An Empirical Analysis of the Green Label Effect
On the euro-issued bonds on the French market
Master Thesis
Copenhagen Business school
Authors:
Bergdís Bergsdóttir (125129) Maria Louise Rasmussen (102162)
Supervisor:
Lars Sønnich Pørksen
Number of pages: 92
Characters including spaces: 163,188 Date: 14.05.2020
Abstract
The green bond market has been growing intensively in recent years, and the total green bond issuance is expected to reach USD350bn in 2020. The intense growth of the green bond market can be traced to increased climate and global warming awareness. However, the market is still quite young and is a relatively small percentage of the total bond issuance. This results in a lack of sufficient available research analysing different aspects of green bonds. In addition, there is a lack of common green bond framework and definition, which can be problematic when assessing green bonds. However, following this green bond boom, researchers seem to be more interested in studying them.
In this thesis, we analyse the effect of the green label on bonds. The method used is the propensity score matching method. We match on the propensity score, using four different matching methods. That is, the nearest neighbour matching, radius matching, stratification matching and kernel matching. The bonds analysed are from the French market, issued in euros. France is the biggest European green bond issuer, and one of the three top ranked issuers globally.
We attempt to capture the effect of the green label by comparing green bonds to their closest conventional peer. We found that investors are receiving lower yields when investing in green bonds compared to investing in conventional bonds. It was not possible to conclude on the yield in relation to credit ratings. Additionally, we analysed the difference in the credit spread, and could not confirm that the negative effect on the yield was
compensated by lower spreads. Moreover, it was not possible to conclude that green bonds are more or less liquid on the secondary market than their conventional peers. Thus, the green investors are accepting lower yields without being compensated. Therefore, the combined results indicate that investing in green bonds might have some psychological benefits compensating for the lower yield or that the market demand is unsatisfied.
Table of Contents
1. Introduction ... 7
1.1 Problem Statement ... 8
1.2 Delimitation ... 9
1.3 Structure ... 10
2. Methodology ... 11
2.1 Study Design and Science-Theoretical Approach ... 11
2.2 Method ... 12
2.2.1 Data ... 12
2.2.2 Quality Criteria ... 13
2.2.3 Table for Quality Report ... 14
2.2.4 Validity ... 14
2.2.5 Reliability ... 15
2.2.6 Sufficiency ... 15
2.2.7 Summary of quality requirements ... 16
3. Green Bonds ... 17
3.1 Green Bond Principles ... 18
3.2 Green Bond Standard ... 20
3.3 Beginning and History ... 24
3.4 Green Bond Market Today ... 25
3.5 Motivation and Criticism ... 27
3.6 Bloomberg’s Green Label ... 29
4. Measures for comparing ... 32
4.1 Yield to Maturity ... 32
4.2 Credit spread ... 33
4.3 Liquidity ... 34
5. Literature Review ... 35
6. Data Set ... 39
7. Empirical Framework ... 43
7.1 Framework Guidelines ... 44
7.1.1 Rubin-Roy Model ... 44
7.1.2 Parameter of Interest and Selection Bias ... 45
7.1.3 Conditional Independence Assumption ... 46
7.1.4 Common Support ... 47
7.1.5 Summary ... 47
7.2 The Propensity Score ... 47
7.3 Evaluation Methods ... 48
7.3.1 Nearest Neighbour Matching ... 50
7.3.2 Caliper and Radius Matching ... 51
7.3.3 Stratification Matching ... 52
7.3.4 Kernel Matching ... 52
7.4. Connection to data ... 53
8. Matching & Analysis ... 55
8.1 Difference in yield to maturity ... 57
8.2 Difference in Credit Spread ... 70
8.3 Difference in Liquidity ... 76
9. Discussion ... 84
9.1 Limitations ... 84
9.2 Yield to Maturity ... 85
9.3 Credit Spread ... 87
9.4 Liquidity ... 88
9.5 Other factors ... 88
9.5.1 Behavioural finance ... 88
9.5.2 Demand and supply ... 89
9.5.3 Clientele effect ... 89
10. Conclusion ... 91
Bibliography ... 93
Lists of figures
Figure 1. Interactions between the four components of the EU GBS. (EU TEG on Sustainable Finance,
2019). ... 23
Figure 2. Total green bond issuance in USD equivalent (billions). *Estimated issue amount in 2020. (Climate Bonds Initiative, n.d. d). ... 26
Figure 3. Use of proceeds allocation. (Climate Bonds Initiative, 2020). ... 27
Figure 4. Yield curves. (Berk & Demarzo, 2017). ... 33
Figure 5. Different matching methods. Self-effected. ... 49
Figure 6. Propensity scores for treated and control groups. (Katchova, 2013). ... 49
Figure 7. Nearest neighbour matching. (Katchova, 2013). ... 50
Figure 8. Radius matching. Inspired by Katchova, 2013. ... 51
Figure 9. Stratification Matching. Inspired by Katchova, 2013. ... 52
Figure 10. Kernel matching. (Katchova, 2013). ... 53
Figure 11. Matching process. ... 55
Figure 12. PS Graph - YTM ... 60
Figure 13. PS Graph – Credit Spread ... 72
List of tables
Table 1. Quality report chart. Source: Self-effect with inspiration from (Olsen, 2018). ... 14Table 2. Acceptable use of proceeds types. (Bloomberg, 2015). ... 30
Table 3. Data Set Variables ... 40
Table 4. Rating sub-classification ... 42
Table 5. Time to maturity sub-classification ... 42
Table 6. Independent variables. ... 56
Table 7. Dependent variables. ... 56
Table 8. PS Test - YTM ... 58
Table 9. Total bonds - YTM ... 58
Table 10. Probit Regression – YTM ... 59
Table 11. Estimated Propensity Score - YTM ... 60
Table 12. Balancing Property - YTM ... 61
Table 13. Nearest Neighbour Matching – YTM ... 62
Table 14. Radius Matching (0,1) – YTM ... 62
Table 15. Radius Matching (0,2) – YTM ... 63
Table 16. Stratification Matching – YTM ... 63
Table 17. Kernel Matching – YTM ... 64
Table 18. Sub-classification YTM: Rating AAA to AA- ... 66
Table 19. Sub-classification YTM: Rating A+ to A- ... 67
Table 20. Sub-classification - YTM: > 10 years ... 68
Table 21. Sub-classification - YTM: =< 10 years ... 69
Table 22. PS Test - Credit Spread ... 70
Table 23. Total bonds - Credit Spread ... 71
Table 24. Probit Regression - Credit Spread ... 71
Table 25. Estimated Propensity Score - Credit Spread ... 72
Table 26. Balancing Property - Credit Spread ... 73
Table 27. Nearest Neighbour Matching - Credit spread ... 74
Table 28. Radius Matching (0,1) - Credit Spread ... 74
Table 29. Radius Matching (0,2) - Credit Spread ... 74
Table 30. Stratification Matching - Credit Spread ... 75
Table 31. Kernel Matching - Credit Spread ... 75
Table 32. Nearest Neighbour Matching – Liquidity ... 76
Table 33. Radius Matching (0,1) – Liquidity ... 77
Table 34. Radius Matching (0,2) – Liquidity ... 77
Table 35. Stratification Matching – Liquidity ... 77
Table 36. Kernel Matching – Liquidity ... 78
Table 37. Sub-classification - Liquidity: Rating AAA to AA- ... 79
Table 38. Sub-classification - Liquidity: Rating A+ to A- ... 80
Table 39. Sub-classification - Liquidity: > 10 years ... 81
Table 40. Sub-classification - Liquidity: =< 10 years ... 82
Appendices
Appendix A. Data Set……….….. 96Appendix B. Yield to Maturity Output………..……… 96
