How Does Green Bond Issuance Facilitate the Spillover Effect of Green Technology Innovation in Industry? Evidence from China

Abstract

As the concept of balancing environmental protection and maintaining sustainable economic development has been widely recognized, the green bond is assuming an increasingly significant role within China’s financial market. We utilize the data from China’s A-share listed enterprises that issued bonds in the period 2010 to 2021 and try to examine whether and how green bond issuance facilitates the spillover effect of green technology innovation in industry. The results show that: (1) Green bond issuance can generate a spillover effect, greatly enhancing green technology innovation within the industry. (2) The spillover effect of green technology innovation from green bond issuance within an industry is more pronounced for state-owned enterprises, and relatively weaker for enterprises in Northeast China in the same industry. Relative to non-high-pollution industries, high-pollution industries reinforce the spillover effect. (3) Financing cost and agency cost are important influencing mechanisms for green bond issuance to improve peer enterprises’ level of green technology innovation. Overall, the results provide theoretical support for encouraging the market for green bonds to maintain their development over the long term and for effectively promoting the transformation of the economy and society to a green and low carbon one.

1. Introduction

In response to worldwide climate change, and to foster sustainable economic growth [1], many countries have taken corresponding measures. China has highlighted enhancements in fiscal taxation, finance, investment, price strategies, and standard frameworks to promote eco-friendly growth. Green finance’s foremost objective is to offer financial offerings centered on economic endeavors that enhance the environment, address climate change, and conserve and utilize resources effectively [2,3,4]. A green bond which meets the requirements of “green” and the standards of “finance” is an example of a fresh financing tool. It is a “star” tool issued by the government, banks, enterprises, and other institutions in the green financial system. As shown in Figure 1, recently, there has been a rising pattern in both the scale and quantity of green bonds issued in China.

Regarding the “financing” aspect of green bonds, enterprises issue them in order to obtain the advantages of lower financing costs [5], capital costs [6], and carbon risk [7], and to pursue maximum economic benefits. In another dimension, considering the “eco-friendly” and “sustainability” attributes of green bonds, issuing green bonds can help enterprises ease funding limitations [8], achieve adequate allocation of resources, and encourage enterprises to innovate in green technology while balancing economic gains with environmental responsibilities [9]. Existing research usually devotes much attention to the impact of bond issuing enterprises themselves, including enterprises achieving green technology innovation.

Existing research usually only focuses on green technology innovations carried out by enterprises which have issued green bonds [8,9,10]. While the release of green bonds is marked by strong information disclosure and can convey a “green signal” to the market [11], transforming the research perspective from within the enterprise to the whole industry and investigating whether green bond issuance produces a spillover effect may be a novel perspective. Therefore, in this study, the first research question is posed: whether green bond issuance encourages enterprises in the same industry to facilitate green technology innovation? Existing research is also mainly focused on the mechanism of enterprises issuing green bonds to promote green technology innovation and has rarely paid attention to the mechanism of generating a spillover effect within the industry. Therefore, the second research question is posed: how does green bond issuance encourage enterprises in the same industry to facilitate green technology innovation?

The paper tries to answer these questions: whether and how green bond issuance facilitates the spillover effect of green technology innovation in industry. In order to investigate these issues, this paper takes the dataset from China’s A-share listed enterprises that issued bonds in the period 2010 to 2021 to verify whether green bond issuance will facilitate enterprises’ level of green technology innovation within the same industry. Moreover, how financing and agency costs have an effect ais discussed by constructing a mediating effect model.

The innovations of our study are as follows: (1) We introduced the concept of the spillover effect of green technology innovation into the field of green bonds. From a novel perspective, we judged whether there was a spillover effect of green bond issuance. (2) We also analyzed the heterogeneity of the spillover effect in enterprises with different characteristics, carefully analyzing the impact of the spillover effect from green bond issuance. (3) Using the mediation effect model, we tested mechanism conduction effects of the financing and agency costs. We systematically constructed a theoretical framework for how green bond issuance affects the peer enterprises’ green technology innovation.

2. Institutional Background and Literature Review

2.1. Institutional Background

Many locations around the world have started extensive environmental protection programs and raised money for environmental protection in response to the harm caused by environmental contamination. The establishment and growth of green finance has established a strong basis for both sustainable economic growth and environmental preservation [12]. The World Bank introduced the idea of the “green bond” for the first time in 2007. It is described by the World Bank as a debt instrument that is issued specially to fund environmental or climate-related initiatives (Green Bonds (worldbank.org), accessed on 30 June 2024). As early as July 2007, the European Investment Bank issued a “climate-conscious bond”. It was the first “green bond” in the world, mainly for energy efficiency projects. As the concept of balancing environmental safeguards and maintaining economic prosperity has been recognized by many countries, some countries and regions have begun to promulgate relevant policies to accelerate the growth of green bond markets. Using information provided by the Green Bond Policy Data Set (Green Bond Policy Data Set|Climate Bonds Initiative, https://www.climatebonds.net/policy/data?page=1, accessed on 30 June 2024), the number of green bond policies was calculated for various nations (regions) between 2016 and 2019 (Figure 2).

