Impact of COP26 and COP27 Events on Investor Attention and Investor Yield to Green Bonds

Abstract

Green bonds are a relatively new financial product that offers investors a variety of alternatives. However, many individuals continue to be suspicious about its long-term returns and risks. To clarify this issue, this study employed two global environment events—COP26 and COP27—to influence investor attention and investor yield of green bonds and conventional bonds. The data are collected from 15,188 bonds, including 779 green bonds and 14,409 conventional bonds issued from 2021 to 2023 worldwide. The event study method has been conducted with pre- and post-event data to estimate the impact of green bond issuance before and after COP26 and COP27 on investor returns, as well as the impact of investor attention on investment returns. The research results show that investors should buy shares of companies that issue green bonds after major environmental events to benefit from the higher CAR of these companies. Investors can also use the S&P 1200 index as a measure to assess risk and abnormal returns when making short-term investments in shares of organizations that issue green bonds.

1. Introduction

Green bonds are one type of bond labeled as “green” requiring confirmation from a third party (the Climate Bonds Initiative and International Capital Market Association). These bonds are very crucial to call capital for green projects. They have quickly proven to be an effective financial tool for mobilizing capital for these green projects in the mid-2000s [1]. For green bonds to develop, it is necessary to demonstrate that they meet the expectations of both investors and issuers in terms of providing capital for environmentally friendly projects [2]. According to Bloomberg statistics, in the first half of 2023 alone, 935 green bonds were issued, raising USD 351 billion [3]. For the green bond market to thrive, it is critical to ensure that these instruments meet the expectations of both investors and issuers in terms of delivering environmental benefits and financial returns.

Event study is an econometric research method that involves the collection, analysis, and interpretation of data to assess the impact of a specific event or a series of specific events on an economic variable [4,5]. This research method uses pre- and post-event data to estimate the impact of the event [6]. Moreover, the event’s impact can occur over a period of time after the event date, so this method is often conducted not only at a single point in time (the event date) but within an “event window”—a period of time before, during, and after the event [5,6]. Depending on the nature of the events, researchers will choose different window lengths and may apply the division of the chosen event window into shorter event windows for observation and calculation.

COP26 and COP27 are two big events that we should pay attention to when it comes to green bonds. In the COP26 Conference taking place from 31 October to 12 November 2021 in Glasgow, UK, Vietnam and more than 100 countries signed the Glasgow Agreement, committing to bring net carbon emissions to zero by 2050 (net zero emissions). The Glasgow Agreement emphasizes the need to mobilize climate finance from all sources to achieve levels needed to realize the goals of the Paris Agreement, including significantly increasing support for developing countries and urging developed countries to urgently complete the committed target of USD 100 billion as well as the target by 2025, emphasizing the importance of transparency in implementing these commitments [7]. COP27, held in Sharm El-Sheikh, Egypt, from 6 to 18 November 2022, is a process following COP26. The COP27 conference agreed on an overarching decision called the “Sharm el-Sheikh Implementation Plan”, emphasizing that the global transition to a low-carbon economy is expected to require investments of at least USD 4000–6000 billion per year [8]. Delivering these funds will require a rapid and comprehensive transformation of the financial system, its structures, and processes involving governments, central banks, commercial banks, institutional investments, and other financial corporations.

Both COP26 and COP27 are two major environmental events organized by the United Nations that have produced significant environmental decisions. For green bonds, these events are significant in attracting funding for green projects. While event studies have been conducted on both events in different areas, none have focused on green bonds. An event study on global green bonds would be more persuasive to investors.

The influence on investor behavior and bond pricing from COP26 and COP27 can be translated through several interconnected mechanisms, such as policy announcements, market expectations, and investor sentiment. Policy announcements at events like COP26 and COP27, including commitments to net-zero emissions and significant climate finance targets (e.g., USD 100 billion annually), enhance the appeal of sustainable finance. These pledges boost demand for green bonds as investors align portfolios with sustainability goals, leading to lower yield spreads and reduced perceived risks compared to conventional bonds. COP events significantly heightened investor awareness of environmental issues, as reflected in the increased interest in terms like “green bonds” during these periods. This heightened attention, combined with ESG mandates, drives investors to prioritize green bonds for their perceived dual benefits of financial returns and alignment with sustainability goals. In addition, COP events emphasize stricter regulations for carbon-intensive industries, increasing perceived risks for conventional bonds while making green bonds more attractive due to lower regulatory and reputational risks. These events may also trigger short-term market volatility, prompting investors to diversify into green bonds as a hedge against long-term environmental and policy risks.

This paper fills a critical gap in the literature by conducting an event study to assess the impact of COP26 and COP27 on investor attention and yields in the global green bond market. By analyzing how these high-profile environmental events influence investor behavior, this study provides insights into the effectiveness of global climate conferences in mobilizing financial resources for green initiatives. Specifically, this research explores whether these events enhance investor confidence, attract greater capital flows, and influence yield dynamics, thereby making green bonds a more attractive investment option. The findings of this paper are intended to inform policymakers, issuers, and investors about the interplay between global climate negotiations and green financial markets, ultimately contributing to the development of more robust strategies for financing the green transition.

2. Literature Review and Hypothesis Development

2.1. Investor Yield

Investor returns can be separated into two components: yield differential and cumulative abnormal return. The yield differential, for instance, benefits bond investors, whereas the cumulative anomalous profit benefits stockholders.

Yield spread represents the difference in yield between a risky asset, such as a corporate bond, and a benchmark risk-free asset, typically a government bond or treasury bill. It serves as compensation for the additional risk undertaken by investors when holding riskier assets compared to the perceived safety of risk-free instruments.