Appendix C. Yield to Maturity Rating Subclass Output………..………….……. 103
Appendix D. Yield to Maturity Time Subclass Output ……….…….. 118
Appendix E. Credit Spread Output ……….….………….. 133
Appendix F. Liquidity Output ……….………….……….. 141
Appendix G. Liquidity Rating Subclass Output ………..….……….. 148
Appendix H. Liquidity Time Subclass Output………..….………….…. 162
1. Introduction
The green bond market has been growing intensely in recent years. Green bond proceeds go towards projects which are meant to have a positive environmental or climatic effects. The specific use of funds raised is what distinguishes green bonds from other types of bonds (World Bank, 2015). Climate Bonds Initiative estimates the total green bond issuance in 2020 to reach USD350bn (Climate Bonds Initiative, n.d.d). Entities have become more and more interested in issuing green bonds to fund sustainable projects, and there has also been a lot of demand coming from investors. This trend can be traced to increased awareness of climate changes and global warming. It is undeniable that the focus on climate change is increasing. With the Paris Agreement in 2015, drastic measures were taken, when the United Nations signed a binding agreement to combat climate change, and to accelerate and
intensify actions and investments needed for a sustainable low carbon future. The main goal of the Paris Agreement is to keep the global temperature rise below 2 degrees Celsius. This action had a major effect on the green bond market, as the goals of the agreement depend to a large extent on green sustainable investments (United Nations. n.d.). Throughout the years, environmental awareness had not been a high priority with regards to investment policies, but that is no longer the case.
The green bond market today is still quite young and comprises a relatively small percentage of the total bond issuance. There is a lack of sufficient research available regarding different aspects of green bonds, even though researchers have become more and more interested in green bonds recently. In addition, the lack of a common green bond definition and framework can be problematic (Pronina, 2019).
Many previous studies have been analysing green bonds using matching methods, comparing green bonds to conventional bonds. That is, comparing them to a bond that shares mostly the same characteristics except for the green factor. Most of the previous studies have been comparing the bonds across both countries and currencies, matching on different characteristics, some with large variation on matching factors. Therefore, it is interesting to analyse green bonds with narrower terms of comparison, using propensity score matching. Propensity score matching is not typically used when analysing bonds, but it is suited for the empirical analysis at hand. The method is suited for an empirical setting where there is a so-called treatment, and that applies to our case, as we view the process of
getting a green label as a treatment. Thus, the green bonds are the bonds that make up the treatment group in our analyses, and bonds without the green label make up for the
comparison group, sometimes referred to as the control group. We believe that the
calculated propensity score results in more unbiased matches compared to other methods, such as regression. Thus, making this method fitting for our analysis of the green label.
1.1 Problem Statement
The purpose of this thesis is to evaluate, through the propensity score techniques, several differences between green bonds and their conventional peers. The thesis contains a theoretical account of green bonds, as well as the mathematical methods used to perform the comparison. In addition, an analytical section applies the aforementioned methods, in the comparison between the green bonds and their conventional peers. We have seen from previous studies that investors seem to be excepting lower yields when investing in green bonds. Therefore, we find it interesting to analyse whether we can detect a difference in the yield between green and conventional bonds. Furthermore, we find that other measures are equally interesting to analyse in order to provide the investor with an overview of which aspects they should take into consideration when they invest in green bonds. The overall research question that will be investigated is stated as follows:
“How does the green label impact bonds?”
To help answer this question a series of sub questions are posed, referring to the difference between green bonds and their conventional peers:
We want to investigate the yield to maturity for green bonds, that reflects the true underlying interest rate of return for the investor.
“Is there a difference in yield to maturity?”
We need to access the risk of the bonds, and we do so by looking at the magnitude of the credit spreads.
“Is there a difference in the credit spread?”
We also find it relevant to look into the liquidity of the bonds, and how easily the bonds can be sold on the secondary market.
“Is there a difference in liquidity?”
Furthermore, to understand the green label in more depth, we find it beneficial to use sub classification in our analysis. We will be looking at sub classes with regards to both ratings
and time to maturity of the bond.
“Do the measures vary in sub classification?”
1.2 Delimitation
As mentioned, this thesis focuses on analysing the impact of the green label that can be applied to certain types of bonds. There are certain challenges facing the green bond market, mainly the lack of a common green bond definition and framework. Since there is no
universal definition of the green label, issuers and external verifiers do not always work in the same manner. There are also several databases that report on green bonds, and they use different methods to determine the green label. To avoid inconsistency, we chose to only collect data from one database. We chose to collect data from the Bloomberg Terminal (Bloomberg LP), which according to CBS, is the largest and best source of financial market data in the world. Bloomberg has their own labelling system, which will be covered in detail in chapter 3.6.
We wanted to analyse the green bond label with narrower terms of
comparison. Therefore, unlike many other previous studies, we decided to eliminate both country and currency exposure and investigate one currency within the same market. We wanted to analyse a European market. The market of France is an appropriate market, as France is one of the top three ranked global issuers, as well as the biggest issuer in Europe.
We wanted to eliminate as much of the currency exposure as possible, and thus chose to only look at bonds issued in euros.
We decided to eliminate solvency capital requirement from our thesis. Even though solvency capital requirements are important for investors, and will affect which assets the investor invests in. The reason for this delimitation is that with the data at hand we would not be able to capture the solvency of the bonds. Furthermore, we did not have the option to look at the individual investors collected debt. Thus, it will not be within the scope of this thesis.