To encourage the harmonious development between the environment and the economy, the Chinese government has also started to promulgate green bond policies since 2015. The Chinese green bond policy system has experienced different stages of development. The period 2015–2019 was the initial stage of the establishment of China’s green bond policy system. In July 2015, the first green bond in China was successfully issued. In the same year, the two policies, Notice on the Issuance of Green Financial Bonds in the Inter-bank Bond Market and Directory of Green Bond Support Projects, were introduced to proactively steer the issuance of green bonds in China. Since 2020, the policy system has been continuously optimized. The promulgation of more policies not only unifies the definition and definition standards of green bonds, but also enriches the relevant support projects. The market is flourishing, and the total value of Chines green bonds was approximately 3 trillion yuan by the conclusion of 2022. By concentrating on environmentally-specific objectives, the proceeds from green bonds can guarantee that enterprises will pursue green technology innovation and other activities subsequent to the issuance of green bonds, facilitate their comprehensive green transformation and restructuring, and enhance their reputation in the market. Furthermore, an analysis of the industries to which listed enterprises issuing green bonds belong has been conducted in China. The finance, manufacturing, electricity, heat, gas and water production and supply sectors accounted for the majority of green bond issuance, indicating that green bonds are primarily utilized for the green transformation of traditional energy-intensive sectors. The appeal of green bonds to investors lies in their capacity to fulfil a social responsibility and to provide a hedge against the risks of environmental protection policies.

In Figure 3, China’s green bond policies promulgated at different stages are shown.

This study focuses on the Chinese setting for three reasons. First, information provided in the Green Bond Policy Data Set shows that China has enacted the highest number of green bond policies between 2016 and 2019. This shows that China attaches great importance to the development of the green bond market. The Chinese green bond market is booming. Second, China, as a player in the transition economy, is promoting a green and low-carbon development model in the face of a “dual carbon” context. The development of green finance is strongly supported in China, promoting enterprises to use green financial instruments for financing and green transformation. Finally, the Chinese corporate economy, with its broad industrial structure, environmental policy implementation and mixed ownership, is able to reflect the development of the green bond market and enterprises’ green transformation more broadly. The implications drawn from the Chinese context, therefore, have the potential to provide valuable insights for other economies undergoing similar environmentally friendly transitions, as well as to inform the development of green finance and green transformation in different economies and sectors.

2.2. Literature Review

In the established literature on green bonds, scholars in China and other countries (regions) focus mainly on the issue value and market response. Some scholars concentrate on green bonds’ premium. Green bonds can significantly reduce credit spreads due to the “green” label [13], and as their credit spread is notably lower compared to those of ordinary bonds, this leads to a notable green premium [14]. The formation of this premium depends mainly on green attributes [14], and the research shows that green bonds had a transaction premium of 63 basis points (BPS) in comparison to corporate bond offerings [15]. In addition, enterprises issuing green bonds can better participate in sustainable financing activities and bring higher long-term value to enterprises [16], contributing to both fiscal and environmental performance [17]. Issuing green bonds is not only an advantageous measure for shareholders to carry out profitable green projects or to reduce risks [18], but also significantly improves the liquidity of stocks. These advantages win the favor of investors [16], and positively affect the company’s existing shareholders [19].

Numerous studies on the spillover effect of green bonds have often linked various financial markets with the green bond market. Stronger spillover effects are commonly present in the same category of commodities [19]. Green bond issuance has been observed to have spillover effects on investments in renewable energy and participation in cryptocurrency markets [20]. Moreover, scholars employed the quantile connectivity framework to investigate the influence of investor attention on the performance of green bonds through spillover effects. Their findings revealed that the spillover effect is time-varying, asymmetric, and influenced by market volatility, including fluctuations in stocks and bonds, as well as economic policy uncertainty [21], and the term “asymmetric” is used to describe the phenomenon whereby, in the event of a significant negative shock, green bonds are more closely monitored by retail investors, while the attention paid to green bonds by institutional investors is less pronounced.

Switching the research perspective to bond issuers, enterprises that issue green bonds encourage the development of green technology innovation [9,22]. Due to the expanded financing channels [6], the improvement of enterprise information transparency [23] and the reduction of agency costs [22], bond issuers may enhance the willingness to engage in green technology innovation.

As an important part of the green financial system, green bonds are one of the important tools to achieve the “dual carbon” goal. The spillover effect of green bond issuance should be studied, and is of great significance in encouraging enterprises to invest in green projects, such as energy saving and emission reduction and the circular economy. Relative to the existing research, we introduced the concept of green technology innovation spillover into the field of green bonds and enriched the research perspective. We systematically constructed a theoretical framework for green bond issuance and the spillover effect of green technology innovation and tried to deconstruct the mechanism of how green bond issuance affects peer enterprises’ green technology innovation.

3. Theoretical Backgrounds and Hypothesis Development

3.1. Green Bond, Spillover Effect, and Green Technology Innovation

A green bond is an instrument that uses its proceeds to support or refinance green projects [24]. Green bonds have bond financing and environmentally friendly attributes compared to ordinary bonds [2]. The term “green technology” primarily denotes nascent technologies aimed at diminishing consumption, curtailing pollution, enhancing the environment, fostering the development of an ecological society, and ensuring peaceful coexistence between humans and the natural world. Consequently, our view is that signals from enterprises releasing green bonds could influence the industry, prompting peers in the sector to actively adopt green technology innovation. The reasons for this are as follows.