Cumulative abnormal return (CAR) is a crucial metric employed in finance to evaluate the impact of specific events on the performance of a security or portfolio. It represents the aggregate sum of abnormal returns (AR) calculated over a selected time window. CAR calculations are particularly effective within a short timeframe, making event studies a popular approach for CAR analysis. These studies focus on a specific event and its impact over a defined period [9]. CAR also plays a significant role in determining yield spreads, which compare the yields of various investment assets (bonds, stocks, mutual funds, etc.). CAR-based studies typically utilize event study methodologies to assess the direction of impact, market response, and the overall influence of an event on a particular security.

2.2. Green Bond Issuance

The term “greenium” is frequently used in the context of green bonds. Greenium, also known as green premium, denotes that investors are willing to spend more on bonds or invest in bonds with lower interest rates in order to have an environmental impact [10]. This indicates that green bonds will have lower yield spreads than standard bonds. For different bond issuers, financial factors such as reputation, brand, search for legitimacy, “social license to operate”, and environmental considerations play a significant role in assessing yield spread of green bond [11]. The institution in which the issuer works constitutes a significant trade-off between the two types of bonds (green and conventional). In contrast, the capacity to issue green bonds depends on the project’s financial prognosis. The level of awareness and competitiveness of the product among investors and consumers influences the issuer’s decision to issue green or conventional bonds [12,13].

In the primary market, several prior studies have found a negative yield differential between green bonds and conventional bonds, ranging from −17 to −29 basis points (bps) [14,15,16]. Other studies have found support for the “greenium” view in the corporate bond segment, estimating a range of −24 bps to −6 bps [14,17]. Conversely, some other studies have found that investors exhibit different behavior when choosing between green bonds and conventional bonds from the municipal sector in the US during the period from June 2013 to July 2018 [18]. In China, green bonds issued by listed companies are traded at a significant discount to conventional bonds (−33 bps) [19], in contrast to the slightly positive “greenium” (0.16 bps) from the non-financial sector in developing countries [15]. Despite being modest, the average green bond premium is substantially negative and equal to −2 basis points for the whole sample, which leads to the financial bonds and bonds with low investment grades having higher negative premiums [13]. In general, Chinese green bonds have a worse rollover than conventional bonds; in particular, the estimated liquidity impact on Chinese green bonds is worse, with an average premium of 28.14 bps, which is higher than the value of the corresponding conventional bond, which is only roughly 19.40 bps [20]

In the global secondary bond market, most studies support the existence of a negative “greenium”, ranging from −14 bps to −1 bps [14,16,21]. However, there are some studies that contradict this view such as Bachelet et al. (2019), who found a positive “greenium” ranging from 2 bps to 6 bps [22]. In the secondary market, the estimated “greenium” for corporate bonds is −63 bps, four times higher than for the issuer groups [23]. This highlights the inconsistency among the studies. Many argue that green bonds are no different from conventional bonds except for the green label, as the return on green bonds is still based on the performance of the issuing company. However, there is a general trend in the research that green bonds have lower yields than conventional bonds [24,25].

The cumulative abnormal return (CAR) of three types of bonds (first-time green bonds, subsequent green bonds, and conventional bonds) would have different reactions in the periods surrounding the issue date [26]. For first-time green bonds, the CAR moved positively for 4 days after the issue date, and then it began to decrease gradually. This positive result was also explained by Tang & Zhang (2020) using the term “green label”, in which investor confidence in green bonds tends to increase strongly due to the “green” factor, especially in the context of environment and climate being the top concerns in society today [17]. It was concluded that green bond issuance has a significantly positive impact on stock prices [27]. Although, at the time of issue, the returns of all stocks were negative for all samples studied, the cumulative abnormal return (CAR) over the next 10 days increased and was no longer negative. A study on the Chinese market over one year (May 2019–May 2020) using event windows of [−10; 10] and [−5; 5] concluded that the COVID-19 pandemic had a significant impact on the Chinese green bond market, significantly increasing the CAR and stock price compared to the pre-pandemic period [28]. However, after the pandemic was controlled, the CAR decreased significantly. Another study analyzed the effect of traditional bond issuance and showed that CAR moved negatively, although not much [29]. The regression model of the study by Jian et al. (2022) included company variables such as ROA, tangible assets, Debt-to-Assets ratio, and company size [30].

Based on previous studies, the hypotheses are proposed as follows:

Hypothesis 1a. 

Green bond issuance has a positive impact on the yield spread of businesses issuing green bonds.

Hypothesis 1b. 

Green bond issuance has a positive impact on CAR of businesses issuing green bonds.

2.3. COP 26 and COP27

Kahneman’s attention theory (1973) [31] posits that attention is a finite resource, leading to selective attention as a result of the overwhelming amount of information and limited processing capacity. This theory has been applied to explain various phenomena in finance. For instance, investors tend to focus on salient information over ambiguous information [32] due to the need for selective and efficient attention [31]. With a vast array of securities to choose from, investors allocate their attention to a subset of securities and actively seek deeper information about them. For the financial industry, news and events have a huge impact on the market.

