The effect of ESG performance on
yield spread?
Master thesis University of Groningen
Faculty of Economics and Business (FEB) Supervisor: A. Dalo
Co-assessor: S. Homroy
Author: Joop Stolle Student Number: S2966247
Abstract: This study provides an empirical analysis of whether the cost of debt is
influenced by ESG performance. This thesis analyses the impact of the combined ESG-, environmental-, social-, and governance pillar score on yield spread. The Paris Climate Agreement of 2015 is handled as an exogenous shock. This work contributes to the literature by using a fixed effects regression instead of a cross-sectional regression. This paper also extends the current research on ESG
performance, as the ESG pillars are researched independently. The sample consists of bond- and ESG data from 2010 to 1019. The main findings of this paper are the significant decreasing effect of the ESG combined, environmental-, social-, and governance pillar score on yield spread. Furthermore, this research finds that the Paris Agreement has a significant negative effect on the cost of debt as well. On the contrary, this study does not find a significant relationship between the ESG
combined, environmental-, social-, and governance pillar score and the corporate yield.
Identification of keywords: ESG, Environment, Social, Governance, Cost of Debt,
Yield Spread, Paris Agreement
1. Introduction
According to a report from Standard & Poor’s (2017), an estimate of more than 1,000 credit rating decisions is linked to Environment, Social, and Governance (ESG) factors. The subject ESG is increasing attention and popularity among people active in the corporate world. Investors, executives, and other stakeholders are
incorporating ESG factors in their strategies. Primarily, the influence on firm value (Fatemi, Glaum & Kaiser, 2017; Harrison & Wicks, 2013) and financial performance (Friede, Busch & Bassen, 2015) is being studied by academia. However, new interest arose, and academics started to research the effect on debt financing from ESG as well (Lee & Faff, 2009). In line with the report from Standard & Poor’s,
scholars believe that the Environmental, Social, and Governance practices should be integrated into the credit scores provided by credit rating agencies (Brindelli et al., 2015; Zeidan et al., 2015; Attig et al., 2013).
maintain a considerable level of sustainable investing. The same applies to Japan, which quadrupled its global sustainable investing assets since 2016.
One of the most comprehensive sustainable investment strategies worldwide is ESG integration. According to GSIR (2018): “ESG integration is the systematic and explicit inclusion by investment managers of Environmental, Social, and Governance factors into financial analysis.” From 2016 to 2018, every region witnessed a rise in ESG integration and sustainability investing. ESG integration has risen by 69 percent in two years to a total of $17.5 trillion in assets. ESG integration also dominates in the United States, Canada, and New Zealand as the most popular strategy. The reason for asset managers to incorporate ESG is to improve the firms’ performance and minimize firm risk. A study from Fitch (2020) showed that 86% of the asset owners instructed their asset managers to show their ESG ratings in the funds managed.
The rise of ESG also influences the bond market. An example is the development of green bonds. The purpose of green bonds is to finance projects with clear
environmental goals. A green bond issuer has to provide a detailed explanation of the allocation of the funds and the environmental impact. In 2008, the World Bank issued the first green bond. Since then, the market has proliferated to $265.4 billion in 2019. Most of these bonds were issued in developed markets (81%). Emerging markets covered 13% of the green bond issuance, while the remaining 6%
happened on a supranational level (Climate Bonds Initiative, 2019).
In 2015, the Paris Agreement was signed by all countries part of the United Nations. The goal of the agreement was to create a common cause to combat climate
to play, and the agreement wants to reward environmentally active firms. Edenhofer et al. (2020) researched the impact of the Paris Agreement on companies’
Environmental- and Social strategies. The authors discover a transition process wherein companies become more efficient in their allocation of resources. Hence, some fixed and financial assets undergo significant devaluation if they have negative Social and Environmental value. Palea & Drogo (2020) extend these findings, while the authors research the effect of carbon risk on the cost of debt, before and after the Paris Agreement of 2015. The authors document a significant positive
relationship between the cost of debt and carbon risk. Against these findings, companies can have higher incentives to invest in Environmental- and Social friendlier projects after the Paris Agreement. Therefore, this research wants to incorporate the potential impact of the Paris Agreement. To summarize, this research wants to address the following question:
What is the effect of ESG on yield spread?
Using a sample of 1,091 firms traded worldwide, this research shows that ESG performance has a significant negative effect on yield spread. Therefore, ESG performance decreases the cost of debt. Moreover, the study finds that both the combined ESG score as the three pillars independently (Environmental, Social, and Governance), have a negative effect on yield spread. Plus, the Paris Agreement decreases yield spread as well.
markets. The goal is to provide a more comprehensive and complete view of the relationship between ESG and debt financing.
Additionally, panel regression is performed to estimate the coefficients in this study. Other studies (Cooper & Uzun, 2015; Barth et al., 2018) use cross-sectional or ordinary least squares regressions. Therefore, this study presents an effect that is determined over a more extended period.
Moreover, the research shows the effect of ESG score as well as the pillars independently. This illustrates the pillar score of most substantial influence and proves that all three pillars are significantly connected to yield spread.
Lastly, this paper includes the Paris Agreement of 2015 and investigates whether this agreement can be identified as an exogenous shock that globally boosts ESG performance. The relation between ESG performance and the Paris Agreement is also a new research subject.
The rest of the paper is structured in the following way: Section 2 provides an overview of existing research related to the relationship between ESG strategy and cost of debt. Section 3 entails the characteristics of the dataset and the
methodological approach, followed by elaborating the results in Section 4. Section 5 provides a discussion, after which the study is concluded in Section 6.
2. Literature review
between ESG and the cost of debt. The third subsection explains the relation between the Environmental pillar score and the cost of debt. Then, the relationship between the Social pillar score and the cost of debt is entailed. The fifth subsection illustrates the relation between the Governance pillar score and the cost of debt. Finally, the moderating influence of the Paris Agreement of 2015 is examined.
2.1. Historical background
The emergence of the Kaplan and Norton balanced scorecards (BSC) back in 1996 could be seen as one of the first corporate social responsibility tools. The scorecards were being used to get increased intention to customer relationship - and human resource management. During the last decades, profit maximization was no longer the sole objective of corporate firms. Multinationals were driven by the shareholder theory, which is slowly getting replaced by the stakeholder theory. This new
innovative trend is taking over the financial market. The extent to which companies are socially responsible can be measured by their Corporate Social Responsibility (CSR) profile, and this CSR profile is based on Environmental, Social, and
Governance (ESG) criteria (Renneboog et al., 2008).
The term ESG originates from the year 2005. A landmark study, “Who Cares Wins”, was presented during a conference that brought institutional investors, asset
Much research has been conducted on the effect of ESG performance. This research is mostly focused on financial performance (Friede et al., 2015; Lins & Servaes, 2017) and firm value (Fatemi et al., 2017; Harrison & Wicks, 2013). The findings highlight benefits on firm value and firm performance from Environmental, Social, and Governance policies. Extensions of ESG research concern the
implementation of ESG and its effect investors’ reward (Revelli, 2017; Revelli & Viviani, 2015) and companies (Orlitzky et al., 2003), by looking at the financial performance after these ESG innovations. The introduced research is mostly focused on short-term reactions from the ESG implementations. More frequently, academia is starting to investigate if and how ESG policies can benefit investors and companies in the long run. ESG considerations reduce long-term risk and the cost of capital (Bassen et al., (2006); Eccles et al., (2014); El Ghoul et al., (2011); Cheng et al., (2014)).