Finally, we did not find any tax incentives for green bonds on the French market, which will be addressed in chapter 5. Thus, our thesis will not look into greenwashing, policy risk or other side effect of tax incentives for green bonds.
1.3 Structure
The remainder of this paper is structured as follows: In chapter 2 we clarify the methodology used in this thesis. Chapter 3 provides a thorough review of green bonds. Chapter 4 reviews relevant measurements, that are used the comparative analysis. Chapter 5 contains literature review and overview of what some of the previous studies show. In chapter 6 the data set is addressed. Chapter 7 addresses the empirical framework that serves as a guideline for our analysis, the propensity score and the different evaluation methods used. In chapter 8 we perform the matching and analyse the output. In chapter 9 we discuss the implications of our results and how it contributes to previous findings. Finally, in chapter 10 we summarize and conclude on our findings.
2. Methodology
The purpose of this section is to clarify the methodology used in this thesis. The aim is to clarify the research design of the thesis, and the essential epistemological, methodological and ontological reflections made in the preparation of the research design, to provide clarity on the project's theoretical baseline.
2.1 Study Design and Science-Theoretical Approach
The research strategy will be based on a quantitative method which aims to produce an empirical analysis of the two types of bonds.
The quantitative data collection consists of data sets collected from the Bloomberg Terminal, and the data is thereby classified as secondary data. The calculations and the use of the statistical model such as the propensity score model points towards a positivistic and mathematical approach (Holm,2016).
The science-theoretical approach of the thesis is identified within the positivist paradigm. Positivism is characterized by the fact that knowledge is obtained as A posterior realization, through sensory experiences. The methodology of positivism is primarily
induction, i.e. based on quantitative observations and nothing else. Thus, a summary of a lot of individual observations becomes general scientific theories (Holm,2016).
In addition, the methodology deduction is used in positivism to prove theories, which can be considered a way to verify. The ontological direction of positivism is realistic, which means that reality exists and therefore positivism seeks to create universals. The epistemological direction is objective and dualistic. Positivism is dualistic because there is true science that is objective and not science that is metaphysics. Positivism is objective because it is based only on specific sensory experiences. Working with positivism as a science builds on past observations, and thus the deductive methodology can help to verify one's theory. According to positivism, the accumulation of knowledge collection produces a steady and continuous evolution, and positivist science is thus characterized by being a cumulative science. As mentioned, positivism works with verification because positivists consider scientifically true theories to be provable (Holm,2016).
Other science-theoretical approaches for instance hermeneutics, critical theory and constructivism are abandoned in the thesis, as the thesis aspires to find a difference between green and conventional bonds based on quantitative data. The scope of the thesis is
not to investigate other factors such as investor preferences and that is why qualitative data should not be used.
2.2 Method
In order to prepare an analysis, it has been necessary to collect secondary data. Data has been collected to identify the difference in yield to maturity, credit risk and liquidity between green and conventional bonds in the French market. When working with secondary data, errors in original data should be taken into account. In section 2.2, a quality description of the data collected will be reviewed.
2.2.1 Data
This section presents and discusses the data that forms the basis for the empirical analysis of this thesis. Initially, thoughts and considerations related to data selection will be described, and then data collection. We would like the reader to know that we mention data limitations serval times throughout the paper because during the making of this thesis, school campus was closed due to COVID-19. Thus, data could no longer be accessed. Faced with two options, either re-designing the research towards a primarily literature-based thesis or work with the data at hand, we chose to work with the data we had already collected.
2.2.1.1 Selection of Data
Our data samples are set up to evaluate, through the propensity score matching techniques, several differences between green bonds and their conventional peers.
The measures that the thesis focuses on are:
1. Difference in yield to maturity between green bonds and their conventional peers.
2. Difference in credit spread between green bonds and their conventional peers.
3. Difference in liquidity between green bonds and their conventional peers.
Furthermore, we also use sub-classification with regards to both ratings and time to maturity for a better analysis.
The analyses are performed on the French bond market and will consist of bonds issued in euros. The data comes from the Bloomberg Terminal. For every bond, the Bloomberg Terminal provides detailed information about the bond issues’ and issuers characteristics.
This approach is chosen in order to examine the difference between green and conventional bonds from an empirical foundation and to avoid country and currency
exposure affecting our analysis.
2.2.1.2 Data Collection
The empirical data for both green and conventional bonds was primarily collected on February 7, 2020 (Bloomberg Terminal, 2020). The issue date period 2014-2019 is chosen based on data availability for French green bonds. Later in the process, measures for both green and conventional bonds were collected via the CBS library from the Bloomberg Terminal and merged with our “original” data set.
The total data sample consist of 43 green bonds and 340 conventional bonds.
The different characteristics of the bonds and measures added are described in depth in chapter 6.
2.2.2 Quality Criteria
In this section, the part of the method that deals with justification is clarified. That is, how the said methods ensure truth in data collection. For this purpose, quality description is used by Olsen (2018). According to Olsen, quality can be divided into three quality types, validity, reliability and sufficiency.
2.2.3 Table for Quality Report
Table 1. Quality report chart. Source: Self-effect with inspiration from (Olsen, 2018).
Purpose Question Data Validity Reliability Sufficiency Compare
green and conventional bonds based on empirical
data
Is the internal rate of green bonds higher or lower than the internal
rate of conventional
bonds?
Yield to
maturity The statistical validity of this measure can
be criticised
All measures are collected
from the Bloomberg-
Terminal.
Bloomberg is a recognized source. Hence,
backing is achieved by retrieving data
from this source.
Good for the purpose, as we gain more
insight into how the green
and the conventional bond market differentiate.
However, not a 100%
sufficient to make a generalization,
but will allow us to investigate the
France market which is one
of the big markets when
dealing with green bonds.
Is the risk associated with investing
in green bonds different from
conventional bond?
I-spread The statistical validity is good for corporate bonds.
Is liquidity lower or higher in the
green bond market than
the conventional bond market?
Difference between bid and ask
price The statistical validity of this measure can be criticised
2.2.4 Validity
Validity covers the scope of a concept or dataset. That is, an assessment of what it covers or what the results can be used for. Thus, this quality criterion is considered the most important of the three. Validity is divided into technical, internal, statistical and external validity.
Technical validity relates to whether operational data representation is chosen correctly in relation to the concept being investigated. The technical validity of this thesis is considered to be respected as the data source selected, are recognized in data collection and analysis.