COP26 and COP27 are two big events that we should pay attention to when it comes to green bonds. In the COP26 Conference taking place from 31 October to 12 November 2021 in Glasgow, UK, Vietnam and more than 100 countries signed the Glasgow Agreement, committing to bring net carbon emissions to zero by 2050 (net zero emissions). The Glasgow Agreement emphasizes the need to mobilize climate finance from all sources to achieve levels needed to realize the goals of the Paris Agreement, including significantly increasing support for developing countries and urging developed countries to urgently complete the committed target of 100 billion USD as well as the target by 2025, emphasizing the importance of transparency in implementing these commitments [7]. COP27, held in Sharm El-Sheikh, Egypt, from 6 to 18 November 2022, is a process following COP26. The COP27 conference agreed on an overarching decision called the “Sharm el-Sheikh Implementation Plan”, emphasizing that the global transition to a low-carbon economy is expected to require investments of at least 4000–6000 billion USD per year [8]. Delivering these funds will require a rapid and comprehensive transformation of the financial system, its structures, and processes involving governments, central banks, commercial banks, institutional investments, and other financial corporations. The government must create a more effective and efficient green funding model and give green initiatives more importance during the evaluation process [33]. Green financing can support the attainment of environmental goals under the COP-26 targets in a variety of ways. For example, investments in clean technologies brought about by the expansion of green financing can both create new jobs and lower emissions from energy production [34]. From a COP-27 perspective, promoting green innovation, efficient green finance, and growing finance can help minimize climatic damage to a sustainable environment [35].

COP26 and COP27 also are two events that have a strong impact on green finance. There has not been much research on green bonds related to these events. Therefore, a study is needed to evaluate the impact of these events on green bond issuance.

Based on previous studies, the research team proposed the following hypotheses:

Hypothesis 2. 

COP26 and COP27 events significantly increased investor attention toward green bonds, as evidenced by heightened search activity and trading volume.

Hypothesis 3a. 

The COP26 and COP27 events led to a reduction in the yield spread for businesses issuing green bonds compared to conventional bonds.

Hypothesis 3b. 

COP26 and COP27 events positively influence the cumulative abnormal returns (CAR) of businesses issuing green bonds within the event window.

2.4. Investor Attention

Existing shareholders benefit from green bond issuance when institutional ownership increases and stock liquidity improves after green bond issuance, which also helps to broaden the investor base because green bond issuance can attract more media attention and be used by impact investors to meet their investment mandates [17]. Some previous studies have found a positive relationship between green bonds and other major asset classes [13,36]. Additionally, Nayak [37] finds that investor sentiment is a significant factor in the determination of corporate bond yield spreads. High-yield bonds are more susceptible to mispricing due to market sentiment, and conversely, low-yield bonds are less sensitive to sentiment. The multidimensional idea of investor attention has a significant impact on market performance and financial decisions. Google searches are a popular way to gauge the level of interest of individual investors, who often use this tool to find information because of its accessibility. The Google Search Volume Index (GSVI) provides a more direct measure of investor attention [38].

The group of investors is emotional, impulsive, and has a significant reaction to short-term events [38,39]. In contrast, institutional investors are primarily attracted to media coverage and often use it to shape more complex, long-term strategies. The media plays an important role in helping institutional investors analyze the market, especially when it comes to initial public offerings [39]. These organizations often use information from the media to identify long-term opportunities and reduce uncertainty. While institutional investors tend to closely monitor and conduct in-depth analyses of media content, individual investors are more susceptible to short-term events [40].

It is more comparable to media coverage of green bond research to use Google Trends to gauge investor interest, particularly for events like COP26 and COP27. By looking at Google search data, it is possible to gain insight into the short-term considerations that drive private investors, whereas media coverage is often more advisor-oriented and fast-paced, requiring time to write and circulate online stories. Furthermore, the media has difficulty conducting an in-depth analysis of changes in annotations at specific points in time (event windows) or in real-time. The search volume for “green bonds” increases significantly after COP events, indicating the level of investor interest.

A similar study was conducted using Baidu’s search index by Yang el al. (2021) specifically for the Chinese market [41]. Moreover, it was founded that the spillover between green bonds and investor attention is limited to the median quantile and becomes stronger at the lower and upper quantiles [42]. Green bond pricing implies that non-economic factors like environmental preferences should be taken into account by future bond pricing [43]. Since there are feedback effects between green bonds and investor attention, investors interested in green bonds can use market attention as a useful tool to predict the performance of these bonds [44].

From previous studies, the research on the impact of investor attention on the investor yield of green bond issuers is still open. There are not many studies that delve into this issue, while theoretically, this is an important factor affecting the investment decision and return of green bonds. Therefore, the authors proposed the following hypothesis:

Hypothesis 4a. 

Increased investor attention, as measured by search volume indices, positively impacts the cumulative abnormal returns (CAR) of businesses issuing green bonds.

Hypothesis 4b. 

Higher investor attention reduces the yield spread of green bonds by increasing demand and lowering perceived risks.

2.5. Market Risk

There are two main types of risk that investors face when participating in the financial market. Based on the level of impact, risk is divided into two categories: market risk and specific risk. Market risk, also known as systematic risk, is a risk factor that can affect a company’s profits and is caused by changes in the market [45]. Since market risk affects all investments, it cannot be avoided. This is in stark contrast to specific risk, which occurs only in a specific company or sector and can be mitigated by diversifying the investment portfolio. Some of the most common systematic risk factors in the financial market include economic recessions, political instability, exchange rate fluctuations, and natural disasters. Compared with conventional bonds, green bonds might be more useful for hedging against stock market risks [46]. Furthermore, of the four main market indexes, it was discovered that only the green bond index is the most sensible and efficient hedge for carbon market risk [47]. Thus, green bonds have certain green label risks, such as the risk of green projects, the issuer’s profile [48], and the unpredictability of new green technologies [24], which investors can utilize to make well-informed investment choices and choose asset classes based on their tolerance and appetite for risk.

The research of Collin-Dufresn et al. (2021) emphasizes that market risk will increase the credit spread of corporate bonds [49]. It was pointed out the importance of measuring market risk in credit spread models and concluded that companies operating in high-market-risk environments would face a higher risk [50]. The breakthrough came from demonstrating the significant impact of market risk on green bonds using 82 green bond issues during the period of 2016–2017 [23]. The common point of the above studies is that they all use indicators and data that are representative of the entire market, the most popular of which are the S&P index and data from large companies around the world. Therefore, the authors proposed the following hypothesis:

Hypothesis 5a. 