With the rise of Corporate Social Responsibility, screens have been added to
measure this phenomenon. Currently, these screens are Environmental, Social, and Governance pillars. Rating agencies started to administer these ESG scores:
Thomson Reuters ESG database since 2003; the KLD (Kinder, Lydenberg, and Domini Research & Analytics) since 1990.
2.2 The effect of ESG on the cost of debt
between the cost of debt and ESG has no clear-cut boundaries yet, which implies that it could use more research.
One of the first scholars investigating the relationship between ESG performance and debt financing is Menz (2010). His findings suggest that firms with a higher ESG rating endured higher corporate bond prices, which results in higher credit ratings. Menz’s (2010) research sample consists of 498 bonds, coming from European companies during 2004-2007. The author discovers that companies face a higher risk premium when including ESG in their strategic processes.
Menz (2010) is one of the few scholars to find a negative relation between ESG performance and the cost of debt. Contrastingly, Ge & Lui (2015) discover a positive relation. In a research consisting of 4260 corporate bond issues, between 1992 and 2009, a higher CSR score correlates with a decrease in yield spreads when a new corporate bond is issued. This correlation eventually evolves into a better credit rating. Another distinct contribution to the literature is delivered by Stellner et al. (2015). The authors investigate the influence of higher Corporate Social
Performance (CSP) on credit risk. The results show that CSP operates as a risk mitigation component, resulting in a higher-quality rating from a rating agency. A more recent article by Barth et al. (2020) investigates the relation between credit spreads of European companies with ESG performance. The authors find that Environmental performance decreases yield spreads by 0.25%. However, the contrary applies to Social performance. Higher social performance leads to 0.22% higher yield spreads due to a waste of resources, ending in higher firm risk. In line with this research, both Cooper & Uzun (2015) and Oikonomou (2014)
scholars review the effect of Corporate Social Responsibility on the price of
corporate bonds. They find a significant negative effect of CSR performance on the cost of debt. This thesis extends their studies by taking ESG performance as a proxy. An overview of the literature is provided in Table A1 in the Appendix. As a result of the previously mentioned research, the first hypothesis is:
H0a: ESG scores do not have an impact on the cost of debt.
2.3 Effect of Environmental pillar scores on the cost of
debt
Most of the studies focus on the ESG scores combined. Environmental, Social, and Governance policies are interlinked with each other. However, this study claims that there are distinctions among them. Therefore, this research shows that mixing the three pillars could have confounding effects, and we should examine each ESG pillar individually. Sharfman & Fernando (2008) conduct one of the first studies that extend the perspective on the environmental-economic performance connection. According to the authors, higher economic achievements can arise from sharpened resource utilization. Sharfman & Fernando (2008) find that firms benefit highly from
environmental risk management, resulting in a lowered cost of capital. Moreover, a shift from equity- to debt financing is discovered, since it is found that environmental risk management strategy lowers the cost of debt of firms (Sharfman & Fernando, 2008).
of yield spreads regarding the environmental management performance of a company. They find that environmental concerns lead to larger bond spreads. Therefore, higher Environmental scores will lead to fewer environmental concerns, resulting in lower yield spreads. A different study concentrates on a more specific Environmental topic (Schneider, 2011). He examines the relationship between firms in polluting industries and their cost of debt. The research revealed that better environmental performance lowered corporate bond spreads in the secondary market. However, the effect decreased bond quality, according to the author, giving mixed results.
In line with the previously mentioned studies, Chava (2014) finds that investors and private lenders take the environmental profile into account. Environmental problems create a higher cost of equity and debt for a company. Nevertheless, Chava (2014) further provides evidence that Environmental strengths decrease neither the cost of equity nor the cost of debt. However, lenders lower the interest rate for firms that are involved in environmental-linked products. An overview of the literature is provided in
Table A2 in the Appendix.Following the empirical evidence, this study supposes that
a higher Environmental score will positive affect the corporate bond price in the primary market. This empirical evidence leads to the second hypothesis.
H0b: E-scores does not have an impact on the cost of debt.
2.4 Effect of Social pillar score on the cost of debt
higher Social performance will reduce financial and operating risks. This reduction in overall risk results in a decrease in firm risk (McGuire et al., 1988).
Bradley et al. (2007) study a more specific social concern, referred to as employee management. They examine the relationship between employees’ management with corporate debt pricing. The results show that a better employees’ management strategy will lead to a less opportunistic investment strategy combined with higher protection of the bondholders’ wealth. The bond market reacts to this development by reducing yield spreads, which leads to a decrease in the price of debt. In line with this research, Bauer et al. (2009) investigate the relationship between employee management and credit risk. The authors show that risk is mitigated because the behavior of dissatisfied employees cannot harm the firm. This effect will result in a lower cost of debt. Therefore, firms with better employee management will benefit significantly from lower bond prices and higher bond ratings.
positively affect yield spread. An overview of the literature is presented in Table A3 in the Appendix.
H0c: S-scores do not have an impact on the cost of debt.
2.5 Effect of Governance pillar score on the cost of debt
The last dimension of the ESG framework is the Governance pillar score. There has been much research into the relationship between a company's governance policies and a corporate bond price. Bhorjaj and Sengupta (2003) were one of the first academists to investigate the relationship between governance and yield spreads. They consider governance to be the most essential factor to determine credit ratings. A higher Governance score implicates lower information asymmetry. In theirresearch, they link corporate governance mechanisms to lower bond yields and higher bond ratings. By reducing information asymmetry between the firms and lenders, a lower default risk comes forth. The research was extended by Ashbaugh-Skaife et al. (2006), who argue that the information asymmetry is between the management and their outside stakeholders. They confirm that firms with strong corporate governance do receive higher credit ratings than firms with weaker governance, which improves the cost of debt of these firms.
Extending the research concerning the Governance pillar score and the cost of debt, Chava et al. (2009) investigate the connection between a company’s takeover
structure. It concludes that a weak takeover defense will increase the cost of debt for a specific company.
A different aspect is studied by Schauten & van Dijk (2011). They examine corporate governance mechanisms’ reactions on the cost of debt by looking into the
relationship between the level of disclosure and yield spreads. The results are
illustrating a significant relationship between an improved disclosure and a decrease in yield spread. Because bondholders need less of a premium when the disclosure is improved, the risk for bondholders is mitigated. Following the empirical evidence, this study considers that the Governance pillar score has a significant influence on the cost of debt. An overview of the literature is displayed in Table A4 in the Appendix.
H0d: G-scores do not have an impact on the cost of debt.
2.6 Moderating effect of The Paris Agreement in 2015 on
the relationship between ESG and the cost of debt
On the other hand, several studies show the effect of not implementing the Paris Agreement into a corporate strategy. For example, Delis et al. (2018) provide evidence that the stock markets penalize firms with high fossil fuel reserves. In line with these findings, Capasso et al. (2020) find that the market perceives firms with a high carbon footprint as more likely to default, especially after the Paris Agreement. Hence, companies encounter a higher yield spread if they have a carbon footprint. These empirical studies are examples of the effect of the Paris Agreement. This study implements the Paris Agreement as an exogenous shock, wherein the
empirical effect before and after 2015 is distinguished. The effect of ESG on the cost of debt before and after 2015 is compared.
Due to a necessary change in the ESG strategy for all companies in 2015, this study holds that the Paris Agreement could influence the cost of debt and ESG
performance. Therefore, this study presumes that the Paris Agreement will mostly influence the combined ESG and the Environmental pillar score. Besides, the effect on both the Social- and Governance pillar scores is investigated.
H0E: The impact of ESG on the cost of debt is not moderated by the Paris Agreement (of 2015).