Internal validity deals with causality. That is, if there is a clear causal
relationship in the observations or if there are any unresolved causal relationships. The thesis
shows a clear relationship between empiricism and theory, which is explained throughout the thesis.
Statistical validity illuminates the statistical analysis design. Here, the project sees a critical point regarding the statistical validity of all measures. Statistical validity of yield to maturity: Optimally we would like to examine both yield at issue and yield to maturity at several different points to detect both if there is a premium at issue and a premium after trading has commenced. However, due to data limitation we were only able to get yield to maturity at one point in time and therefore the results will only be valid at this point.
Statistical validity of I-spread: the I-spread could be problematic in relation to government bonds as one would have to assume that the government bonds share the same risk as corporate bonds. This issue was dealt with by only using the I-spread for analysing corporate bonds and therefore our risk analysis is missing the government bond perspective. The G- spread was available to us and would be considered a good measure to analysis the risk of government bonds. However, we only had 14 green government bonds and 18 conventional government bonds, thus making the amount of conventional government bonds inadequate for propensity score matching. Statistical validity of liquidity: the difference between the bid and ask price is one measure that can illuminate liquidity of a bond. Optimally we would like to examine both difference between bid and ask price and LOT liquidity. However due to data limitation we were only able get data necessary to calculate the bid-ask price liquidity.
External validity addresses whether the study provides an opportunity to generalize. As mentioned, the thesis seeks to compare green and conventional bonds on the France market, and it is therefore uncertain whether the same results will be applicable to similar studies.
2.2.5 Reliability
Reliability assesses the soundness of data relative to the method of collection. As mentioned, this is secondary data from reliable sources, and therefore reliability is considered to be high.
2.2.6 Sufficiency
The sufficiency sheds light on whether the study is comprehensive, i.e. whether the sub- questions constitute satisfactory answers to the research question. Data is considered to be good for the purpose, as it provides an insight into how the model is influenced by historical
empirical data. However, it will not be 100% sufficient to make a generalization but will provide an opportunity to elucidate several aspects of the models.
2.2.7 Summary of quality requirements
Based on the above, the data quality is considered moderate. The requirement for validity is generally considered to be complied with, but the quality is weakened by uncertainty about statistical validity and the generalizability of the results. Reliability is considered high as data is collected from reputable sources. Since data only provides an opportunity to shed light on historical aspects, sufficiency is not 100%.
3. Green Bonds
A green bond is a debt security. Like other bonds, it is issued to raise capital, but what makes it green is the fact that the capital raised is specifically used to support climate related or environmental projects. The specific use of the funds raised, the financing of specific projects, is what distinguishes green bonds from other types of bonds. This gives investors, in addition to evaluating the bonds standard financial characteristics, access to the specific
environmental purpose of the projects that they intend to support (World Bank, 2015). The purpose behind green bonds is for the proceeds to go to green assets. The green bond label can be applied to all debt formats. The bonds can be issued by both central and local
governments, banks as well as corporations (Climate Bonds Initiative, 2019).
The lack of a common green bond definition and framework is one of the biggest challenges with regards to the development of the global green bond market.
Without a common definition and framework, it is difficult to state what is considered
“green” and when a bond is truly green. Answering those questions is not an easy task. Many issuers say that they follow the Green Bond Principles (GBP), endorsed by the International Capital Market Association (ICMA) in 2014 (Pronina, 2019). The GBP are voluntary process guidelines that recommend transparency and disclosure. They promote integrity in the development of the green bond market by clarifying the approach for issuance (ICMA, 2018).
There are also companies that offer services to independently assess, verify or certify the green factor on the debt instrument. They include for example, rating companies such as Moody’s Investors Services, The Climate Bonds Initiative which will be addressed in detail later in this chapter, as well as specialized firms such as Viego Eiris, Sustainalytics and Cicero Shades of Green, a unit of the Norwegian climate research institute Cicero, to name a few (Pronina, 2019).
This chapter is organized as follows: First, the Green Bond Principles will be addressed, which are voluntary guidelines that clarify the issuance approach for green bonds, in chapter 3.1. Following with an introduction to the relatively new EU Green Bond Standards in chapter 3.2. Chapter 3.3 covers the green bonds history, beginning and how this specific type of debt instrument has evolved. Chapter 3.4 briefly covers the green bond market today.
After, in chapter 3.5 the main motivation behind both green bond issuance and investment is
addressed as well as criticised. Finally, chapter 3.6 addresses the Bloomberg Terminal green label process.
3.1 Green Bond Principles
GBP are voluntary process guidelines that were established in 2014 by a consortium of investment banks. Monitoring and development of the guidelines has since been moved to an independent department hosted by the ICMA (Climate Bonds Initiative, n.d.a). The guidelines recommend transparency and disclosure, and they promote integrity in the development of the green bond market. They do so by clarifying the issuance approach for green bonds, enabling all stakeholders to understand the characteristics of any given green bond. The GBP have four core components;
1. Use of proceeds
2. Process for project evaluation and selection 3. Management of proceeds
4. Reporting
(ICMA, 2018).
1. Use of proceeds; The cornerstone of a green bond is the utilization of the proceeds to green projects. The use of proceeds needs to be described in the legal
documentation for the security. The green projects should provide clear environmental benefits, which needs to be assessed and quantified by the issuer. The guidelines recognize several broad categories that fall under green projects. Eligible green bond categories are listed below:
• Renewable energy
• Pollution prevention and control
• Environmentally sustainable management of living natural resources and land use
• Terrestrial and aquatic biodiversity
• Clean transportation
• Sustainable water and wastewater management
• Climate change adaptation
• Eco-efficient and/or circular economy adapted products, production technologies and processes
• Green buildings
It should be noted that the listing is not final, and other projects can also be eligible. What they all have in common is that they contribute to environmental objectives such as climate change mitigation, climate change adaption, natural source conservation, biodiversity conservation, and pollution prevention and control (ICMA, 2018).
2. Process for project evaluation and selection; Another really important factor is the communication between the issuer and the investor regarding the process for the project evaluation and selection. Issuers should inform investors about the environmental sustainability objectives, the process by which the issuer determines how the projects fit within the categories, and the related eligibility criteria. In addition, Issuers are encouraged to disclose any green standards or certifications referenced in the project selection (ICMA, 2018).