Market risk affects the yield spread of businesses issuing green bonds.

Hypothesis 5b. 

Market risk affects the CAR of businesses issuing green bonds.

3. Data and Methodology

3.1. Research Data

Sample data for this analysis comprises 15,188 bonds issued between 2021 and 2023, including (i) 779 green bonds, certified as “green” by the Climate Bonds Initiative or similar entities and (ii) 14,409 conventional bonds with no green certification, issued by the same companies within the time span of 2021–2023. Sources of dataset include:

• Bond characteristics and financial data from Refinitiv Eikon, providing detailed information on coupon rates, issue size, and yields, and capturing financial metrics (e.g., Return on Assets (ROA), Debt-to-Assets ratio (D/A), and total assets) of companies issuing the bonds.
• Market returns from S&P Global 1200 Index data, representing global equity market trends.
• Investor attention data from Google Trends, measuring public interest or searching activity for terms related to green bonds or specific environmental themes.

• Yield Spread Model

The yield spread is a critical measure used to evaluate the difference in yields between green bonds and conventional bonds. It serves as a proxy for understanding the additional risk or premium investors associate with green bonds compared to traditional bonds and is calculated as the difference between a bond’s yield and the benchmark government bond interest rate for the corresponding month:

Yield Spread = Bond Yield − Government Bond Interest Rate

(1)

To calculate the yield spread, it is first necessary to calculate the monthly government bond interest rate:

Monthly government bond interest = (Sum of interest of all government
bonds)/Number of bond issues in a month

(2)

A matching method is employed to integrate data from different sources into a unified dataset. The merged dataset ensures that each observation includes bond-specific data, corresponding government bond interest rates, and issuer-level financial information.

To ensure the accuracy and reliability of the analysis, several filtering steps are applied to remove incorrect, incomplete, or extraneous data:

• Remove duplicates: rows with duplicated entries across data sources are eliminated.
• Exclude missing data: observations missing any critical information (e.g., bond yields, issuer financials, or government bond interest rates) are removed.
• Filter based on plausibility criteria: observations with unrealistic or implausible values are excluded, including the following:

  –

  Bonds with yield spread lower than 0 (indicating nonsensical pricing).

  –

  Bonds with negative interest rates exceeding 30% (highly unrealistic scenarios).

  –

  Government bonds with negative interest rates (a rare and extreme condition).

  –

  Issuer financials with

  –

  Tangible Index greater than 1, reflecting inconsistencies in financial reporting.

  –

  Total assets below USD 100,000, excluding small or atypical enterprises.

  –

  ROA below −50% or above 50%, capturing extreme financial performance.

  –

  Debt-to-Assets (D/A) ratio greater than 1, indicating excessive leverage beyond feasible financial structures.

The cleaned dataset is used to perform a regression analysis, with the yield spread as the dependent variable. Key independent variables include the following:

• Bond characteristics: coupon rate, maturity, issue size, and whether the bond is green or conventional.
• Issuer financial metrics: ROA, total assets, D/A ratio.
• Investor attention: Google Trends data, particularly around significant events like COP26 and COP27.
• Event dummy variables: indicators for COP26 and COP27 to capture their impact on yield spreads.

• CAR Model

The CAR model evaluates the impact of specific events, i.e., COP26 and COP27, on cumulative returns. By isolating abnormal returns (AR) from actual returns, the model highlights deviations from expected performance. This process involves data matching, estimating expected cumulative returns, calculating abnormal returns, and aggregating these values over a defined event window.

The estimation window is set to 60 trading days, from 70 days to 11 days before the bond’s issue date [−70, −11], to avoid contamination from preissuance effects that might influence returns closer to the issue date. An Ordinary Least Squares (OLS) regression model is used to estimate the expected cumulative return (CR) as follows:

Cumulative return (CR) = cons + β*Number of days from 70 days before the issue
date + ε

(3)

Whereas

• α: constant term representing the baseline return.
• β: coefficient capturing the trend in cumulative returns over the estimation window.
• ϵ: residual term accounting for unexplained variations.

Abnormal return (AR) measures the deviation of actual returns from the expected returns on a given day and is calculated as

Abnormal return (AR) = Actual cumulative return day n − expected cumulative return day n

(4)

The CAR aggregates abnormal returns over the event window, capturing the total impact of the event on returns:

$$\mathit{Cumulative}\mathit{abnormal}\mathit{return}\left(\mathit{CAR}\right)={\sum }_{-10}^{10}Abnormal return$$

(5)

The event window spans 21 trading days, from 10 days before to 10 days after the issue date [−10, 10], which ensures that both pre- and post-issuance effects are captured, allowing for a comprehensive analysis of the event’s influence.

3.2. Research Model

• Yield Spread Model and Variables

$$\begin{array}{c}\mathrm{Yield} \mathrm{Spread}=\mathrm{cons}+{\mathrm{\beta }}_1\mathrm{Green}+{\mathrm{\beta }}_2\mathrm{IA}+{\mathrm{\beta }}_3\mathrm{GreenxIA}+{\mathrm{\beta }}_4\mathrm{Market} \mathrm{risk}+\\ {\mathrm{\beta }}_5\mathrm{Maturity}+{\mathrm{\beta }}_6 \mathrm{AmountIssue}+{\mathrm{\beta }}_7\mathrm{Guaranteed}+{\mathrm{\beta }}_8\mathrm{Callable}+{\mathrm{\beta }}_9\mathrm{Putable}+\\ {\mathrm{\beta }}_{10}\mathrm{ROA}+{\mathrm{\beta }}_{11}\mathrm{Tan}\mathrm{gibility}+{\mathrm{\beta }}_{12}\mathrm{D}/\mathrm{A}+{\mathrm{\beta }}_{13}\mathrm{I}\mathrm{size}+{\mathrm{\beta }}_{14}\mathrm{COP}26+{\mathrm{\beta }}_{15} \mathrm{COP}27+\\ {\mathrm{\beta }}_{16}\mathrm{Green} \times \mathrm{COP}26+{\mathrm{\beta }}_{17} \mathrm{Green} \times \mathrm{COP}27+\epsilon \end{array}$$

(6)

Yield Spread: Coupon spread of bonds and government bonds with the same maturity and issue year.