3. Data and Methodology
3.1 Data
3.1.1 ESG performance
Following Cheng et al. (2014), the ESG data is retrieved from the Thomson Reuters database, collected through Datastream. Thomson Reuters provides transparent, accurate, and comparable ESG ratings for a global universe of more than 4,500 firms (Stellner et al., 2015). The sample consists of all the companies in the ASSET4 database, which consists of 7,200 firms. The stock ISIN codes of the ASSET4
sample firms are matched to corresponding ESG performance on Datastream. The ratings from Thomson Reuters are comparable and transparent while utilizing the percentile rank methodology. This leads to a minimum rating of 0 (lowest level of ESG) and a maximum of 100 (highest ESG) (Stellner et al., 2015). In line with Galema et al. (2008), I employ both the individual and aggregate ESG ratings.
3.1.2 Yield spread
The daily bond data is obtained from Thomson Reuters Eikon Fixed Income Indices, which provides all the corporate bond data between January 1, 2010, and December 31, 2019. The Fixed Income Indices does not supply bond data before January 1, 2010. Moreover, this research only includes corporate bonds because the effect of a company’s ESG performance on the cost of debt is investigated. Following
in percentage, between a treasury bond yield and a corporate bond yield in a specific year.
3.1.3 Control variables
According to the works of Ge & Lui (2015), Oikonomou et al. (2014), Shi (2003), Ertugrul & Hehe (2008), and Cooper & Uzun (2015), I have to control for bond characteristics that relate to yield spreads. Hence, this study controls for the bond size at the issue of the bond. Issue size can have positive as well as negative effects on the risk premia, and therefore have both a positive and negative effect on yield spreads (Shi, 2003). As claimed by Ge & Lui (2015), Oikonomou et al. (2014), Shi (2003), and Cai et al. (2011), I have to control for firm-level variables. Therefore, this research uses control variables to increase the validity of the research. Thus, firstly, this study controls for the firm size (Size) and is measured by taking the logarithm of a company’s total assets. Secondly, this study controls for a firm's debt ratio
(leverage), which is calculated by dividing the book value of debt by the book value of assets. Thirdly, this study controls for the Return On Assets (ROA), measured by dividing the operating income before depreciation by the book value of assets.
3.2 Descriptive Statistics
& Uzun (2015), the corporate credit spread (the difference between yield spread and corporate yield) is within 2% and 3%.
Table B1, B2, and B3 are available in the Appendix. In Table B1, the bonds per country throughout the sample are provided. Most bonds are issued in Europe and North America. The United States covers 73% of all bonds in the study. As for company industries, Table B2 shows the differences in bonds per SIC. SIC stands for Standard Industrial Classification, which is a sector indicator. The highest number of bonds are issued in the manufacturing sector, followed by the transportation & utilities industries (Cooper & Uzun, 2015; Oikonomou et al., 2014). Lastly, Table B3 provides an overview of the bonds per year. Between 2010 - 2012, there is a
significantly lower volume of issues compared to the following years, which could be due to the aftermath of the financial crisis during that period. The descriptive
statistics indicate that most bonds were issued in 2015.
This study finds ESG data for 1,091 firms in the period from 2010 - 2019. Because ESG scores are highly dependent on a company’s disclosure, many firms only have scores in the last few years. The ASSET4 database is constructed so that the average ESG scores over the whole rating universe have an average of 50. In the sample, this assumption holds for both the combined ESG score and the
Environment score (51.71 and 49.17, respectively). Overall, the companies’ combined ESG score in the ASSET4 database is slightly higher than the average ESG score of all companies worldwide. Moreover, the Social- and Governance pillar scores are above average (59.49 and 57.85, respectively).
Environmental-, Social- and Governance pillar scores. All of these correlations are significant at the 5% level. Considering the control variables, total assets, as well as Return On Assets (ROA), are negatively correlated with yield spread. On the other hand, leverage is positively correlated with yield spread, implicating that an increase in assets and ROA will decrease yield spread while higher leverage increases yield spread.
By reviewing the correlation matrix, this study test for potential multicollinearity throughout the sample. Due to the linkage of the ESG variables, multicollinearity issues can arise. This study finds a high correlation between the combined ESG-, Environment-, Social- and Governance pillar scores. The correlation between the variables is above this thesis threshold of 0,5. Therefore, this study excludes a combined model with the combined ESG-, Environment-, Social- and Governance pillar present, but only create models with the proxies independently.
Table 1. Descriptive statistics
Panel A: Summary Statistics
Panel B: Correlation Matrix
Panel A reports summary statistics (Mean, Percentile 50, Percentile 25, Percentile 75, Standard Deviation, and Number of observations) of the bond-, ESG, and firm-level data. IssueAmount refers to the issue amount of the bond (x 1000). Yield Spread is the corporate yield minus the risk-free rate. BondCount is the number of bonds per firm in a year. ESG is the combined ESG score; Environment refers to the environmental pillar score; Governance refers to the governance pillar score; Social refers to the social pillar score. Total Assets is the logarithm of the total assets. Return on Assets is the operating income before depreciation by the book value of assets. Leverage is the book value of debt divided by the book
value of assets.
Panel B reports the correlation matrix among the variables. *,**,*** indicate significance based on a t-test at 10%, 5%, and 1% significance level, respectively.
3.3 Empirical framework
Following Cooper & Uzun (2015) and Oikonomou et al. (2014), this study uses the following empirical models to test the relationship between ESG and yield spreads of corporate bond issues.
YieldSpreadijt = β0 + β1 ESGit-1 + β2 Paris Agreement + β3 ESGit-1 * Paris Agreement + β4 Sizeit-1
+ β5 Leverageit-1 + β6 ROAit-1 + β7 IssueSizeijt + ɳi + Øt+ εijt (1)
Yield spread is the difference between corporate yield and U.S. yearly treasury yield of the bond j for firm i at time t (Jiang, 2008; Ortiz-Molina, 2006). ESG is the
combined, environmental, social, or governance score for firm i at time t-1. Paris Agreement is a dummy variable, with the value one after December 31, 2015, and zero before the December 31, 2015. Size is the logarithm of total assets for firm i at time t-1 (Oikonomou et al., 2014; Ortiz-Molina, 2006; Cooper & Uzun, 2015).
Leverage is the book value of debt divided by the book value of assets for firm i at time t-1 (Oikonomou et al., 2014; Ortiz-Molina, 2006; Cooper & Uzun, 2015). Return On Assets is the operating income before depreciation divided by the book value of assets for firm i at time t-1 (Oikonomou et al., 2014; Ortiz-Molina, 2006; Cooper & Uzun, 2015). Issue Amount is the issue amount of the bond j for firm i at time t (Ge &
Lui, 2015; Khurana & Raman, 2003). ɳ represents the firm-fixed effects. Ø
represents the year-fixed effects.
In line with previous studies on the relationship between the cost of debt and ESG performance, this study lags the independent variables (Ge & Lui, 2015; Oikonomou et al., 2014; Cooper & Uzun, 2015). There are multiple reasons for this. Firstly, the primary scope is to investigate the causal relationship between ESG performance and the cost of debt. This study desires ESG performance as the cause, whereas the cost of debt is the subsequent effect. Secondly, the simultaneity bias and reverse causality can be reduced by lagging the ESG- and control variables. These problems can arise due to bidirectional causality between ESG performance and cost of debt. Lastly, the common practice of Thomson Reuters is to assemble the data at the end of each calendar year. Therefore, lagging the ESG performance variables ensures that the ratings are already public knowledge at time t and incorporated in bond prices (Godfrey et al., 2009).