3. Management of proceeds; The management of the proceeds, or an amount equal, should be credited to a sub-account, moved to a sub-portfolio or tracked by the issuer in any appropriate manner. In addition, it should be linked to the issuer´s lending and
investment operation for the specific project. It is recommended to use an auditor or another third party to verify the internal tracking method and the allocation of the funds (ICMA, 2018).
4. Reporting; The last core component is reporting. Issuers should keep available up to date information on the use of proceed and allocation. Its recommended to use qualitative performance indicators as well as quantitative performance measures (ICMA, 2018).
GBP recommends external review. There are a few ways to receive such outside input. Issuers can seek advice from both consultants and/or institutions with expertise in the field. The external reviews may vary in scope and address different processes regarding the green bond issuance. GBP groups them into the four types;
1. Second party opinion 2. Verification
3. Certification
4. Green bond scoring/rating
(ICMA, 2018).
1. Second party opinion; An independent institution with environmental
alignment with the GBP. For example, an assessment of the issuers objectives, strategy and policy. It can be assessment of the processes relating to environmental sustainability, as well as evaluation of the environmental features of the projects (ICMA, 2018).
2. Verification; Issuers can obtain an independent verification. The verification is usually set against a designated set of criteria. That is, connected to specific business
processes or environmental criteria. The verification can vary and focus on different aspects of the process, both external and internal (ICMA, 2018).
3. Certification; Issuers can have their green bonds, green bond framework or use of proceeds certified. The certification is set against a recognized external standard or label, focused on some specific criteria. Usually performed by qualified and accredited third party (ICMA, 2018).
4. Green bond scoring/rating; Finally, issuers can have their green bonds, green bond framework, use of proceeds, or other key features evaluated or assessed by a qualified third party. The third party is using an established scoring/rating methodology and can be a specialized research provider or a rating agency (ICMA, 2018).
It is encouraged that the external review providers disclose their credentials and relevant expertise, and communicate clearly the scope of the review (ICMA, 2018).
The principles do not create any rights or liabilities. They are adopted and implemented independently. Issuers are solely responsible for the issuance, underwriters are not responsible if issuers do not fulfil their commitments (ICMA, 2018).
These institutions, performing the external review, do not all work in the exact same manner, even though they follow GBP. This can be problematic when assessing green bonds. The European Union (EU) is creating a Green Bond Standard (GBS). The GBS will build on current market practices, such as GBP. This means that issuers worldwide will be able to cite compliance, if their plans are independently verified by an EU-Accredited assessor. The new standard will not be legally binding, it will be voluntary. The International Organization for Standardization is also preparing a Green Bond Standard, based on existing principles (Pronina, 2019).
3.2 Green Bond Standard
In 2018, the European Commission established a Technical Expert Group (TEG) on sustainable finance. The group was formed following the European Commission action plan of financing
sustainable growth, published earlier that year. A part of this action plan was the prepare a report on an EU GBS, building on current best practices (European Commisson, 2019).
In June 2019, the TEG published its report on the GBS. The proposed scope and objective is to create a voluntary, non-legislative GBS. The purpose behind the standards is to enhance the effectiveness, transparency, credibility and comparability of the green bond market. Additionally, market participants are encouraged to both issue and invest in green bonds. The GBS report builds on the interim report, a report following the action plan. More than 100 organizations provided feedback on the report. The feedback received was positive and a large majority of the respondents supported the standard creation. TEG studied the feedback from these organizations and created an improved version of the standards. They also published, in March 2020, based on recommendation, a usability guide for the GBS. The guide offers market participants guidance on the use of the proposed standards and the setup of a market-based registration scheme for external verifiers (European Commisson, 2019).
The European Commission is now exploring the possibility of a legislative initiative for the GBS in the context of the public consultation on the renewed sustainable finance strategy, taking place from March to May 2020 (European Commisson, 2019).
TEG’s main recommendations conclude a proposed scope and objectives, a proposed definition of an EU green bond and proposed core components regarding the standards. They recommend that a EU green bond can be any type of both listed and unlisted bond or capital market debt instrument. It can be issued by a European or international issuer aligned with the EU GBS. The proposed core components are built on best market practices. They consist of four critical elements;
1. Alignment with EU-taxonomy
2. Publication of a green bond framework 3. Mandatory reporting
4. Mandatory verification
(EU TEG on Sustainable Finance, 2019).
1. Alignment with EU-taxonomy; Proceeds from the EU green bond need to fulfil certain requirements. The proceeds should finance or refinance projects or activities that contribute significantly to at least one of the six taxonomy environmental objectives,
• Climate change mitigation
• Climate change adaptation
• Sustainable use and protection of water and marine resource
• Transition to a circular economy
• Waste prevention and control
• Protection of healthy ecosystems
Additionally, none of the other objectives should be significantly harmed. The proceeds should comply with the minimum social safeguards represented by the principles and rights set out in the eight fundamental conventions identified in the Internal Labour Organization.
The TEG wants to develop a technical screening criteria, that projects and activities should fulfil, allowing however for specific cases where these may not directly apply (EU TEG on Sustainable Finance, 2019).
2. Publication of a green bond framework; TEG proposes publication of the green bond framework, which confirms the voluntary alignment of green bonds issued with the GBS. It explains how the issuer’s strategy aligns with the environmental objectives, provides details on all key features regarding use of proceeds, processes as well as reporting of the green bond (EU TEG on Sustainable Finance, 2019).
3. Mandatory reporting; TEG proposes a mandatory reporting on use of proceeds, an allocation report, and on environmental impact, an impact report. The reports can be in a project-by-project level or on a portfolio level, and should be published on the issuer’s website or other communication channels (EU TEG on Sustainable Finance, 2019).
4. Mandatory verification; TEG proposes as well, a mandatory verification of the green bond framework and final allocation report by an external reviewer. TEG recommends the use of external verifiers that are formally accredited and supervised. They believe that the most suitable European authority to oversee that process would be the European Securities and Markets Authority (ESMA) (EU TEG on Sustainable Finance, 2019).
The flowchart below illustrates the interactions between the components throughout the issuance process and until full proceeds allocation.
Figure 1. Interactions between the four components of the EU GBS. (EU TEG on Sustainable Finance, 2019).
The report also includes recommendations, suggesting how the European Commission, European Union Member State governments and other market participants can support the uptake of the GBS (EU TEG on Sustainable Finance, 2019).
They believe that the standard will solve several barriers in the current market.
It will do so by reducing the uncertainty on what is green, by linking it with taxonomy.
Furthermore, standardizing both the verification and reporting processes, and having an official standard, may motivate stakeholders to both issue and invest in green bonds (EU TEG on Sustainable Finance, 2019).
