅𝑖𝑠𝑘𝑖,𝑡 ∗

Electronic copy available at: https://ssrn.com/abstract=4336082

5.1 Descriptive statistics

Electronic copy available at: https://ssrn.com/abstract=4336082

N Mean Median Std Dev. p5 p25 p75 P99 Max.

Electronic copy available at: https://ssrn.com/abstract=4336082

Table 2. Breakdown by industry.

5.2 Regression models

Electronic copy available at: https://ssrn.com/abstract=4336082

Variable ESG Disclosure Score Normal ESG ESG Deviation

Electronic copy available at: https://ssrn.com/abstract=4336082

Electronic copy available at: https://ssrn.com/abstract=4336082

Electronic copy available at: https://ssrn.com/abstract=4336082

6.1 Earnings forecast dispersion and error

It seems plausible that (potential) bond investors are particularly interested in

Electronic copy available at: https://ssrn.com/abstract=4336082

Variable ESG Deviation ESG Deviation

Electronic copy available at: https://ssrn.com/abstract=4336082

Variable ESG Deviation ESG Deviation

6.2 Financial benefits of ESG disclosure

Electronic copy available at: https://ssrn.com/abstract=4336082

6.3 ESG performance

Whether increased ESG disclosure is credible in translating into improved ESG

Electronic copy available at: https://ssrn.com/abstract=4336082

Naturally, increasing disclosure is easier than increasing performance. Thus, it is

Electronic copy available at: https://ssrn.com/abstract=4336082