All variables of Yield spread model are described in

Table 1. Yield spread model.

Variables	Description	Data Source
Green (dummy variable)	Green = 1 if the green bond Green = 0 if a conventional bond	Refinitiv Eikon
IA (lagged variable)	Investor attention: attention of investors to green bond	Google Trend
GreenxIA	Interaction variable between green bond and IA
Market risk	Market risk measured by the standard deviation of the S&P Global 1200 index	S&P Global
Maturity	Maturity of the bond	Refinitiv Eikon
ln AmountIssue	Natural logarithm of the issued bond volume	Refinitiv Eikon
Guaranteed (dummy variable)	Guaranteed = 1 if the bond is a guaranteed bond Guaranteed = 0 if the bond is not a guaranteed bond	Refinitiv Eikon
Callable (dummy variable)	Callable = 1 if the bond is a callable bond Callable = 0 if the bond is not a callable bond	Refinitiv Eikon
Putable (dummy variable)	Putable = 1 if the bond is a Putable bond Putable = 0 if the bond is not a Putable bond	Refinitiv Eikon
ROA (lagged variable)	Return on Assets	Refinitiv Eikon
Tangibility (lagged variable)	Tangible Assets to Total Assets	Refinitiv Eikon
D/A (lagged variable)	Debt to Total Assets	Refinitiv Eikon
lsize (lagged variable)	Natural Logarithm of Total Assets	Refinitiv Eikon
COP26 (dummy variable)	Period = 0 if the firm issues green bonds before COP26 or after COP27 Period = 1 if the firm issues green bonds during or after COP26
COP27 (dummy variable)	Period = 0 if the firm issues green bonds before COP27 Period = 1 if the firm issues green bonds during or after COP27
Green×COP26	Interaction variable between green bond and COP26
Green×COP27	Interaction variable between green bond and COP27

Source: authors.

as below:

• CAR Model and Variables

CAR = cons +β1 Green + β2 IA + β3 Green x IA + β4 Market risk + β5 ROA +
β6 Tangibility + β7 Leverage + β8 lsize + β9 COP26 + β10 COP27 +
β11 GreenxCOP26 + β12 GreenxCOP27 + ε

(7)

CAR: Cumulative abnormal return of corporate stock.

All variables of CAR model are described in

Table 2. CAR model.

Variables	Description	Data Source
Green (dummy variable)	Green = 1 if the green bond Green = 0 if a conventional bond	Refinitiv Eikon
IA (lagged variable)	Investor attention: attention of investors to green bond	Google Trend
GreenxIA	Interaction variable between green bond and IA
Market risk	Market risk measured by the standard deviation of the S&P Global 1200 index	S&P Global
ROA (lagged variable)	Return on Assets	Refinitiv Eikon
Tangibility (lagged variable)	Tangible Assets to Total Assets	Refinitiv Eikon
Leverage = D/A (lagged variable)	Debt to Total Assets	Refinitiv Eikon
lsize (lagged variable)	Natural Logarithm of Total Assets	Refinitiv Eikon
COP26 (dummy variable)	Period = 0 if the firm issues green bonds before COP26 or after COP27 Period = 1 if the firm issues green bonds during or after COP26
COP27 (dummy variable)	Period = 0 if the firm issues green bonds before COP27 Period = 1 if the firm issues green bonds during or after COP27
GreenxCOP26	Interaction variable between green bond and COP26
GreenxCOP27	Interaction variable between green bond and COP27

Source: authors.

as below:

3.3. Hypothesis Testing Methods

The study was conducted in six steps, in the following order:

F-test—Model fit: To assess the adequacy of the regression equation, or the percentage of the variance of the dependent variable explained by the independent variables in the regression equation, the coefficient of determination R² is used. The closer R² is to 1, the more meaningful the equation is.

t-test—Significance of regression coefficients: Since the data used to determine the parameters in the sample regression equation is based on the results of a specific sample survey, the next step is to test the significance of the regression coefficients. For a single-variable regression model, the regression coefficient is tested with the hypothesis that the regression coefficient is equal to 0, meaning that there is no relationship between X and Y in the population [51].

Multicollinearity test: Multicollinearity is a phenomenon in regression analysis where two or more independent variables are highly linearly correlated with each other. Since variables that are highly linearly correlated with each other do not provide any new information, it is not possible to determine the individual effect of each independent variable on the dependent variable [51]. Therefore, it is necessary to re-test the results to see if there is multicollinearity. This study uses the variance inflation factor (VIF) to test for multicollinearity.

Durbin–Watson test for first-order autocorrelation: Autocorrelation is the phenomenon where there is a correlation between the components of observations arranged in time order, then there is a relationship between the consecutive errors (residuals). This test is used to detect this phenomenon.

Test for heteroscedasticity of error variance: When two quantitative criteria—X and Y—satisfy the conditions of normal distribution, we can use the linear correlation coefficient to test whether there is a linear correlation relationship between the two criteria.

4. Results

4.1. Impact of COP26 and COP27 on Investor Attention

The descriptive statistics in

Table 3. Descriptive statistics of investor attention to green bond.