First, bond issuers themselves may benefit from issuing green bonds. Green technology innovation refers to innovations in technology that enhances ecological quality [25]. Enterprises’ efficiency in resource utilization is enhanced and enterprises can be helped to face technical barriers and forge core competitiveness [26]. Enterprises need green technology innovation to foster sustainable growth and enhance their competitive edge [27]. By issuing green bonds, enterprises can spur innovation in green technology [22]. Green bonds, on the one hand, act mainly as an instrument for direct profit, substantially reducing information asymmetry in financial markets [9]. Enterprises’ issuance of green bonds increases the channels for external financing, which makes the internal capital flow more abundant [28]. On the other hand, due to government support for green projects, companies receive government subsidies due to the environmentally friendly nature of green bonds, resulting in lower financing costs than those of general credit bonds [29]. Enterprises’ release of green bonds has the potential to lower the expenses associated with financing [30], improve capital maturity mismatches, and ease corporate financing constraints [8]. Therefore, the capital allocation of enterprises is alleviated by external financial support and lower financing costs, which promotes technological innovation in enterprises [31].

Second, the term “spillover effect” denotes the tendency of decision-makers to replicate or assimilate the decisions made by their peers in a context of uncertain information [32]. Enterprises within the same sector often encounter comparable market environments and growth opportunities and, when one of them issues green bonds, it naturally draws the interest of other enterprises in the industry. Analyzed in the context of green bond characteristics, first, enterprises seek economic benefits. When enterprises successfully issue green bonds, this provides ideas for enterprises in the same sector to open up new financing channels. Besides offering financial backing for enterprises’ eco-friendly initiatives, green bonds also assist businesses in enhancing their reputation for sustainability and boosting their environmental performance and enterprise’s worth. Given the advantages and insights obtained by green bond issuers, enterprises within the same sector are eager to investigate the possibilities of green bonds as a funding option. Green bonds represent a source of funds earmarked for green projects, including green technology innovation. In anticipation of future opportunities, these enterprises within the same industry prepare to issue green bonds and thereby to access new financing channels. Secondly, from the standpoint of technological spillovers, issuing green bonds encourages enterprises to engage in green innovation [22] and boosts overall green technology standards in the industry. Under the logic of the positive effect of technological innovation spillover, their innovative technologies and concepts can be used as references and learnt from by peer enterprises, which will lead to the upgrading of new technologies in the industry. Furthermore, green bonds direct the influx of capital in support of green projects within the industry, which will facilitate the development of green projects. Concurrently, this will elevate the threshold of awareness regarding environmental legitimacy among enterprises within the same industry, thus increasing the willingness to pursue green technology innovation. Whether for the purpose of issuing green bonds to open up new financing channels, or due to technological spillovers and pressure for environmental legitimacy, peer enterprises will make the choice to implement green technology innovation under the dual considerations of pursuing economic efficiency and sustainable development.

Analyzed in the context of the theory of “spillover effect”, based on dynamic competition theory, enterprises in the same industry have spillover effects due to their interactive competitive relationships [33]. If enterprises launch green bonds to enhance their competitive edge in green technology innovation, their contemporaries in the sector will swiftly follow these green innovation practices. This reduces the barriers to competitors and the enterprise’s advantages. Then, when enterprises issuing green bonds are under competitive pressure from peer enterprises, the utilization of green bond financing will facilitate the acquisition of additional green innovation resources within the market. Competitive pressure leads to repeated dynamic games between the management of the two groups of enterprises, leveraging green bond issuance and green innovation behavior to build a business model that competes with enterprises in the same industry. This drives the whole industry to accelerate green technology innovation. Additionally, stemming from the demonstration effect and peer learning theory, enterprises are likely to mimic the manufacturing and operational behaviors of industry peers, particularly industry leaders, to augment their overall value [34,35].

Then, we propose Hypothesis 1:

H₁. 

Green bond issuance can produce spillover effect and enhance the level of green technology innovation of enterprises within the same industry.

3.2. Mechanism Analysis

3.2.1. Mechanism Effect of Financing Cost

Due to its information asymmetry and high-risk characteristics, the green innovative method makes it difficult to carry out a large amount of financing using the traditional financial model, thus limiting its development space [36]. Even if green innovation activities find it more difficult to obtain external financing, enterprises need enough funds to make internal capital allocation more abundant and solve the problem of environmental externalities, such as market failure [37]. Therefore, provided that green bond issuance can lower the reliance on external financing from enterprises and channel funds into industries that promote environmental safeguarding and sustainable growth, it can produce spillover effects and enhance green innovation in enterprises within the same industry.

Green bonds possess qualities akin to bond financing as well as environmental friendliness. Enterprises can issue green bonds, indicating a high legitimacy level [28], and they need to finance their green projects. However, the homogeneity of enterprises in the same industry is an objective and practical basis for spreading green signals. Therefore, investors infer that, when in the same industry, even companies that do not issue green bonds will also have environmentally friendly tendencies and green development potential. This reduces the credit risk expectations of enterprises in the same industry, thus helping enterprises to carry out external financing. Stemming from peer learning theory [38], enterprises are likely to mimic the manufacturing and operational behaviors of industry peers. Peer enterprises may take the initiative to make strategic decisions conducive to environmental protection and low financing costs. The reduction in financing costs for enterprises operating within the same industry [39] provides an additional advantage in the realization of green technology innovation.

Then, we propose Hypothesis 2:

H₂. 

Green bond issuance reduces enterprises’ financing cost and enhances green technology innovation within the same industry.