4. Empirical Results
spread, whereas Table 3 the relationship between the Environmental-, Social-, and Governance pillar score on yield spread presents. Table 4 covers the relationship between the combined ESG-, Environmental-, Social-, and Governance pillar score variables, interacted with the Paris Agreement, on yield spread. The estimated coefficients are shown with the corresponding p-values in parenthesis.
4.1 Impact of combined ESG score on Yield Spread
Table 2 provides the relationship between yield spread and the combined ESG score. Following Ge & Lui (2015) and Oikonomou et al. (2014), this study lags theindependent variables to isolate the effect on the dependent variable.1
Model 1 shows a regression without fixed effects. This model’s outcome provides a significant negative relation between the combined ESG score and yield spread (0.025, respectively). In Model 2, a performed panel regression outcome is illustrated with firm fixed effects, however excluding year fixed effects. Consequently, Model 3 contains the addition of the year fixed effects. The combined ESG score has a significant negative effect on yield spread of 0.024 and 0.017 percentage points at the 1% significance level. Because both coefficients are significant, this study implements firm- and year-fixed effects throughout the regressions. Due to these findings, the study can reject the first null hypothesis, which states that ESG-active firms do not have a reduction in the cost of debt.
Therefore, this study advises companies to enhance their ESG strategy. A gain of one combined ESG score point should provide a reduction in yield spread of 0.017
percentage points, ceteris paribus. A lower yield spread results in a lower cost of debt. Consequently, firms' financing possibilities improve, which will result in growth opportunities. The findings are in line with Cooper & Uzun (2015), suggesting that firms with a higher ESG combined score have a lower cost of debt. According to their findings, firms that improve their ESG strategies will also have a lower yield spread and a lower cost of debt. Moreover, the results should incentivize companies to continuously improve their ESG strategy. Ge & Lui (2015) also show that improved Corporate Social Responsibility results in a lower yield spread, and a lower cost of debt.
Table 2. Relationship between yield spread and the combined-ESG pillar score
(1) (2) (3)
Yield Spread Yield Spread Yield Spread
L.ESG -0.025*** -0.024*** -0.017*** (0.00) (0.00) (0.00) L.TA -0.150*** -0.157*** -0.255*** (0.03) (0.03) (0.03) L.ROA -0.041*** -0.041*** -0.046*** (0.01) (0.01) (0.01) L.LEV 0.021*** 0.020*** 0.018*** (0.00) (0.00) (0.00) IssueAmountAnnual -0.115** -0.116** -0.036 (0.05) (0.05) (0.05) _cons 9.822*** 9.956*** 7.356*** (1.04) (1.06) (1.04)
Firm_Fixed_Effect No Yes Yes
Year_Fixed_Effect No No Yes N 1092 1092 1092 r2 0.201 r2_b 0.283 0.279 chi2 54.51*** 248.55*** 254.11*** p_value 0.000 0.000 0.000
Robust standard errors in parentheses
="* p<.10 ** p<.05 *** p<.01"
This table depicts the regression results of the relationship between the combined ESG score and yield spread. Model 1 performs a standard regression. Model 2 performs a regression with firm-fixed effect. Model 3 performs a regression with both firm- and year-fixed effects. LagESG is
the 1-year lagged combined ESG score. LagTotalAssets is the 1-year lagged logarithm of total assets. LagROA is the 1-year lagged operating income before depreciation by the book value of assets. LagLeverage is the 1-year lagged book value of debt divided by the book value of assets.
4.2 Impact of Environmental-, Social-, and Governance
pillar score on Yield Spread
Table 3 provides the regression results of the three pillars independently.
Concerning the three pillars independently, the estimation results reveal significant negative effects on yield spread for all three. The results implicate that an
improvement of one of the three pillar scores will reduce yield spread. The effect of the Social pillar score on yield spread is the highest. The Social pillar score has a significant negative effect of 0.014 on yield spread at the 1% level. The
Environmental- and Governance pillar scores effects are also negative and significant at the 1% level (0.009 and 0.007, respectively).
The outcomes of Table 3 support this thesis contends that a higher Environmental-, Social- and Governance pillar score results in a lower yield spread. Moreover, the findings are in line with previous research. For example, other scholars discovered that the Environmental, Social, and Governance practices should be integrated into the credit scores provided by credit rating agencies (Brindelli et al., 2015; Zeidan et al., 2015; Attig et al., 2013).
For the Environmental pillar score, the results extend the findings of Sharfman & Fernando (2008). If a firm incorporates or improves an environmental-active
strategy, this will decrease the cost of debt. The reduction in the cost of debt occurs a year after the renewed strategy. Therefore, we can reject the second null
The effect of the Governance pillar score on yield spread is also in line with previous research. Following Bhorjaj and Sengupta (2003) and Ashbaugh-Skaife et al. (2006), this research finds that improved governance will reduce the cost of debt. Previous research already showed that improved governance led to improved credit ratings. This study extends the findings by revealing that it will also lead to a decrease in yield spread. Therefore, we can reject the third null hypothesis. Namely, an increase in the Governance pillar score will lead to a lower cost of debt.
Lastly, the effect of the Social pillar concerning the cost of debt is examined. In line with Bradley et al. (2007) and Bauer et al. (2009), the research proves that improved social strategy reduces the cost of debt. Those studies focused on more specific aspects of the Social pillar score, like employee management. This research extends this by finding that a higher Social pillar score will lead to a lower yield spread.
Table 3. Relationship between yield spread and ESG pillars independently
(1) (2) (3)
Yield spread Yield spread Yield spread
L.Environment -0.009*** (0.00) L.Governance -0.007** (0.00) L.Social -0.014*** (0.00) L.TA -0.243*** -0.287*** -0.242*** (0.04) (0.04) (0.05) L.ROA -0.049*** -0.050*** -0.046*** (0.01) (0.01) (0.01) L.LEV 0.019*** 0.020*** 0.018*** (0.00) (0.00) (0.00) IssueAmountAnnual -0.004 -0.029 0.010 (0.05) (0.05) (0.05) _cons 8.840*** 10.032*** 9.173*** (0.99) (0.93) (0.95)
Firm_Fixed_Effect Yes Yes Yes
Year_Fixed_Effect Yes Yes Yes
N 1092 1092 1092
r2_b 0.313 0.312 0.320
chi2 3216.98*** 2993.17*** 3345.72***
Robust standard errors in parentheses
= *p<.10 ** p<.05 *** p<.01
This table depicts the regression results of the relationship between the Environmental, Social, and Governance score and yield spread. All three models perform a firm- and year-fixed effects regression. In model 1, the impact of the environmental pillar score on the yield spread is presented. In model 2, the impact of the governance pillar score on the yield spread is presented. In model 3, the impact of the social pillar score on the yield spread is presented. L. Environment is the 1-year lagged Environment Pillar score. L.Social is the 1-year lagged SocialPillar score. L.Governance
is the 1-year lagged Governance Pillar score. LagTotalAssets is the 1-year lagged logarithm of total assets. LagROA is the 1-year lagged operating income before depreciation by the book value of assets. LagLeverage is the 1-year lagged book value of debt divided by the book value
of assets. IssueAmount is the issue amount at the start of the bond. Standard errors are adjusted for heteroskedasticity.