The main difference between the Green Bond Principles (GBP) and the Green Bond Standard (GBS) is that the principles do not provide details on what is considered
“green”. They leave the green definitions to the issuer to determine. The GBS are much more detailed, prepared by experts in the field and in cooperation with over 100 institutions, resulting in an improved version of the standards. As previously mentioned, the European Commission will explore the possibility of a legislative initiative for the standards. If that goes through it will have a major impact on the green bond market, but it is difficult to predict the future and only time will tell how these matters evolve. Either way, the GBS will hopefully
have a positive impact on the green bond market, since one of the biggest challenges the market is facing is the lack of common definition and framework, with regards to market development. Even facing this difficult challenge, green bonds have been on the rise in recent years. Before addressing the state of the green bond market and its growth, it is important to look into how this all started.
3.3 Beginning and History
It is undeniable that focus on climate change is increasing. People are to a greater extend starting to believe that humanity has had a major impact on the Earth’s climate with increased emission of carbon dioxide. Evidence of this evolution can be seen in the Earth’s rising temperature (NASA, n.d.).
This issue has been addressed in different ways in the recent years. In 2015, drastic measures were taken, when the United Nations signed a binding agreement to combat climate change and to accelerate and intensify actions and investments needed for a sustainable low carbon future. The Paris Agreement aims is to strengthen the global response to the threat of climate change. The goal is to keep the global temperature rise to a
maximum of 2 degrees Celsius above pre-industrial levels and pursue to limit the increase even further to a maximum 1.5 degrees Celsius. In addition, the agreement also aims to increase countries abilities to deal with the impacts of climate change and make finance flows consistent with a low greenhouse gas emission and climate-resilient pathway (United Nations. n.d.). This action was taken to combat global warming and has a major effect on the green bond market, but the term green bond had already seen the light of day by then.
The term “green bond” first appeared in 2007-2008 when the World Bank issued a green themed bond for the first time. The request for this kind of debt instrument came from a group of Swedish pension funds, following a report from the United Nations providing scientific data on climate change. The pension funds wanted to find a way to invest money in a way where they could have a positive climate impact. The pension funds
contacted the World Bank to see what could be done. Soon, the first so-called green bond was issued. That bond created the foundation for today’s green bond market. It defined criteria for projects eligible for green bond support, included a second opinion provider and had impact reporting as a part of the process (World Bank, 2019). However, the wider bond market did not react until the year 2013. In March that year, the first USD1bn green bond
was sold within an hour of issue by IFC. That moment seemed to be a turning point, as different types of issuers started to issue green bonds. Since then the momentum has continued, with now over USD500bn bonds currently outstanding (Climate Bonds Initiative, n.d.b).
3.4 Green Bond Market Today
The main statistical data available on the issue and status of the green bond market today comes from the Climate Bonds Initiative (CBI) organization, previously
mentioned. CBI is an international not-for-profit organization working solely to mobilize the bond market for climate change solutions. They promote investments in projects and assets essential for a speedy transaction to a low carbon and climate resilient economy. Their strategy is to develop a large and liquid green and climate bonds market. The aim is to drive down the cost of capital for climate projects, grow aggregation mechanisms for different sectors, and support governments who seek to use the bond market. Their work is an open source public good and falls into three work streams. The first is market intelligence, which consists of reporting on the green bond market evolution, sizing the climate bonds universe and demonstrating green infrastructure pipelines. The second work stream is developing a trusted standard, and the third is providing policy models and advice (Climate Bonds Initiative, n.d.c).
CBI issues yearly, a state of the market report for green bonds. The report focuses mainly on what they consider labelled green bonds. It uncovers the latest
developments and trends in the green bond market. According to CBI’s latest report, the green bond market has been growing intensely in the recent years, as previously mentioned (Climate Bonds Initiative, 2019). CBI estimates the total green bond issuance in 2020 to reach USD350bn, as presented in figure 3.
Figure 2. Total green bond issuance in USD equivalent (billions). *Estimated issue amount in 2020. (Climate Bonds Initiative, n.d. d).
On March 9, 2020, the issuance had already reached USD29.5bn, USD14.7bn in January and 14.8bn in February (Climate Bonds Initiative, n.d.d).
A new global record was made in 2019 when the global green bond and green loan issuance reached UDS257.7bn. It should be noted, that green loans, where at least 95%
of the proceeds must be dedicated to green assets or projects aligned with the CBI, only accounted for 4% of the issuance – or USD10bn. 1788 green bonds were issued from 496 issuers, of which 250 are new issuers accounting for USD67.8bn. The green bond issuers come from a total of 51 jurisdictions, of which 8 are new. The volume in 2019 was mainly driven by the European market, which accounted for 45% of the global issuance. The Asia- Pacific market accounted for 25% and the North American accounted for 23%. The total amount of green bonds issued in Europe increased by 74% or USD49.5bn year-on-year in 2019, reaching a total of 116.7bn. USA, China and France continue as the top-ranked countries, accounting for 44% of global issuance in 2019. US issuers accounted for
USD51.3bn, the Chinese accounted for USD31.3bn and the French for USD30.1bn. The top three issuers were Fannie Mae, Kfw and the Dutch State Treasury Agency (DSTA). CBI uses the Climate Bond Taxonomy, featuring eight different sectors. Energy and buildings are the largest sectors with regards to use of proceeds. In 2019, both sectors had similar share of the
0.8 0.4 0.9 4.3 1.3 3.5 11
37 45
86
159 168 258
350
0 50 100 150 200 250 300 350
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020*
USD equivalent (billions)
Years
Total issuance
market. Energy accounted for 31% and buildings accounted for 30%. Figure 4. shows the use of proceeds allocation for 2019 (Climate Bonds Initiative, 2020).
The green bond market is still quite young and a relatively small percentage of the total bond issuance. Undeniably, the market has been growing intensively in the recent years, and thus, it is interesting to look at what has been motivating this growth.
3.5 Motivation and Criticism
The main motivation behind green bonds, and other likewise debt instruments, is probably to reach the climate, environmental and social sustainability goals that have been set forward (European Union, 2019). As previously mentioned, humanity has had a major impact on the Earth’s climate which needs to be addressed. The Paris Agreement was a step in that direction, binding nations to combat climate change, accelerate and intensify actions and investments needed for a sustainable low carbon emission future. These factors naturally influence the green bond market. Green bonds can be used as a tool in the mission to keep global warming under two degrees. Green bonds can accelerate the adoption of new innovative technologies, finance green projects and projects that provide green jobs (Go Green Bonds, n.d.).