3.2.2. Mechanism Effect of Agency Cost

The success rates of enterprises’ green technology innovation are frequently impacted by the effectiveness of fund utilization, as these activities do not give priority to the distribution of enterprise resources and capabilities based on their unique characteristics. Consequently, in contrast to the pursuit of maximizing enterprise value, enterprise managers are inclined to focus on enhancing control rights throughout their tenure [40]. That is, the agency cost of enterprises will hinder enterprises’ green technology innovation.

Green bond issuance can effectively limit the short-term opportunistic behavior of managers, thereby ensuring the effective use of internal funds and the effective supervision of external investors [22], helping manage internal funds effectively to comply with laws and regulations while also supporting green projects through refinancing. In addition, according to the free cash flow hypothesis, this approach can also effectively prevent managers from abusing their power. Conversely, in terms of external investor oversight, green bond issuance has garnered public interest and mitigated the information gap between investors and enterprises [6,41]. Green bond issuance can help reduce the agency costs for enterprises, as enterprises in the same sector tend to have comparable preferences in their business practices [35], and managers of enterprises in the same industry will also take corresponding measures to reduce the agency cost of enterprises. When green bonds are issued within the industry, peer enterprises take the initiative or have to reduce their agency costs in order to obtain competitive advantages, under the competitive pressure of the same industry.

Then, we propose Hypothesis 3:

H₃. 

Green bond issuance reduces enterprises’ agency cost and enhances green technology innovation within the same industry.

The theoretical analysis path is shown in Figure 4.

4. Data and Methodology

4.1. Sample

In order to examine the longitudinal comparison of the time series before and after the issuance of green bonds, we took the dataset from China’s A-share listed enterprises that issued bonds in the period 2010 to 2021. The sample in this study was filtered based on specific criteria; we excluded enterprise samples that were ST or *ST (ST or *ST indicates that the enterprise is under special treatment by the exchange), enterprises in the financial sector, enterprises listed for less than one year and enterprises lacking essential financial data. To reduce the influence of outliers, this paper employed two-sided shrinkage (Winsorize) on the continuous variables at the 1% threshold. Finally, 11,428 “firm-year” observations were ultimately obtained. Data were sourced from the CNRDS database and the CSMAR database.

4.2. Variable

4.2.1. Explained Variable

Green technology innovation (GTI). Certain academics have noted that the delay in patent licensing hinders the precise assessment of innovation in GTI [42]. Typically, it takes 3 to 5 years to receive patent examination decisions. Patent applications effectively reflect the innovation efforts of enterprises in the current period. Thus, based on the studies conducted by Ma et al. (2021) [43] and Li and Xiao (2020) [44], the measure of GTI is taken as the number of green inventions independently applied by the enterprise in the year plus 1 as the natural logarithm. Greater values correspond to increased levels of green technology innovation.

4.2.2. Explanatory Variable

Whether the industry issued green bonds (Green × Post). Based on the study conducted by Wu et al. (2022) [39], the term “Green” signifies the presence of enterprises within the sector that release green bonds. If there are enterprises in an industry that issue green bonds publicly, the value of “Green” for the enterprises in that sector is assigned as 1; otherwise, it is assigned as 0. The variable “Post” indicates the time of green bond issuance. For the year when green bond issuance first appeared in the industry and in the following years, the “Post” value for enterprises within the same sector is assigned as 1; otherwise, it is assigned as 0.

4.2.3. Control Variables

Control variables (Controls): detailed explanations can be found in

Table 1. Variable definition.

Variable Symbol	Variable Definition
GTI	Ln (the number of independent applications for green inventions in that year + 1)
Green × Post	takes the value of 1 when the industry issues green bonds and is in a year in which green bonds have already been issued, otherwise takes the value of 0
Rd	the percentage of enterprise R&D expenditure to total assets
Size	the natural logarithm of enterprise total assets scale
Lev	net profit divided by average balance of shareholders’ equity
Roa	net profit divided by average balance of total assets
Cashflow	net cash flow from operating activities divided by total assets
Fixed	ratio of net fixed assets to total assets
Growth	(current year’s operating income/previous year’s operating income) − 1
Board	natural logarithm of the number of board members
Indep	independent directors divided by the number of directors
Dual	the chairman and the general manager are the same person 1, otherwise 0
Top	the number of shares held by the top five shareholders/the total number of shares
TobinQ	(circulating shares market value + number of non-tradable sharesmber of non-tradable shares × net assets per share + debt book value)/total assets
Big4	the company was audited as 1 by the Big Four (PWC, Deloitte, KPMG, Ernst and Young), otherwise it was 0
Opinion	if the company’s financial report of the year is issued a standard audit opinion, the value is 1, otherwise it is 0
Year	year dummy variable
Industry	industry dummy variable

.

4.3. Basic Model

Model (1) is developed to verify whether green bond issuance promotes the spillover effect of green technology innovation.

$$GT{I}_{i,t}={\beta }_0+{\beta }_{1}Green\times Pos{t}_{i,t}+{\beta }_{i}Control{s}_{i,t}+Year+Industry+{\epsilon }_{i,t}$$

(1)

${\epsilon }_{i,t}$ represents the random perturbation term. Furthermore, this paper employs firm-level clustering on the robust standard errors of the regression results. If ${\beta }_1$ is significantly positive, then Hypothesis 1 is verified.