4.3 Effect of Paris Agreement interacted with the combined
ESG-, Environmental-, Social- and Governance pillar score
on Yield Spread
zero if the scores were presented before December 31, 2015. The interaction variable contains the variable Paris interacted with the independent variable (Combined ESG-, Environment-, Social-, and Governance pillar score). This study uses the Paris Agreement as an exogenous shock. In Table 4, Paris has a significant negative effect on yield spread in all the models, except for the Governance pillar score. In Models 1, 2, and 4, Paris has a significant negative effect on yield spread (0.863, 0.813, and 0.960 percentage points, respectively) at the 1% level. Moreover, the ESG combined, Environmental- and Social pillar score has a significant negative effect on yield spread (0.016, 0.007, and 0.015 percentage points, respectively) at the 1% level. Therefore, this study finds that the variables independently have a significant negative effect on yield spread. However, all the interaction variables are insignificant.
Additionally, this study performs a joint hypothesis test to examine the overall significance of the model. The JointF-statistic in Table 4 is significant for both the Environmental- and Social pillar score. Therefore, this study finds significant results for the overall F-test. These conflicting results can come from the difference in assessing the coefficients jointly instead of individually. It could be that the
interaction variables are not predictive enough on their own, and therefore are only significant collectively. Overall, we cannot reject the fifth hypothesis because all the interaction terms are insignificant.
There are a few possible explanations for the results of Table 4. Firstly, we should not forget the incorporation of ESG strategy of companies before the Paris
effect is insignificant. The Paris Agreement tried to create a common cause for firms to enhance environmental and social strategies.
Nevertheless, numerous firms already incorporated an ESG strategy before the end of 2015. As stated before, this study implements the Paris Agreement as an
exogenous shock. Still, a dispersion among firms’ ESG performance before the exogenous shock was present. Moreover, the joint hypothesis test results show the significance of the Environmental and Social pillar scores. This study argues that the ESG strategy before the Paris Agreement creates an inequality, which creates a blur. The blur arose while some firms reacted to the Paris Agreement, and other firms did not have to respond. Therefore, it could be that the new and already existing firms cancel each other out.
Secondly, the Paris Agreement mainly discusses environmental and social causes, which may indicate insignificant outcomes for the Governance Pillar, and
presumably, combined ESG scores. As a result, the joint hypothesis test did not provide significant results at the combined ESG-, and Governance pillar score models.
Nevertheless, the study finds significant influences of the Paris Agreement on the relationship between ESG performance and yield spread. Therefore, this thesis holds that a more in-depth investigation into the Paris Agreement’s impact as
Table 4. Relationship between yield spread and ESG combined score and ESG pillars independently interacted with Paris Agreement
(1) (2) (3) (4)
Yield Spread Yield Spread Yield Spread Yield Spread
L.ESG -0.016*** (-2.92) Paris*L.ESG -0.000 (-0.06) L.Environment -0.007** (-1.94) Paris*L.Environment -0.002 (-0.61) L.Governance -0.005 (-0.63) Paris*L.Governance -0.004 (-0.46) L.Social -0.015*** (-3.33) Paris*L.Social 0.001 (0.25) Paris -0.863*** -0.813*** -0.765 -0.960*** (-2.63) (-3.10) (-1.55) (-3.16) L.TA -0.252*** -0.242*** -0.287*** -0.242*** (-5.92) (-5.61) (-7.04) (-5.32) L.ROA -0.047*** -0.049*** -0.051*** -0.046*** (-4.80) (-4.93) (-5.02) (-4.69) L.LEV 0.019*** 0.019*** 0.020*** 0.018*** (5.69) (5.76) (5.71) (5.35) IssueAmountAnnual -0.023 -0.001 -0.029 0.009 (-0.50) (-0.01) (-0.64) (0.20) _cons 7.626*** 6.599*** 7.869*** 6.843*** (7.18) (5.65) (6.96) (6.04)
Firm_Fixed_Effect Yes Yes Yes Yes
Year_Fixed_Effect Yes Yes Yes Yes
N 1092 1092 1092 1092 r2_b 0.327 0.314 0.312 0.320 F 3334.84*** 3236.85*** 3001.45*** 3339.64*** JointF 0.28 0.38 0.21 0.06 JointP 0.105 0.050 0.647 0.091 t statistics in parentheses = *p<.10 ** p<.05 *** p<.01
This table depicts the regression results of the relationship between the combined ESG, Environmental, Social, and Governance pillar score, and yield spread, with the Paris Agreement's moderating effect. All three models perform a firm- and year-fixed effects regression. In Model 1, the
moderated by the Paris Agreement. L.ESG is the 1-year lagged combined ESG score. L.Environment is the 1-year lagged Environment Pillar score. L.Social is the 1-year lagged SocialPillar score. L.Governance is the 1-year lagged Governance Pillar score. "Paris" is a dummy variable
for the Paris Agreement. LagTotalAssets is the 1-year lagged logarithm of total assets. LagROA is the 1-year lagged operating income before depreciation by the book value of assets. LagLeverage is the 1-year lagged book value of debt divided by the book value of assets. IssueAmount
is the issue amount at the start of the bond. JointF is the overall significance of the model. JointP is the P-value of the overall significance. Standard errors are adjusted for heteroskedasticity.
4.4 Robustness checks
For robustness purposes, this research performs some additional tests. Following the Fama-French 5 factors for Developed Markets, this analysis includes only developed countries in a robustness regression sample. The total number of 2,140 bonds remains in the sample. Appendix Table C1 provides the results of the panel regression for developed countries. The results are very similar to the complete sample regression. All four independent variables still significantly negatively affect yield spread, at the 1% significance level. Therefore, this study finds the same effect of ESG on yield spread in developed and undeveloped countries.
Moreover, following Shaw (2012) and Cooper & Uzun (2015), this study uses companies’ corporate yield as a robustness check. Corporate yield is the same as the price of a corporate bond. The Appendix Tables D1 and D2 show the impact of combined ESG-, Environmental-, Social-, Governance pillar scores, and the
interaction effect with the Paris Agreement on the corporate yield. The results show smaller coefficients, although the coefficients are still negative. These findings are aligned with Cooper & Uzun (2015). However, the impact of the combined ESG score-, Environmental-, Social- and Governance pillar score are all insignificant. Therefore, this study concludes that ESG performance has an insignificant influence on the corporate yield. Previous literature did find a significant relationship between corporate yield and ESG performance. Hence, we should be cautious with
spread, the cost of debt still decreases from an enhanced ESG performance. The difference between the yield spread and the corporate yield is the risk-free rate, or in this study, the U.S. treasury bond. Due to the insignificant results for the corporate yield, we can conclude that the risk-free rate is crucial for the relationship between ESG performance on yield spread. Therefore, this study recommends a more in-depth investigation of the risk-free rate.
5. Discussion
This research analyzes the relationship between ESG performance and the cost of debt. The results indicate a significant relationship, but we should be careful with interpreting these results. There are, however, some limitations to the underlying database and the Bond Fixed Income Indices at the Thomson Reuters database. This section outlines the limitations that are associated with this research. Moreover, it discusses the implications for academia and contributes suggestions for future research directions.
Secondly, the ratings of ESG performance have a backward-looking nature. The applicable rating agency based the ESG scores on the metrics in the latest available year. Although this study implements lagged variables, this can be problematic from an investor’s point of view. The ESG commitment in the past does not guarantee the same Corporate Social Responsibility performance in the future. From an investor’s point of view, future ESG commitment is more important than ESG commitment in the past.
Furthermore, this study only uses the companies present in the ASSET4 database. This database covers 7,200 firms globally. However, there are, by definition, more companies unavailable in the database. Therefore, we should take into account that the evidence is not conclusive in covering all companies globally.