	N	Range	Minimum	Maximum	Mean	Std. Deviation
Investor attention to green bonds (worldwide)	157	94	6	100	23.83	12.985
Valid N (listwise)	157

Source: authors.

show the results of investor attention (IA) under the impact of two events, COP26 and COP27, for 157 weeks surrounding the two events. The IA values range from 6 to 100, with an average of 23.83. The standard deviation of IA is 12.985, implying a coefficient of variation of approximately 0.5. This suggests that while IA is evenly distributed around the events, the overall level of attention is relatively low.

The line chart illustrates that for both COP26 and COP27, IA surges abruptly within 5 and 8 months after each event, respectively. However, after the events, IA for green bonds declined compared to the pre-event period. This could stem from investor concerns about greenwashing practices, leading to a decrease in IA after COP26. However, the release of ICMA’s “Green, Social and Sustainable Bonds: A High-Level Mapping to the Sustainable Development Goals” in June 2022 rekindled investor interest in green bonds. Additionally, the success of green bond issuances in 2022, such as the HK$20 billion retail green bond issuance in Hong Kong in May 2022, kept IA for green bonds stable until COP27.

Regarding COP27 in Figure 1, Fairless stated, “I’d be surprised if COP27 delivered anything more meaningful than the US Inflation Reduction Act or the EU’s RePowerEU package. Both plans lay out ambitious decarbonization targets and billions of dollars in funding to support them”. As a result, global investor expectations for green bonds were tempered after COP27. In the first half of the year, S&P Global Ratings released its mid-year update on global bond forecasts for 2023: “Credit Trends: Global Financial Conditions: Market Resilience Supports Stronger-Than-Expected Issuance in 2023”, published on 26 July 2023, along with the green bond issuance volume of USD 310 billion in the first half of 2023, marking the highest half-year issuance since the green bond market’s inception, reignited strong investor interest in green bonds just ahead of COP28.

4.2. Impact of Green Bond Issuance on Yield Spread Before and After COP26 and COP27

The descriptive statistics in Appendix A show yield spread (YS) under the impact of two events, COP26 and COP27, for 70 days surrounding the two events. The YS is an average of 7.847. The histogram indicates that the mean of the residuals is approximately zero (−3.63 × 10⁻¹³), and the standard deviation is 0.999 (approximately 1). These results suggest that the distribution of residuals closely resembles a normal distribution in Figure 2.

The scatter plot has most points in the range [−3; 3], showing that the assumption of homoscedasticity is not violated in Figure 3.

Table 4. Regression result of yield spread model.

		B	Std. Error	Beta	VIF
	(Constant)	28.981 ***	0.746
	Green bond	−0.445 **	0.217	−0.024	5.291
	IA-1	−0.012 ***	0.001	−0.085	1.377
	GreenxIA	0.003	0.002	0.006	1.222
	Market risk	0.009 ***	0.001	0.077	1.558
	Maturity	−0.095 ***	0.007	−0.093	1.763
	ln amount	−0.490 ***	0.010	−0.349	1.887
	Guaranteed	1.942 ***	0.145	0.070	1.088
	Callable	−1.250 ***	0.081	−0.101	1.708
	Putable	−1.226 **	0.533	−0.012	1.035
	ROA	−0.162 ***	0.012	−0.074	1.145
	Tangibility	3.355 ***	0.647	0.029	1.293
	D/A	−8.188 ***	0.272	−0.174	1.341
	Isize	−0.672 ***	0.017	−0.298	2.162
	COP26	−0.528 ***	0.116	−0.047	4.371
	COP27	0.176 *	0.100	0.018	4.154
	GreenxCOP26	0.646 **	0.272	0.021	3.014
	GreenxCOP27	0.692 **	0.259	0.025	3.397
a. Dependent variable: yield spread b. ***, **, and * indicate significance levels at 1%, 5%, and 10%, respectively c. Durbin–Watson: 1.301 d. F statistic: 1468.353 ***

Source: authors.

shows the regression results impact of green bond issuance on yield spread before and after COP26 and COP27; the R-square is 0.622. The impact results are as follows:

The test results show that the VIF values for all independent variables are less than 10. This implies that multicollinearity is not present among the independent variables.

4.3. Impact of Green Bond Issuance on Cumulative Abnormal Return and After COP26 and COP27

The descriptive statistics in Appendix B show cumulative abnormal return (CAR) under the impact of two events, COP26 and COP27, for 70 days surrounding the two events. The CAR is an average of −62.137%. The histogram indicates that the mean of the residuals is approximately zero (−1.56 × 10⁻¹⁴), and the standard deviation is 0.999 (approximately 1). These results suggest that the distribution of residuals closely resembles a normal distribution.

There are few data points in the scatter plot that fall outside the range [−3; 3]. Based on this observation, it can be concluded that the assumption of homoscedasticity is not violated in Figure 4 and Figure 5.

The multicollinearity test results show that the VIF values for all independent variables are less than 10. This implies that multicollinearity is not present among the independent variables.

Table 5. Regression result of CAR model.