5. Empirical Results

5.1. Descriptive Statistics

In

Table 2. Statistical results of main variables.

VarName	Observed Value	Mean	SD	Min	Median	Max
GTI	11,428	0.9273	1.2067	0.0000	0.6931	5.0499
Green × Post	11,428	0.1160	0.3203	0.0000	0.0000	1.0000
Rd	11,428	0.0214	0.0216	0.0001	0.0168	0.1412
Size	11,428	22.8538	1.3968	20.2300	22.6925	26.8319
Lev	11,428	0.5048	0.1948	0.0779	0.5118	0.9208
Roa	11,428	0.0418	0.0550	−0.1768	0.0378	0.2027
Cashflow	11,428	0.0440	0.0673	−0.1607	0.0450	0.2198
Fixed	11,428	0.2385	0.1725	0.0017	0.2066	0.7082
Growth	11,428	0.1965	0.4103	−0.5098	0.1285	2.6647
Board	11,428	2.1527	0.2025	1.6094	2.1972	2.7081
Indep	11,428	0.3762	0.0551	0.3333	0.3636	0.5714
Dual	11,428	0.2409	0.4276	0.0000	0.0000	1.0000
Top	11,428	0.5435	0.1607	0.1928	0.5415	0.9075
TobinQ	11,428	1.6892	0.9183	0.8282	1.3882	6.3924
Big4	11,428	0.0897	0.2858	0.0000	0.0000	1.0000
Opinion	11,428	0.9721	0.1647	0.0000	1.0000	1.0000

, the average GTI is 0.9273, with a standard deviation of 1.2067 and a median of 0.6931 for the variables that were described. This shows that the level of GTI among bond issuers has certain differences. The average Green × Post is 0.1160, showing that approximately 11.60% of enterprises will be impacted. The average value of Rd is 0.0214, ranging from a low of 0.0001 to a high of 0.1412, indicating a notable variation in enterprise R&D spending. The enterprise has an average Lev of 50.48%, with a maximum of 92.08%, suggesting that the majority of enterprises issuing bonds are exposed to higher financial risks and are responsible for more than half of the debt. The average value of Roa is 4.18%, indicating that most bond-issuing enterprises’ return on total assets is not high.

5.2. Analysis of the Empirical Results

Table 3. Baseline regression results.

Variable	(1)	(2)
	GTI
Green × Post	0.1589 ***	0.1534 ***
	(3.39)	(3.66)
Rd		7.3757 ***
		(9.33)
Size		0.5440 ***
		(21.52)
Lev		−0.4475 ***
		(−3.60)
Roa		−0.3893
		(−1.28)
Cashflow		−0.7333 ***
		(−5.28)
Fixed		−0.0346
		(−1.46)
Growth		0.1230
		(1.04)
Board		0.6086
		(1.63)
Indep		0.0178
		(0.43)
Dual		−0.0604
		(−0.51)
Top		0.0441 ***
		(2.65)
TobinQ		0.2386 ***
		(2.71)
Big4		0.2913 ***
		(4.28)
Opinion		−12.8638 ***
		(−21.33)
Constant	0.3498 *	7.3757 ***
	(1.77)	(9.33)
Year/Industry	Yes	Yes
N	11,428	11,428
adj. R2	0.206	0.462

Note: *** and * denote significance at the 1%, and 10% levels, respectively. Standard errors are clustered at the firm-level, and t values are reported in parentheses. The criteria in the table below are the same. Column (1) shows the regression result without adding control variables. Column (2) shows the regression result adding control variables.

indicates that the regression coefficients for Green × Post remain significantly positive, irrespective of the inclusion of control variables. This suggests that the level of peer enterprises’ green technology innovation enhances once the industry has issued green bonds, and Hypothesis 1 is verified. From an economic perspective, green bond issuance has led to a 15.34% boost in GTI among enterprises in the industry.

5.3. Robustness Tests

5.3.1. Propensity Score Matching Method

The PSM model is employed for testing, in order to avoid the estimation error caused by the differences noted between the experimental and control groups. Firstly, variables are screened by maximum likelihood estimation, and the variables related to enterprise characteristics, such as Size, Lev, Cashflow, Indep, Dual, Top, TobinQ, and FirmAge, are selected as covariates. Secondly, the kernel and 1:4 nearest neighbor matching techniques are employed to pair both the treatment cohorts with the control group. Finally, the PSM models are used to re-match the samples for the regression test. In

Table 4. PSM regression results.

Variable	The Kernel Matching	1:4 Nearest Neighbour Matching
	(1)	(2)
	GTI
Green × Post	0.1575 ***	0.0711 *
	(4.08)	(1.70)
Constant	−12.8114 ***	−12.5761 ***
	(−46.47)	(−40.76)
Controls/Year/Industry	Yes	Yes
N	11,399	9369
adj. R2	0.460	0.438

Note: *** and * denote significance at the 1%, and 10% levels, respectively. Column (1) shows the regression result for kernel matching. Column (2) shows the regression result of the nearest neighbor 1:4 matching method.

, Green × Post are positive and the aforementioned findings indicate that the outcomes are strong.