If the ESG rating agency cannot find proportionate criteria in the disclosed
information, the firm is rated zero. A consequence could be biased results because larger, publicly listed firms are required to disclose more information. Therefore, large companies who may perform worse, but can disguise this, will receive higher scores. On the contrary, smaller companies who do good in ESG performance do not disclose enough information, resulting in a rating of zero. Further research is required in examining the difference between larger and smaller firms and the effect of disclosure on ESG ratings. A possible solution to the issue is developing an ESG rating methodology that includes the effect of size.
Concerning the bond data, this thesis finds a mixed relationship between the ESG performance and the cost of debt. Furthermore, a significant impact of ESG
rate. This study uses the U.S. Treasury bond as the risk-free rate. Future research should incorporate a more comprehensive risk-free rate and investigate ESG performance’s role on the risk-free rate.
Moreover, the dispersion of the bond data per year can result in biased results. The Thomson Reuters Fixed Income Indices provides bond data from 2010 onwards, but the first three years only covered thirteen bonds. Therefore, the bond data before 2012 is not comprehensive and makes conclusions over this period not valid.
To conclude, this thesis implemented the Paris Agreement of 2015 as an exogenous shock. By creating a dummy variable, the interaction effect of the Paris Agreement on the relationship between ESG performance and yield spread is measured. The only significant effect is present with a joint hypothesis test. The interaction effects are not significant when measured individually. Therefore, further research on the effect of the Paris Agreement on the cost of debt can provide fruitful outcomes.
6. Conclusion
This thesis empirically researches the influence of ESG performance on the cost of debt. By performing a panel regression concerning the combined ESG ratings and the three pillar ratings independently, I investigate the potential reactions in yield spread of bonds. Despite previous empirical studies focusing on a single-country or continental sample, this thesis uses a worldwide sample. Furthermore, this study researches the moderating effect of the Paris Agreement in 2015 on the relationship between ESG performance and cost of debt.
Following Cooper & Uzun (2015) and Oikonomou et al. (2014), this research creates an empirical model to research the effect of the combined ESG score on yield
spread. Additionally, this study focuses, and thereby extends current research, on both the ESG combined score and the pillar ratings independently. Most studies take a combined ESG score as a proxy to research the effect on the cost of debt or a particular element of the pillar ratings, like employee management.
I retrieve ESG performance data from the ASSET4 Thomson Reuters Eikon
Datastream database. The bond data is retrieved from the Thomson Reuters Fixed Income Indices database.
The main findings of this paper are in line with previous empirical findings. The study finds a significant negative effect of ESG performance on yield spread (Cooper & Uzun, 2015; Oikonomou et al., 2014; Ge & Lui, 2015). Furthermore, this study researches the relationship between Environmental-, Social- and Governance pillar scores and yield spread. A significant negative effect of all three pillars on yield spread is observed. These results are also in line with previous research (Sharfman & Fernando, 2008; Bauer et al., 2009; Bhorjaj & Sengtupa, 2003). The results also show that the Social pillar score has the highest effect on yield spread.
For robustness purposes, this thesis creates a new model wherein only developed countries are included. Following the Fama-French 5 factors for developed countries, this study excludes all undeveloped countries from the sample. The results remain significant, and the coefficients almost unchanged.
Secondly, this thesis implemented corporate yield as a dependent variable. Contrary to yield spread, the impact of ESG on corporate yield is insignificant. This
contradiction is not in line with Cooper & Uzun (2015), who find significant relationships for both the yield spread and the corporate yield. Hence, this study concludes that there should be further research on the relationship between ESG performance and the risk-free rate, while this is the difference between both proxies. The evidence from this research could be conducive for corporate managers. The implications show a special relationship between an improved ESG performance and a lower yield spread. Therefore, this study recommends corporate managers to improve their responsible behavior. A lower yield spread results in a lower overall risk due to a reduced cost of capital.
References
Amiraslani, H., Lins, K. , Servaes, H., Tamayo, A., 2019. The bond market benefits of corporate social capital, Unpublished Working Paper.
Ashbaugh-Skaife, H., Collins, D., & LaFond, R., 2006. The effects of corporate governance on firms’ credit ratings. Journal of Accounting and Economics, 42(1), 203-243.
Attig, N., El Ghoul, S., & Guedhami, O., 2013. Corporate social responsibility and credit ratings. Journal of Business Ethics, 117(4), 679-694.
Barth, F., Huebel, B., Scholz, H., 2020. Esg and corporate credit spreads, Unpublished Working Paper.
Bassen A., Meyer K., Schlange J., 2006. The influence of corporate responsibility on the cost of capital. An empirical analysis. Unpublished Working Paper.
Bauer, R., Derwall, J., & Hann, D., 2009. Employee relations and credit risk. Unpublished Working Paper. Bauer, R., Hann, D., 2010. Corporate environmental management and credit risk, Unpublished Working Paper. Bessembinder, H., Kahle, K., Maxwell, W. & Xu, D., 2009. Measuring abnormal bond performance. Review of Financial Studies, 22(10), 4219-4258.
Bhojraj, S., & Sengupta, P., 2003. Effect of corporate governance on bond ratings and yields: The role of institutional investors and outside directors. The Journal of Business, 76(3), 455-475.
Bradley, M., Chen, D., Dallas, G., & Snyderwine, E., 2007. The relation between corporate governance and credit risk, bond yields and firm valuation.
Birindelli, G., Ferretti, P., Intonti, M., & Iannuzzi, A., 2015. On the drivers of corporate social responsibility in banks: Evidence from an ethical rating model. Journal of Management and Governance, 19, 303-340.
Cai, Y., Jo, H. & Pan, C., 2011. Vice or Virtue? The Impact of Corporate Social Responsibility on Executive Compensation. Journal of Business Ethics, 104, 159-173
Cantiono, V., Devalle, A., Fiandrino S., 2018 ESG sustainability and financial capital structure: where they stand nowadays? International journal of business and science. Vol. 8, No. 5
Capasso, G., Gianfrate, G., and Spinelli, M., 2020. Climate change and credit risk, Journal of Cleaner Production, 266 (2020) 121634.
Caprio, G. and Klingebiel., 2003. Episodes of Systemic and Borderline Financial Crises, The World Bank, January 22nd:
Chava, S., 2014. Environmental Externalities and Cost of Capital. Management Science 60(9):2223-2247.
Chava, S., D. Livdan, and A. Purnanandam., 2009. Do shareholder rights affect the cost of bank loans?, The Review of Financial Studies 22, 2973-3004.
Cheng B, Ioannou I, Serafeim G., 2014. Corporate social responsibility and access to finance. Strategic Management Journal 35:1–23
Climate Bonds Initiative., 2016a. Bonds and climate change: The state of the market in 2016.
Cooper, E., Uzun, H., 2015. Corporate social responsibility and the cost of debt. Journal of Accounting and Finance. Vol.15, No.8, 11-29
Delis, M., de Greiff, K., Ongena, S., 2018. Being Stranded on the Carbon Bubble? Climate Policy Risk and the Pricing of Bank Loans. CEPR Discussion Paper No. 12928
Dorfleitner, G., Halbritter, G., Nguyen, M., 2015. Measuring the level and risk of corporate responsibility – An empirical comparison of different ESG rating approaches. Journal of Asset Management, 16(7), 450-466.
Eccles R., Ioannou I., Serafeim, G., 2014. The impact of corporate sustainability on organizational processes and performance. Management Science 60:2835–2857
El Ghoul S., Guedhami O., Kwok C., Mishra D., 2011. Does corporate social responsibility affect the cost of capital? Journal Banking and Finance 35:2388–2406
Fatemi, A., Glaum, M., & Kaiser, S., 2017. ESG performance and firm value: The moderating role of disclosure. Global Finance Journal.