		B	Std. Error	Beta	VIF
	(Constant)	−1014.939 ***	99.890
	Green bond	99.058 ***	20.856	0.108	3.688
	Period COP27	338.284 ***	13.825	0.522	3.268
	Period COP26	87.910 ***	14.802	0.131	3.488
	GreenxCOP27	−109.659 ***	26.979	−0.071	2.174
	GreenxCOP26	−24.445	29.760	−0.013	1.763
	IA	0.456 ***	0.088	0.077	1.594
	GreenxIA	−0.629 ***	0.178	−0.057	1.890
	Market risk	0.828 ***	0.102	0.115	1.426
	ROA	699.748 ***	166.061	0.055	1.218
	Tangibility	114.707	75.110	0.020	1.233
	Leverage	164.410 ***	36.647	0.062	1.350
	Isize	20.170 ***	2.164	0.134	1.475
a. Dependent variable: CAR b. *** indicate significance levels at 1% c. Durbin–Watson: 1.441 d. F statistic: 95.021 ***

Source: authors.

shows the regression results impact of green bond issuance on CAR before and after COP26 and COP27 as follows:

5. Discussion

5.1. Impact of Green Bond Issuance on Investor Yield Before and After COP26, COP27

5.1.1. Impact of Green Bond Issuance on Yield Spread Before COP26

The study shows that green bonds have a lower yield spread than traditional bonds (−44.5 bps), consistent with hypothesis 1a and in line with previous research (−66 bps) [30] and (−34 bps) [19] for the Chinese market; (−17 to −29 bps) [16]; (−6 to −24 bps) [17]; (−1 to −14 bps) [13], further demonstrating the general trend in research that green bonds have lower yields than conventional bonds [24,25].

This suggests that investors accept a lower yield when holding green bonds because they value the green factors related to ESG (environmental, social, and governance) and the sustainability of green bonds. Green bonds help reduce CO₂ emissions, increase the proportion of renewable energy consumption, and help countries achieve Sustainable Development Goals (SDGs) [52], similar to the study of the Joint Research Center (JCR), helping to reduce CO₂ emissions by an average of 4%. This reduction is doubled for new green investments (not refinancing) at 8%. Investors appreciate the stringent governance process of green bonds [53]; the yield spread increases when green bonds have higher sustainability benefits, minimizing the adverse impact of property damage due to climate change and environmental degradation [12].

From a business perspective, green bonds with a lower yield spread help to promote corporate issuance and create an advantage due to lower debt capital costs. With a lower yield spread of 44.5 bps compared to traditional bonds, this is appropriate and much higher than the green bond certification cost of 0.1 bps from the Climate Bond Initiative (CBI). Issuing green bonds not only reduces debt costs but also reduces equity costs [54] by reducing information asymmetry between investors and businesses, increasing stock liquidity, and reducing the risks faced by businesses.

Governments also encourage businesses to issue green bonds through policy incentives such as a preferential monetary policy of 5%, a subsidy policy and a tax incentive of 4% in the Asian region (Asian Development Bank, ADBI), a 50% discount on bond issuance costs up to HKD 2.5 million for the first time bond issuance, 100% of external assessment costs up to HKD 800,000 of the Green and Sustainable Finance Grant Scheme (GSF Grant Scheme). This creates an increasing trend of green bond issuance due to the benefits it brings to both investors and businesses.

5.1.2. Impact of Green Bond Issuance on Cumulative Abnormal Return Before COP26

The study shows that the issuance of green bonds has an abnormally higher cumulative return than traditional bonds, confirming hypothesis 1b, consistent with the findings of [30]. This suggests that the issuance of green bonds is good news for investors, who expect sustainable corporate development and a commitment to corporate environmental, social, and governance goals. In fact, research by [55] shows that green bond issuance increases a firm’s ROA due to reduced debt costs and government subsidies. Investors view the issuance as an indicator that the firm is performing well and has high profits associated with green and sustainable factors, which in turn increases the abnormal return.

5.1.3. Impact of Green Bond Issuance on Investor Yield After COP26 and COP27

The COP26 event increased the yield spread of green bonds while decreasing the yield spread of conventional bonds. For the COP27 event, both green and conventional bonds experienced an increase in yield spread after the event, with the increase in green bond yield spread being higher. These findings confirm hypothesis 3a.

Figure 6 indicates that the COP26 and COP27 events both have a strong impact on the CAR, confirming hypothesis 3b. After each event, both green and conventional bond issuance have higher returns than before the event. This suggests that each COP event creates a strong incentive for investors to invest in green bonds, raising investor awareness of sustainable development in the context of climate change hurting the world.

However, Figure 7 indicates that while at COP26, green bond issuance has an increase in CAR and Yield spread compared to conventional bonds, COP27 saw the opposite trend. A plausible explanation is that while COP26 made a big step forward with over 200 countries agreeing to the Glasgow Climate Pact, increasing the likelihood of keeping the temperature rise to 1.5 degrees Celsius, with 90% of countries committing to net zero emissions by 2050, COP27 only made further progress by establishing a loss and damage fund to support countries most affected by climate change, whereas agreements on peaking emissions by 2025 and phasing out fossil fuels, including coal, and switching to cleaner fuels such as wind and solar power were not mentioned at this event.

Experts also share their gloomy expectations about COP27. Andy Howard, the Global Head of Sustainable Investment, does not expect huge things from COP27. It seems very improbable that major steps forward or statements in COP27. While expectations for COP27 are low, policy progress in other areas is more expected. Isabella Hervey-Bathurst, Global Sector Specialist, Multi-Region Equity, was more hopeful about the US Inflation Reduction Act or the EU’s RePowerEU package. Both plans represent ambitious decarbonization targets and billions of dollars of funding to back them up. As a result, investors have lower expectations for green bonds compared to the previous COP26 event, reducing the CAR of green bond issuance. Nevertheless, both COP events have driven green bond issuance by companies and other organizations.

5.2. Impact of Investor Attention on Investor Yield

The increase in investor attention on green bonds reduces both the yield spread of green bonds and conventional bonds (−1.2 bps), confirming hypothesis 4a, and reduces the CAR of green bond issuance while increasing the CAR of conventional bond issuance, confirming hypothesis 4b.