5.3.2. Placebo Test

A placebo test was adopted in this study by simultaneously randomizing the time of issuing green bonds and the treatment group sample. Referring to Bai et al. (2022) [45], Green_random of the pseudo-processing group and Post_random of the pseudo-issuance of green bonds were both randomly generated, i.e., the enterprise samples of 17 industries were randomly selected from the enterprise samples of all 44 industries, and the time of issuing green bonds was randomly generated to construct a new processing group, along with random industry and issuing green bonds. Five hundred repeated regressions were performed on the enterprise samples in 44 industries to guarantee the accuracy of the placebo test. For each trial, a random selection of enterprise samples from 17 different industries was assigned to the treatment group, concerning the timing of green bond issuance, also randomly determined. Finally, 500 sets of dummy variables were obtained. Figure 5 displays the p value and kernel density distribution of the coefficients of 500 Green_random × Post_random. In the random processing, the dummy variables’ coefficients predominantly cluster near 0, with the majority lacking significance. However, the calculated spillover effect coefficient from the Green × Post is 0.1534, which is notably distinct from the placebo test outcomes. The above conclusions show that the results are robust.

5.3.3. Change the Fixed Effect Models

To enhance the accuracy of the findings in this paper, individual, provincial, and time–provincial interaction fixed effects are included to mitigate the influence of other missing enterprise characteristics on the results. From

Table 5. Results of changing the fixed effect models.

Variable	(1)	(2)	(3)
	GTI
Green × Post	0.1692 ***	0.1530 ***	0.1561 ***
	(4.09)	(3.66)	(3.69)
Constant	−8.4223 ***	−12.0651 ***	−12.6226 ***
	(−9.46)	(−19.60)	(−21.31)
Controls/Year	Yes	Yes	Yes
Id	Yes	No	No
Industry	No	Yes	Yes
Province	No	Yes	No
Year#Province	No	No	Yes
N	11,428	11,428	11,428
adj. R2	0.741	0.474	0.462

Note: *** denotes significance at the 1% level. Column (1) shows the regression result for introducing the individual fixed effect. Column (2) introduces the province fixed effect. Column (3) introduces the time-provincial interaction fixed effect.

, the research discovered that Green × Post is notably favorable. Including extra potential control variables, does not substantially change the study’s conclusions.

5.3.4. Adjust the Interval of Sample

Firstly, it is necessary to exclude any other forms of policy interference. In order to ensure the reliability of the research results, it is crucial to account for other pilot policies that might have influenced enterprises’ GTI during the sample period. The Green Credit Guidelines, introduced in 2012, could influence how enterprises manage their green governance. The Guiding Opinions on Building a Green Financial System issued in 2016 may influence the volume of green bond issuance and encourage enterprises to engage in environmentally friendly practices, potentially affecting the empirical regression outcomes.

Secondly, it is necessary to reduce the influence of industry policies. The construction, engineering and electrical power industries represent the primary sectors in which issuers of green bonds are concentrated. Accordingly, we exclude the samples pertaining to policy time points and related industries, and re-run the regression. In

Table 6. Results of excluding other policy interferences.

Variable	Excluding the “Green Credit Guidelines”	Excluding the “Guiding Opinions on Building a Green Financial System”	Excluding Industry Policies
	(1)	(2)	(3)
	GTI
Green × Post	0.1422 ***	0.1721 ***	0.1106 ***
	(3.47)	(3.99)	(2.63)
Constant	−12.8263 ***	−12.5209 ***	−12.4774 ***
	(−21.57)	(−21.42)	(−20.62)
Controls/Year/Industry	Yes	Yes	Yes
N	10,508	10,408	10,878
adj. R2	0.464	0.466	0.461

Note: *** denotes significance at the 1% level. Column (1) excludes the effect of “Green Credit Guidelines”. Column (2) excludes the effect of “Guiding Opinions on Building a Green Financial System”. Column (3) excludes the effect of industry policies.

, the coefficients for Green × Post after excluding the effects of other policies are significantly positive.

5.3.5. Replace with Tobit Model

Since Green × Post is a dummy variable, it takes on values of either 0 or 1. We used the Tobit model to replace the benchmark regression OLS model for empirical testing. This conclusion is shown in

Table 7. Result of the Tobit model.

Variable	(1)
	GTI
Green × Post	0.1521 ***
	(3.63)
Constant	−12.5476 ***
	(−21.33)
Controls/Year/Industry	Yes
N	11,428

Note: *** denotes significance at the 1% level. Column (1) shows the result of the Tobit model.

. The finding indicates that the outcomes are strong.

6. Further Analysis

6.1. Heterogeneity Analysis

Enterprises show mixed reactions to industry’s green bond issuance, due to their own nature, the characteristics of the industry, the external conditions of the region and other factors affecting their decision on whether make more green technology innovation. Therefore, we analyze, from heterogeneous perspectives, i.e., at micro, meso and macro levels, specifically including the nature of enterprises, industry characteristics and regional characteristics.

6.1.1. Enterprise Characteristics

We conducted a grouping test on the type of property rights. In

Table 8. Heterogeneity analysis: the nature of property rights.

Variable	State-Owned Enterprises	Non State-Owned Enterprises
	(1)	(2)
	GTI
Green × Post	0.2172 ***	0.0333
	(3.37)	(0.62)
Constant	−14.2083 ***	−10.4777 ***
	(−15.91)	(−12.41)
Controls/Year/Industry	Yes	Yes
N	4757	6671
adj. R2	0.561	0.352

Note: *** denotes significance at the 1% level. Column (1) shows the result of the sample with state-owned enterprises. Column (2) shows the result of the sample with non-state-owned enterprises.