Investors have been increasingly demanding socially responsible investments (SRI) in recent years. Retail investors have been demanding sustainable investments from their brokers and their fund managers while institutional investors have been demanding
31%
30%
20%
9%
4% 3%
2019 USD 257.7 bn
Energy 31%
Buildings 30%
Transport 20%
Water 9%
Waste 4%
Land Use 3%
Industry 1%
ICT 1%
Unalloc. A&R 1%
Figure 3. Use of proceeds allocation. (Climate Bonds Initiative, 2020).
green bonds to address environmental, social and governance (EGS) mandates. Before green bonds, it had been difficult to address EGS mandates with fixed income tools. It is possible to say that green bonds have attracted new investors as well as new types of investors (Go Green Bonds, n.d.). It still depends on the investors and what motivates them. Are they driven by the belief that this kind of investing can truly promote sustainability? Are they mainly driven by profits, or can it be a combination of both? This topic falls under behavioural finance, which is the study of the influence of psychology on the behaviour of the investors. It focuses on the fact that investors are not always rational, that they have limits to their self- control, and are influenced by biases (CFI, n.d.).
Green bonds can enable projects at a lower cost of capital, to support environmental investments that may not otherwise be an option, or investments where it may be uneconomic using expensive capital. They are well suited for large-scale sustainability projects, which often require capital investment ahead of revenues and generate fair
revenue over a longer investment period. Green bonds issuers today are paving the way for the rest of the nation in financing green projects. They can take part in shaping the market and developing standards, as well as encourage others to participate. Issuers can also use green bonds to brand themselves, as the press has been covering green bonds favourably (Go Green Bonds, n.d.).
Green bonds can provide tax incentives for issuers and investors, and there are several types of tax incentives policy makers can put in place. The incentives for green bonds, can be divided into three main types. That is, tax credit bonds, direct subsidy bonds and tax- exempt bonds. With tax credit bonds, issuers do not have to pay interest on their green bond issuance. Bond investors receive tax credits instead of interest payments. With direct subsidy bonds, issuers receive cash discounts from the government to fund their net interest
payments. Tax-exempt bonds, exempt investors from paying income tax on interest from their green bonds (Climate Bonds Initiative, n.d.e).
One of the biggest challenges with regards to green bonds is the lack of a common definition and framework. Therefore, investors have been facing the challenge of judging what is green and what is not. Regulators are working on standards to help with regards to this problem, and they help guarding against green washing. Green washing refers to misleading claims about how environmental friendly the issuer is. This is a problem that is currently being addressed but has not yet been fixed. Investors, that invest in green bonds,
usually want to know exactly where their money is going. This leads to a cost of reporting for the issuers (Pronina & Freke, 2019).
There are also other issues at hand with green bonds. Even though the market is on a rise, it is still relatively small and therefore there is a lack of liquidity in the market.
Usually, investors that buy green bonds might need to hold it until maturity. Other risks might include low yields, mispricing, and a lack of sufficient research available to make an educated investment decision (Colonial First State, n.d).
With most financial instruments, there are pros and cons to both issuance and investing. There is a lack of sufficient research available regarding different aspects of green bonds, but recently, following the green bond boom, researchers seem to be more interested in studying them. Information regarding green bonds can be found in several different data bases, where the bonds have a specific green label separating them from other bonds.
3.6 Bloomberg’s Green Label
Identifying green bonds can be a real challenge as there is no common green bond definition and framework yet. There are many different shades of green bonds, in addition to different types of issuers and debt instruments (Bloomberg, 2015).
The Bloomberg terminal is the most powerful and flexible platform for financial professionals. The terminal contains of real-time data, news, detailed research, analytics, communication tools and execution capabilities. The terminal delivers real-time coverage of markets and securities with information across asset classes. Therefore, it contains
information about the green bond market (Bloomberg, n.d).
Bloomberg aims to improve transparency in the green bond market, and they do so by labelling the bonds. Their green bond working definition is adopted from the GBP and other organizations’ pioneering both issuance and analysis. According to their
methodology, while the use of proceeds can vary, all issuers must commit to deploying 100%
of the bonds proceeds for environmental sustainability-orientated activities if they want to have their bond labelled green on the terminal. Such activities are outlined briefly in table 3.
Table 2. Acceptable use of proceeds types. (Bloomberg, 2015).
Use of proceeds type Example
Renewable energy Solar, wind, hydro, geothermal, biomass/biofuel/bioenergy, waste-to-energy, tidal and other renewables
Energy smart technologies and energy efficiency
Electric and hybrid vehicles, fuel cells, energy storage, digital energy and energy efficiency (buildings, industry, LED and smart lighting)
Green buildings and infrastructure New LEED, Energy Star, BREEAM or other certified building and energy efficient social infrastructure
Agriculture and forestry Forestry management, reforestation, afforestation and land- use
Other sustainability Other climate change adaptation/mitigation, including waste management, clean water, pollution control, etc.
The labelling in the terminal is in the “use of proceeds” field, labelled as green bonds. It should be noted that Bloomberg observes specific exclusions from the categories listed in the table above, using the tag “use of proceeds including - but not limited to“, those involving coal and nuclear (Bloomberg, 2015).
Bloomberg labels a bond green when issuers self-labels a bond as green or identifies it as an environmental sustainability-orientated bond. In addition, the issuer needs to turn on a clear statement about the company’s commitment to deploy the funds towards the right projects. A review of a green bond issue takes place upon the announcement or issuance of the bond, depending on information availability. All green bonds require a term sheet or prospectus containing a “use of proceeds” disclosure aligning with the accepted categories listed in table 2. Issuers can also include optional documentation about the use of proceeds. The optional documentation is recognized by Bloomberg as supplementary
information. A side letter, syndicate email, official statement from issuer, and published review of issuer’s bond or bond framework by an external party are considered as
supplementary information. Bloomberg recognizes these documents but they are insufficient for labelling (Bloomberg, 2015).
Bloomberg recognizes that the green bond market is an early stage market, which is still evolving. Therefore, they need to evolve and update their work on as-needed basis (Bloomberg, 2015).
4. Measures for comparing
This chapter addresses the measures for comparing, previously mentioned in the problem statement. These measures will be used later in this thesis, were we attempt to analyse the differences between green bonds and their conventional peers, to determine the effect of the green label. We will address the measurements in detail and their relevance with regards to this thesis.