Friede, G., Busch, T., & Bassen, A., 2015. ESG and financial performance: Aggregated evidence from more than 2000 empirical studies. Journal of Sustainable Finance & Investment, 5(4), 210-233.
Galema, R., Plantinga, A. and Scholtens, B., 2008. The stocks at stake: Return and risk in socially responsible investment. Journal of Banking and Finance, 32(12), 2646– 2654.
Ge, W., & Lui, M., 2015. Corporate social responsibility and the cost of corporate bonds. Journal of Accounting and Public Policy, 597-624.
Godfrey, P., Merrill C., Scholtens, B., 2009. "The relationship between corporate social responsibility and shareholder value: an empirical test of the risk management hypothesis." Strategic Management Journal 30(4): 425-445
Görgen, M., Jacob, A., Nerlinger, M., Riordan, R., Rohleder, M. and Wilkens, M., 2018. Carbon Risk. Augsburg, Germany: University of Augsburg.
Goss, A., Roberts, G., 2011. The impact of corporate social responsibility on the cost of bank loans. Journal of Banking and Finance, 35(7), 1794-1810.
Halling, M., Yu, J., Zechner, J., 2020. Primary corporate bonds markets and social responsibility, Unpublished Working Paper. Harrison, J., & Wicks, A., 2013. Stakeholder theory, value, and firm performance. Business Ethics Quarterly, 23(1), 97–124. Jiang, J., 2008. Beating earning benchmarks and the cost of debt. Accounting review, 83(2), 377-416.
Justensen, P., 2009. Ratings recall: will new reform make lasting impact? Journal of Corporation Law, Vol.35, pp.193-213
Kaplan, R., Norton, D., 1996. “Using the balanced scorecard as a strategic management system”, Harvard Business Review, Vol. 74 No. 1, pp. 75-85.
Lee, D., Faff, R., 2009. Corporate sustainability performance and idiosyncratic risk: A global perspective. The Financial Review, 44, 213-237.
Lins, K., Servaes, H., 2017. Social capital, trust, and firm performance: The value of corporate social responsibility during the financial crisis. The Journal of Finance.
Martynova M, Renneboog L., 2008. A Century of Corporate Takeovers: What Have We Learned and Where Do We Stand? Journal of Banking and Finance 32: 2148-2177.
McGuire, J., Sundgren, A., & Schneeweis, T., 1988. Corporate social responsibility and firm financial performance. Academy of Management Journal, 31(4), 854-872.
Menz, K., 2010. Corporate social responsibility: Is it rewarded by the corporate bond market? A critical note. Journal of Business Ethics, 96(1), 117-134.
Oikonomou, I., Brooks, C. & Pavelin, S., 2014. The effects of corporate social performance on the cost of corporate debt and credit ratings. Financial Review, 49, 49-75.
Orlitzky, M., Schmidt, F., & Rynes, S., 2003. Corporate social and financial performance: A meta-analysis. Organization Studies, 24(3), 403-441.
Ortiz-Molina, H., 2006. Top Management Incentive and the Pricing of Corporate Public Debt. Journal of Financial and Quantitative Analysis, 41(2), 317-340.
Palea, V., and Drogo, F., 2020. Carbon emissions and the cost of debt in the eurozone: The role of public policies, climate-related disclosure and corporate governance Business Strategy and the Environment, 1–20.
Refinitiv., 2020. Environmental, Social and Governance scores from Refinitiv.
Revelli, C., & Viviani, J., 2015. Financial performance of socially responsible investing (SRI): What have we learned? A meta-analysis. Business Ethics: A European Review, 24(2), 158-185.
Schauten, M., van Dijk, D., 2010. Corporate governance and the cost of debt of large European firms. (Report Series; No. 2010-025-F&A). Rotterdam: Erasmus Research Institute of Management.
Schneider, T., 2011. Is environmental performance a determinant of bond pricing? evidence from the u.s. pulp and paper and chemical industries. Contemporary Accounting Research 28, 1537–1561.
Shaw, K., 2012. CEO incentives and the cost of debt. Review of Quantitative Finance and Accounting, 38, 323-346.
Shi, C., 2003. On the trade-off between the future benefits and riskiness of R&D. Journal of Accounting and Economics, 35(2), 227-254.
Stellner, C., Klein, C., Zwergel, B., 2015. Corporate social responsibility and eurozone corporate bonds: The moderating role of country sustainability. Journal of Banking and Finance 59, 538– 549.
Utz, S. and Wimmer, M., 2014. Are they any good at all? A financial and ethical analysis of socially responsible mutual funds. Journal of Asset Management, 15(1), 72-82.
Wang, A. & Zhang, G., 2009. Institutional ownership and credit spreads: An information asymmetry perspective. Journal of Empirical Finance, 16(4), 597-612.
Appendix
Variable Definition
Variable Definitions
IssueAmount The amount of the bond at issuance (x 1000), in dollars
Corporate Yield % Corporate bond price
Yield Spread % The difference between the corporate bond price and the risk-free rate
ESG The combined ESG pillar score
Environment The Environmental pillar score
Governance The Government pillar score
Social The social pillar score
Total Assets The logarithm of the total assets per firm
Return on Assets % The operating income before
depreciation divided by the book value of the assets
Table A1. Previous literature on the relation between ESG and cost of capital
Study Methodology Time
Period Sample Data collection Findings Menz, 2010 Quantitative (OLS regression) 2004 - 2007 Total of 498 bonds Merrill Lynch Index System Due to CSR a higher bond spread in Europe. Ge & Lui, 2015 Quantitative (Multivariate regression) 1992 - 2009 Total of 4260 bond issues, which come from 2317 different firms RiskMetric s group, KLD database, Compustat CSR creates bonds at lower cost. Barth et al., 2020 Quantitative (Time-Series analysis) 2009 - 2016 108 Cedit Defaults Swaps each referring to a single firm Thomson Reuters Eikon Reduction in credit spreads for firms with high ESG scores. Cooper &
Uzun, 2015 Quantitative (Cross-sectional regression) 2006 - 2013 Total of 2252 bonds. Mergent Fixed Income Securities Database and Bloomberg database Firms with strong CSR have a lower cost of debt. Oikonomou et al., 2015 Quantitative (Cross-sectional) 1991 - 2008 3240 bonds issued KLD Stats Database Good CSR performance improves bond prices and ratings. Cantino et
al., 2017 Qualitative research No clear-cut boundaries
Table A2. Previous literature on the relation between Environmental pillar and cost of capital
Study Methodology Time
Period Sample Data collection Findings Bauer & Hann, 2010 Quantitative research 1995 - 2006 582 U.S. public corporation s KLD Stats Database Environmental concerns leads to larger bond spreads Schneider, 2011 Quantitative research 1999 - 20005 533 Corporate Bonds KLD Stats Database Better environmental performance lower corporate bond spreads in secondary market Chava,
2014 Quantitative research 1992-2007 13,114 firms. Compustat Environmental problems
Table A3. Previous literature on the relation between Social pillar and cost of capital
Study Methodolog
y
Time Period Sample Data
collection Findings Amisraslani et al., 2019 Quantitative research 2005 - 2013 2212 bonds Trace Database Social capital only positively influenced bond spreads during the financial crisis. Halling et
Table A4. Previous literature on the relation between Governance pillar and cost of capital
Study Methodolog
y
Time Period Sample Data
collection Findings Bhorjaj & Sengupta, 2003 Quantitative research 1991 - 1996 2098 bond issues Warga Fixed Income Database Higher governance leads to lower credit yields and higher bond ratings Chava et al., 2009 Quantitative research 1990 - 2004 6000 loans issued Compustat Global A lower takeover defense results in higher interest loans Schauten & van Dijk, 2011
Quantitative 2000 - 2008 186 firms Deminor
Table B1. Bonds per country
Country Average Yield Yield Spread No. of Bonds Issue Amount (Bil.)