, the results indicate that state-owned enterprises are seen as crucial contributors to energy efficiency and emission reduction efforts, playing a unique and essential part in reaching these goals. If the executives of government-owned enterprises do not diligently carry out energy-saving and emission-reducing initiatives, these enterprises will consequently incur stricter penalties compared to their private counterparts [46]. Consequently, the prominent role of state-owned enterprises in meeting their environmental duties and promoting sustainable development is clear. When industry players issue green bonds to signal eco-friendly initiatives and boost green technology innovation, state-owned enterprises will proactively answer the call to elevate their own green technology innovation.

6.1.2. Industry Characteristics

To investigate the differences in the spillover effect, we divided enterprises into high-pollution or non-high-pollution industries. In

Table 9. Heterogeneity analysis: whether in high-pollution industries or not.

Variable	High-Pollution Industries	Non High-Pollution Industries
	(1)	(2)
	GTI
Green × Post	0.2339 ***	0.0831
	(3.94)	(1.44)
Constant	−9.8586 ***	−12.9178 ***
	(−11.71)	(−16.91)
Controls/Year/Industry	Yes	Yes
N	4701	6727
adj. R2	0.372	0.505

Note: *** denotes significance at the 1% level. Column (1) shows the result of the sample from high-pollution industries. Column (2) shows the result of the sample from non-high-pollution industries.

, the coefficient of Green × Post only in high-pollution industries is positive, suggesting that these enterprises are more attuned to the green signal within their sector, and they will also actively innovate in green technology to reduce competitors’ barriers. Enterprises in high-pollution industries have the characteristics of high energy consumption and high discharging pollutants. In order to avoid the penalties associated with environmental pollution, these enterprises are more inclined to implement green technology innovation. The above proves that the green signal generated by issuing green bonds significantly impacts these enterprises in industries with high-pollution levels.

6.1.3. Regional Characteristics

We conducted a grouping test on different regions. China’s economic regions can be roughly divided into the eastern, central, western, and northeast regions.

Table 10. Heterogeneity analysis: economic regional differences.

Variable	Eastern Region	Central Region	Western Region	Northeast Region
	(1)	(2)	(3)	(4)
	GTI
Green × Post	0.1456 ***	0.2336 **	0.1844 *	0.0441
	(2.64)	(2.41)	(1.76)	(0.27)
Constant	−13.3440 ***	−9.3928 ***	−9.4724 ***	−9.8002 ***
	(−17.51)	(−6.87)	(−7.22)	(−3.87)
Controls/Year/Industry	Yes	Yes	Yes	Yes
N	7343	1729	1853	503
adj. R2	0.510	0.421	0.425	0.443

Note: ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively. Column (1), (2), (3) and (4) show the results of the sample for eastern region, central region, western region and northeast region, respectively.

shows the grouping regression results for the four economic regions. Only the coefficient of Green × Post in the northeastern region group is insignificant. This shows that most regions can improve GTI when they issue green bonds to transmit green signals in the industry. Due to the enormous pressure on energy structure adjustment in Northeast China, the pressure on carbon emissions from increasing urbanization and large-scale infrastructure construction in the region, the long way to go for the formation of regional awareness of low-carbon lifestyles, and the undeveloped policy system for green transformation of enterprises, the energy structure in Northeast China has not yet been fully developed. In contrast, enterprises in the northeast region cannot promptly respond to the green signals transmitted in the industry.

6.2. Mechanism Test

We examined the transmission process of the green bond issuance’s spillover effect in relation to GTI by establishing models referred by Wen and Ye (2014) [47]:

$$Mediatio{n}_{i,t}={\beta }_1+{\beta }_{2}Green\times Pos{t}_{i,t}+{\beta }_{i}Control{s}_{i,t}+Year+Industry+{\epsilon }_{i,t}$$

(2)

$$GT{I}_{i,t}={\gamma }_1+{\gamma }_{2}Green\times Pos{t}_{i,t}+{\gamma }_{3}Mediatio{n}_{i,t}+{\gamma }_i{\mathrm{Controls}}_{i,t}+Year+Industry+{\epsilon }_{i,t}$$

(3)

There are two mediating variables in Mediation_i,t: financing cost (Cost) and agency cost (Agcost). Debt financing cost (Cost) refers to the practice of Zheng et al. (2013) [48]: Cost is calculated by dividing financial expenses by the overall debt value at the period’s conclusion. Agency cost (Agcost) refers to the practice of Bae et al. (2002) [49], measured by the proportion of other receivables to the enterprise’s total assets. Other variables are the same as for Model (1).

6.2.1. Financing Cost

We tested the transmission mechanism: green bond issuance—lowing peer enterprises’ financing cost—improving peer enterprises’ GTI. The findings in column (2) of

Table 11. Mechanism analysis: decrease financing cost.