The results show that investor sentiment is an important factor affecting the yield spread of corporate bonds [37]. When investor sentiment is high, it leads to a higher yield spread, while the impact of sentiment on highly rated bonds is lower than on lower-rated bonds. High investor attention can create high (positive) sentiment—where investors are optimistic and confident in the market, leading to higher required returns—or low (negative) sentiment—where investors are pessimistic and prefer to hold bonds with lower yield spreads and stability.

Companies can attract investors to green bonds and increase demand for green bonds by increasing communication about green bonds, third-party certification and assurance, and using government green bond policy incentives, such as green bond investors in the US not having to pay income tax on bond interest and tax-exempt bonds in Brazil to finance wind power projects (tax incentives for issuers and investors, CBI). In addition, research by [56] shows that the green bond market is oversubscribed, meaning that investor demand exceeds supply. Therefore, investors accept a lower yield when holding green bonds due to the increased demand for green bonds and the tax advantage over conventional bonds.

5.3. Impact of Market Risk on Investor Yield

The study shows that when market risk increases, the yield spread of bonds also increases (0.36 bps), consistent with hypothesis 5a proposed, and this result is consistent with the study of the increase in volatility of S&P 500 futures contracts increases the yield spread [49]. A limitation of this study is the sample size of 688 bonds from Lehman Brothers through the Warga database, and the change in yield spread is explained very little by the independent variables along with contract volatility, suggesting that the volatility of S&P 500 futures contracts does not accurately represent market risk and has a low impact on bond yield spreads.

The study shows that market risk has a strong impact on the CAR of bonds. Increased risk creates a higher cumulative abnormal return, confirming hypothesis 5b. Increased market risk due to increased economic and social volatility makes investments riskier; therefore, investors will require a higher return to compensate for taking on additional risk. This is consistent with the theory of the Capital Asset Pricing Model (CAPM).

Our study uses the standard deviation of the S&P 1200, which is consistent with the sample size at the global level and is a direct measure of global market risk. When market risk increases, investors will also require a higher yield to compensate for the increased credit risk of the bond. The results are consistent with the studies [23,49]. The previous results conclude that increased market risk increases the credit spread of the issuer, meaning the bond risk issued by the company increases.

5.4. Impact of Callable and Putable Bond on Investor Yield

Our findings indicate that callable bonds exhibit a higher yield spread compared to non-callable bonds (32 bps), consistent with previous research [57], and aligned with the results of the time value of a call option contributes to higher yields [58]. This is reasonable as investors face increased risk due to the possibility of early bond redemption, leading to a loss of income and potentially limiting their ability to find alternative investments with comparable yields. Furthermore, callable bonds offer issuers greater flexibility in future investment planning. Firms issue callable bonds to mitigate risk in scenarios where investment opportunities deteriorate, enabling them to reduce debt costs while experiencing a decline in profits [59].

Our research also reveals that puttable bonds have a lower yield spread compared to non-puttable bonds (−164 bps), in line with the findings of (−13.3 bps) [60], (−144 bps) [61], and (−16.2 bps) [19]. Put options allow investors to sell the bond back to the issuer at a predetermined price should they wish to recover their principal early for personal reasons or to invest in another asset offering a more attractive yield. For issuers, puttable bonds can help reduce borrowing costs. However, according to the global financial information provider Cbonds, puttable bonds also have the potential to increase issuance costs if a significant number of bondholders decide to sell back to the issuer simultaneously. This can have a substantial impact on the firm’s cash flow and operations.

6. Conclusions and Policy Implications

6.1. Conclusions

Our study shows that green bonds had lower yield spreads compared to traditional bonds before COP26. However, the issuance of green bonds had higher cumulative abnormal returns compared to traditional bonds after COP26. The COP26 event increased the yield of green bonds while decreasing the yield of traditional bonds. For the COP27 event, after the event, both green bonds and traditional bonds increased their yields, with green bonds experiencing a higher increase.

Our research also shows that when market risk increases, the yield spread of bonds also increases. The study indicates that secured bonds have higher yield spreads compared to unsecured bonds.

6.2. Policy Implications

The study acknowledges the limitations of the model used, particularly regarding the inclusion of fixed and random effects. Many studies have used additional fixed and random effects models. However, this model requires each company to correspond to only one bond. Due to limited manpower and time, the research team could not proceed in this direction. Moreover, the research team has not fully addressed all the factors affecting a company’s cumulative abnormal returns. The authors only propose a research model with some key influencing factors. In addition, there are other factors that also affect the dependent variable.

Based on the obtained results, this article suggests the following recommendations. Firstly, for investors, the study suggests that equity investors should choose to buy shares of companies that issue green bonds after major environmental events to benefit from the higher CAR of these companies. Additionally, investors can use the S&P 1200 index as a measure to assess risk and abnormal returns when making short-term investments in shares of organizations that issue green bonds.

Governments and regulators can further encourage the issuance of green bonds by expanding financial incentives through tax exemptions for investors, subsidies for green projects, and reducing issuance costs. Additionally, to increase investor confidence, policymakers should establish stringent reporting and governance standards for green bonds. Third-party certifications and regular disclosures can assure investors of the legitimacy and effectiveness of green initiatives funded by these bonds. This would address concerns about “greenwashing” and increase demand for green financial instruments. Moreover, regulators should establish mechanisms to mitigate the adverse effects of market volatility on bond yields, such as liquidity support during periods of high market risk and promoting market stabilization measures. Lastly, lessons from COP26 and COP27 highlight the need for consistent progress in climate commitments to sustain investor enthusiasm for green bonds, where incorporating more ambitious and actionable plans in global agreements (e.g., clear strategies for phasing out fossil fuels) and complement COP initiatives with robust regional policies, which provide tangible support for decarbonization.

Future research can explore the relationship between green bond issuance and cumulative abnormal returns in various economic contexts and consider additional factors influencing these returns.