This chapter is organized as follows: First the yield to maturity is explained in chapter 4.1. Following with the credit spread in chapter 4.2, and the liquidity in chapter 4.3.
4.1 Yield to Maturity
The yield is a commonly used measurement when comparing different types of bonds, as can be seen in the literature review later this thesis. Yield to maturity (YTM) is the internal rate of return (IRR) of an investment in a bond or the true underlying interest rate of return for the investor. We can address if there is a lower yield for investors financing green projects. That is, if green bonds are associated with a lower rate of return. To evaluate the difference between green bonds and conventional bonds, the yield to maturity is an important factor because it consists of both the purchase price of the bond and the coupon rate. It is the market convention for return on interest rate investment, and can be used to compare investments independently of their time to maturity and other factors influencing the current price of the bond (Berk & Demarzo, 2017).
In the Bloomberg Terminal, the following yield to maturity types are available:
• Mid yield to maturity
• Bid yield to maturity
• Ask yield to maturity
In this thesis, we have chosen to use the mid yield to maturity when evaluating the differences between green and their conventional peers. The bid yield to maturity focuses on investors, and the ask focuses on issuers at issuance and sellers on the secondary market. Therefore, we find it most relevant to use the mid yield to maturity, which solves for the mid-price when valuing the security to maturity (Bloomberg terminal, 2020). Another reason behind our decision to analyse the yield to maturity, is since investors typically buy green bonds for impact investment reason, they tend to hold them until maturity. Thus, the secondary green bond market is less liquid than the secondary market for conventional
bonds. However, green bonds are still traded, and as the green bond market grows, so does the secondary market (World Bank, 2017).
It should be noted that additionally we wanted to analyse the yield at issuance as well but were unable to, due to data limitation and COVID19, which will be covered in chapter 6. The measurement is relevant since it can show if investors are willing to pay a premium for a bond with a green label.
4.2 Credit spread
A yield curve can be plotted for a risk-free Treasury security and for corporate bonds as shown in figure 4.
Figure 4. Yield curves. (Berk & Demarzo, 2017).
The difference between the corporate yield curve and the treasury yield curve is known as the default spread or credit spread. The magnitude of the credit spread will depend on investors estimation of the probability that a firm will default. The credit spread is high for bonds with low credit ratings, which are more likely to default (Berk & Demarzo, 2017). The credit spread can be used as a measure of risk related to a certain bond because it allows for a comparison between a bond and a risk-free alternative.
In our analysis, we capture this risk by using the Interpolated spread (I-spread) from the Bloomberg Terminal as a proxy. The I-spread is characterized as the difference between the yield on a bond and the swap rate. A bond that has a higher credit risk will also
0 1 2 3 4 5
0 5 10 15 20 25 30 35
U.S. Treasury Yield Curve U.S. Industrials (AAA) U.S. Industrials (A)
Yield to Maturity (%)
Time to Maturity (Years)
be accosted with a higher I-spread (Forjan, 2020). The I-spread is used as a proxy for the credit spread related to corporate bonds.
It should be noted, that additionally we also wanted to use the G-spread from the Bloomberg Terminal as a proxy for the credit spread related to green government bonds.
Which would be appropriate since the government bonds are usually not as risky as corporate bonds. The G-spread (nominal spread) is the difference between the yield of a bond and a Treasury bond yield with the same maturity (Forjan, 2020). We were unable to do so, due to data limitation which will be covered in chapter 6.
4.3 Liquidity
Bonds liquidity refers to the bonds ability to sell without having to reduce its price significantly. Bachelet et.al. 2018 computes the liquidity as:
𝑝𝑟𝑖𝑐𝑒&'( – 𝑝𝑟𝑖𝑐𝑒+,-
As can be seen from the formula this is the difference between the highest price a buyer in the market is willing to pay, and the lowest price a seller in the market is willing to sell for.
The higher this value is, the higher the bonds liquidity is assumed to be (Bachelet et al., 2018). Wulandari, et.al. 2018, uses this formula to capture the liquidity. They also use LOT liquidity measure. The LOT liquidity captures both market impact costs, commission costs and opportunity costs and is calculated as the difference between the percent buying cost and the percent selling cost (Wulandari et al., 2018).
In our analysis, we capture the liquidity by calculating the difference between the bid price and the ask price. We would have preferred to use both measures, but we are limited with regards to data. That is, we cannot get the percent buying cost and the percent selling cost. We are analysing how easily the bonds can be sold on the secondary market. The bid/ask spread does not take into account market depth, but it is still the best available proxy to capture liquidity. As previously mentioned, investors tend to hold them until maturity and we want to put that statement to the test. Additionally, if the bond cannot easily be sold at a fair market value, how does that relate to the yield with regards to liquidity premium.
5. Literature Review
There is a lack of sufficient research available regarding different aspects of green bonds, as previously mentioned. However, researchers seem to draw more interest in the topic in recent years. This chapters contains an overview of what some of the previous studies have shown, comparing green and conventional bonds.
Many studies have attempted to capture if green bonds are issued at a premium or discount compared to non-green bonds.
Gianfrate and Peri (2019) investigated whether green bonds were priced at a premium, compared to their conventional peers. That is, compared to bonds that shared similar characteristics except for the green label. They defined the premium as the difference in the spread at issuance between a bond with the green label and a peer without. The goal was to explore the convenience of issuing green bonds, how the market prices them and if issuers can achieve lower yields by issuing them, resulting in a lower cost of capital. The method used was the propensity score matching, the same method used in this thesis. They analysed Euro denominated green bonds issued between 2013-2017. They show that there exists a green bond premium as the green bonds are issued with lower yields. On average, they find that the yield of green bonds is 18 basis points lower than for conventional bonds Additionally, they showed that there was also a negative premium on the secondary market (Gianfrate & Peri, 2019).
Baker et.al. (2018) research was in line with Gianfrate and Peri (2019), showing a green bond premium. They studied the U.S. corporate and municipal green bond market.
They analysed the pricing and ownership pattern of green bonds through regression. Looking at green municipal bonds issued between 2010 and 2016 as well as green corporate bonds issued between 2014 and 2016. They confirm that green municipal bonds are issued at a premium, with lower yields by several basis points. Additionally, the study showed that some investors were holding more green bonds that would be considered optimal. Thus, investors seem to be accepting less returns in order to hold the green bonds (Baker et al., 2018).
Zerbib (2017) uses matching method to analyse the difference in yield between a green bond and a comparable synthetic conventional bond, by analysing investment grade senior bullet fixed rate bonds issued worldwide. Zerbib’s research, like the previously