Variable	(1)	(2)	(3)
	GTI	Cost	GTI
Green × Post	0.1521 ***	−0.0030 ***	0.1486 ***
	(3.62)	(−3.52)	(3.53)
Cost			−1.1485 *
			(−1.81)
Constant	−12.5476 ***	0.0301 ***	−12.5052 ***
	(−21.27)	(3.23)	(−21.31)
Controls/Year/Industry	Yes	Yes	Yes
N	11,428	11,428	11,428
adj. R2	0.462	0.340	0.462

Note: *** and * denote significance at the 1% and 10% levels, respectively. Column (1) shows the result for model (1). Column (2) shows the result for model (2) and examines that issuing green bonds can lower their industry’s overall financing cost. Column (3) shows the result for model (3).

show a notable decrease in debt financing expenses, suggesting that companies issuing green bonds can lower their industry’s overall financing cost. The third column displays the regression outcomes for the Cost that was incorporated in the model. Peer enterprises have a notably negative coefficient of 1% for the Cost. Green × Post is notably positive. Green bond issuance sends a signal, reducing peer enterprises’ financing cost and enhancing GTI. Hypothesis 2 was verified.

6.2.2. Agency Cost

We tested the transmission mechanism: green bond issuance—reducing peer enterprises’ agency cost—improving peer enterprises’ GTI. The results in

Table 12. Mechanism analysis: decrease agency cost.

Variable	(1)	(2)	(3)
	GTI	Agcost	GTI
Green × Post	0.1521 ***	−0.0027 ***	0.1500 ***
	(3.62)	(−3.20)	(3.87)
Agcost			−0.8195 **
			(−2.13)
Constant	−12.5476 ***	0.0192 ***	−12.5319 ***
	(−21.27)	(2.77)	(−47.76)
Controls/Year/Industry	Yes	Yes	Yes
N	11,428	11,428	11,428
adj. R2	0.462	0.187	0.462

Note: *** and ** denote significance at the 1% and 5% levels, respectively. Column (1) shows the result of model (1). Column (2) shows the result of model (2) and examines that green bond issuance can lower agency costs for companies in the same industry. Column (3) shows the result for model (3).

, column (2) from model (2) shows a notable decrease in Agcost, indicating that green bond issuance can lower agency costs for companies in the same industry. Column (3) displays the agency cost within the same industry in the model, suggesting a mediating effect. Enterprises issuing green bonds send a signal, reducing peer enterprises’ agency cost and enhancing GTI. Hypothesis 3 was verified.

7. Conclusions and Suggestion

The green bond has both “green” and “financial” attributes and is a “star” tool in the green financial system. Enterprises will seek direction from the indicators provided by green bond issuance to foster GTI and enhance the robust growth of the industrial sector. The aim of this paper is to analyze whether and how green bond issuance promotes the spillover effect of green technology innovation in industry. For our empirical analysis, we used the data from China’s A-share listed enterprises that issued bonds in the period 2010 to 2021 and established the mediating effect test method. We questioned whether and how green bond issuance promotes the spillover effect of green technology innovation. Our conclusions are as follows. First, green bond issuance can generate spillover effect in the industry, i.e., green bond issuance enhances peer enterprises’ GTI. Second, findings from the heterogeneity analyses reveal that, high-pollution enterprises often pursue the “green signal” of green bond issuance to improve GTI, similar to state-owned enterprises, whereas enterprises in Northeast China do not follow this trend. Third, further mechanism analyses show that green bond issuance reduces the financing and agency costs of enterprises within the same industry, which in turn improves enterprises’ level of green technology innovation. These suggestions follow.

First, it is worth advocating that, regulators fortify the information sharing system of green bond market and enhance its clarity, motivating enterprises to seek economic gains while focusing on environmental sustainability.

Second, innovations in green technology within enterprises are deeply intertwined with external resources. In order to address the issue of inadequate funds for eco-friendly development, enhance the support policy for eco-friendly technology innovation, and offer financial aid to companies engaging in eco-friendly innovation initiatives, the government should speed up the advancement and enhancement of the eco-friendly financial system. At the same time, enterprises shift their approaches to development and funding, adopt green financial mechanisms to support green innovation, and take into account the goal of sustainable development while pursuing economic benefits, so as to pursue sustainable development to achieve long-term growth.

Third, environmentally friendly enterprises are not inherent. This illustrates that the transformation of conventional enterprises into environmentally sustainable entities is a protracted process of transitioning from a conceptual framework to a behavioral one. To enhance this transformation, government provides incentives for green bonds and offer rewards to companies that participate in eco-friendly initiatives in order to motivate them to fulfill their obligation to safeguard the environment cannot be ignored.

Lastly, from an investor point of view, attention could be paid to promoting the concept of socially responsible investment, and more attention may be paid to green bonds as an innovative way of financing environmental protection in the investment process. Faced with the economic and social impacts of climate change, other countries could also learn from Chinese green finance policies in order to achieve environmental protection and green transformation while promoting economic growth. At the same time, with the relative abundance of business types in China, enterprises in other countries can learn from how Chinese enterprises of all types are making the green transition and improving their environmental performance.

It can be acknowledged that our study is not without limitations and that there is scope for further research in this area. Firstly, while the potential for green technology innovation spillovers from green bond issuance among enterprises in the same industry has been discussed, the impact of these spillover effect on the quality of green technology innovation remains unaddressed. This represents an avenue for further investigation that would be worthy of consideration in future research. Secondly, the impact of whether or not green bonds are issued within the industry selected for this study can only be observed at the macro level, in terms of the overall extent of green innovation spillovers. Subsequent research could attempt to examine the quantity or number of green bonds issued within the industry, with a view to exploring the impact of different dimensions. Thirdly, the present study is limited in that it employs a sample comprising solely Chinese listed enterprises. It would be beneficial for future research to conduct multi-sample studies to explore the green innovation spillovers of enterprises from different countries and industries.