THAILAND 5.764 2.294 2 0.75 TURKEY 5.279 4.094 4 2.00 UNITED KINGDOM 4.270 2.985 76 58.77 UNITED STATES 3.933 1.543 1,674 1,342.80 All 3.808 1.513 2,280 1,806.41
The bold countries are the developed countries according to the Fama-French 5 Factors model for Develop Countries
Table B2. Bonds by SIC_GEN
SIC_GEN Average Yield Yield Spread No. of Bonds Issue Amount (Bil.)
Construction 7.121 3.852 63 31.70 Finance 3.567 1.322 214 140.61 Manufacturing 3.609 1.366 965 734.50 Mining 5.466 3.298 173 114.67 Retail Trade 3.548 1.103 174 145.44 Services 3.696 1.400 212 207.07 Transportation & Utilities 3.675 1.369 441 407.98 Wholesale Trade 3.820 1.414 38 24.44 All 3.808 1.513 2,280 1,806.41
Table B3. Bonds by year
year Average Yield Yield Spread No. of Bonds Issue Amount (Bil.)
Table C1. Effect of ESG on yield spread with developed countries.
(1) (2) (3) (4)
Yield Spread Yield Spread Yield Spread Yield Spread
L.ESG -0.016*** (0.00) L.Environment -0.009*** (0.00) L.Governance -0.005*** (0.00) L.Social -0.015*** (0.00) L.TA -0.155*** -0.300*** -0.360*** -0.298*** (0.02) (0.05) (0.05) (0.06) L.ROA -0.044*** -0.054*** -0.056*** -0.051*** (0.01) (0.01) (0.01) (0.01) L.LEV 0.022*** 0.018*** 0.018*** 0.016*** (0.00) (0.00) (0.00) (0.00) IssueAmountAnnual -0.105** 0.054 0.028 0.069 (0.06) (0.05) (0.04) (0.05) _cons 9.810 8.674*** 9.985*** 9.068*** (1.04) (1.08) (0.99) (1.05)
Firm_Fixed_Effects Yes Yes Yes Yes
Year_Fixed_Effect Yes Yes Yes Yes
N 1025 1025 1025 1025
r2_b 0.353 0.348 0.362
Standard errors in parentheses
= *p<.10 **p<.05 ***p<.01
This table depicts the regression results of the relationship between the combined ESG, Environmental, Social, and Governance score, and yield spread, but only including developed countries according to the Fama-French 5 factors model. All three models perform a firm- and year-fixed
effects regression. In model 1, the impact of the combined ESG score on the yield spread is presented. In model 2, the impact of the environmental pillar score on the yield spread is presented. In model 3, the impact of the governance pillar score on the yield spread is presented.
In model 4, the impact of the social pillar score on the yield spread is presented. L. Environment is the 1-year lagged Environment Pillar score. L.Social is the 1-year lagged SocialPillar score. L.Governance is the 1-year lagged Governance Pillar score. LagTotalAssets is the 1-year lagged
logarithm of total assets. LagROA is the year lagged operating income before depreciation by the book value of assets. LagLeverage is the 1-year lagged book value of debt divided by the book value of assets. IssueAmount is the issue amount at the start of the bond. Standard errors are
Table D1. Effect of the ESG pillars on the corporate yield
(1) (2) (3)
Corporate Yield Corporate Yield Corporate Yield
L.Environment 0.006 (0.01) L.Governance -0.001 (0.01) L.Social -0.018 (0.02) L.TA -0.232 -0.210 -0.151 (0.37) (0.38) (0.40) L.ROA -0.029* -0.030* -0.030* (0.02) (0.02) (0.02) L.LEV 0.015 0.014 0.011 (0.01) (0.01) (0.01) IssueAmountAnnual 0.014 0.012 0.010 (0.05) (0.05) (0.05) _cons 6.940 6.982 7.175 (5.83) (6.12) (5.59)
Firm_Fixed_Effect Yes Yes Yes
Year_Fixed_Effect Yes Yes Yes
N 1092 1092 1092 r2_b 0.147 0.207 0.280 Chi2 5.25*** 5.61*** 5.49*** Standard errors in parentheses = *p<.10 ** p<.05 *** p<.01
This table depicts the regression results of the relationship between the Environmental, Social, and Governance score and corporate yield. All three models perform a firm- and year-fixed effects regression. In model 1, the impact of the environmental pillar score on the corporate yield is
presented. In model 2, the impact of the governance pillar score on the corporate yield is presented. In model 3, the impact of the social pillar score on the corporate yield is presented. L. Environment is the 1-year lagged Environment Pillar score. L.Social is the 1-year lagged SocialPillar score. L.Governance is the 1-year lagged Governance Pillar score. LagTotalAssets is the 1-year lagged logarithm of total assets. LagROA is the 1-year lagged operating income before depreciation by the book value of assets. LagLeverage is the 1-year lagged book value of debt divided by
Table D2: Moderating effect of the Paris Agreement on the impact of the ESG pillars on the corporate yield
(1) (2) (3) (4)
Corporate Yield Corporate Yield Corporate Yield Corporate Yield
L.ESG -0.004 (-0.41) Paris*L.ESG 0.000 (0.01) L.Environment 0.008 (1.02) Paris*L.Environment -0.004 (-0.99) L.Governance -0.000 (-0.03) Paris*L.Governance -0.001 (-0.07) L.Social -0.020 (-1.22) Paris*L.Social 0.005 (1.04) Paris -0.142 0.032 -0.116 -0.311 (-0.41) (0.13) (-0.18) (-0.90) L.TA -0.202 -0.283 -0.213 -0.126 (-0.55) (-0.79) (-0.58) (-0.32) L.ROA -0.030* -0.029* -0.030* -0.030** (-1.93) (-1.82) (-1.86) (-1.98) L.LEV 0.014 0.014 0.014 0.012 (1.51) (1.51) (1.55) (1.21) IssueAmountAnnual 0.010 0.019 0.012 0.007 (0.18) (0.37) (0.21) (0.14) _cons 7.191 7.832 7.173 6.943 (1.30) (1.40) (1.23) (1.25)
Firm_Fixed_Effects Yes Yes Yes Yes
Year_Fixed_Effect Yes Yes Yes Yes
N 1092 1092 1092 1092 r2_b 0.226 0.166 0.208 0.272 F 5.28*** 4.84*** 6.71*** 5.16*** JointF 0.00 0.98 0.00 1.08 JointP 0.990 0.322 0.944 0.300 t statistics in parentheses = *p<.10 ** p<.05 *** p<.01
This table depicts the regression results of the relationship between the combined ESG, Environmental, Social, and Governance pillar score and corporate yield, with the moderating effect of the Paris Agreement. All three models perform a firm- and year-fixed effects regression. In model 1, the impact of the ESG combined score on the corporate yield is presented, moderated by the Paris Agreement. In model 2, the impact of the environmental pillar score on the corporate yield is presented, moderated by the Paris Agreement. In model 3, the impact of the governance pillar
Environment Pillar score. L.Social is the 1-year lagged SocialPillar score. L.Governance is the 1-year lagged Governance Pillar score. Paris a dummy variable for the Paris Agreement. LagTotalAssets is the 1-year lagged logarithm of total assets. LagROA is the 1-year lagged operating income before depreciation by the book value of assets. LagLeverage is the 1-year lagged book value of debt divided by the book value of assets.
