Do Green Bonds Help to Improve Enterprises’ Financing Efficiency? Empirical Evidence Based on Chinese A-Share Listed Enterprises

Abstract

This study investigates the relationship between green bonds and enterprises’ financing efficiency. A three-stage data envelopment analysis (DEA) model and a fixed effects model are used to achieve the research objectives. This paper analyzes the dual dimensions of theoretical analysis and empirical investigation. By fully considering the sub-stages of the financing process, it introduces green bonds into the analytical framework of financing efficiency issues. This paper uses data from China’s A-share listed enterprises from 2000 to 2022, uses a three-stage DEA model to measure the efficiency of each sub-stage of enterprises’ financing, and uses a fixed effects model for empirical testing. The study found that issuing green bonds can significantly improve the financing efficiency of enterprises, especially in the total and repayment stages. Furthermore, this paper uses the intermediary effect model to discuss the inherent mechanism of green bonds affecting financing efficiency. Green bonds promote the financing efficiency of enterprises and promote green transformation by affecting investor recognition and financing costs. However, the impact of green bonds is not obvious during the fund use stage and may be related to transparency and accountability mechanisms. This result indicates that expanding investor recognition, financing costs, and green transformation through green bonds is crucial to successfully promoting financing efficiency. The moderation effect model shows that the effect of green bonds issued by state-owned enterprises and highly polluting enterprises is more significant. This study highlights that green bonds positively impact financing efficiency and help promote sustainable economic development. This study also has policy implications for stakeholders.

1. Introduction

As the second largest global economy, China’s capital market is pivotal in refining the economic structure and fostering sustainable growth [1]. Despite significant reforms and innovations over the past decades [2], challenges persist in boosting financing efficiency, primarily due to information asymmetry and suboptimal credit assessments that disproportionately affect enterprises. Consequently, enterprises face burdensome, costly, and inefficient financing processes [3]. Moreover, enterprises predominantly rely on traditional bank loans, underutilizing alternative financing avenues such as equity and bond markets [4].

On the other hand, China faces severe environmental challenges as one of the world’s top carbon emissions and energy consumer countries [5]. In response, China has vigorously promoted green finance since 2016, notably through issuing green bonds. By 2020, China’s green bond balance reached CNY 813.2 billion, positioning it as the second largest market globally. This market is essential for funding climate initiatives, offering significant opportunities for businesses and investors [6]. However, a substantial gap in green financing persists, with incentives and support for green investment still requiring enhancement [7]. The CICC Global Institute (CGI) 2021 estimates a yearly green finance shortfall of approximately CNY 0.54 trillion to meet China’s carbon peak targets and around CNY 1.3 trillion to achieve carbon neutrality, highlighting green bonds’ crucial role and potential in advancing China’s green finance agenda.

Green bonds, a novel financial instrument in green finance, have attracted significant academic attention due to their comprehensive impact on the environmental, economic, and financial sectors. The research covers various aspects, including ecological outcomes [8], economic benefits [9], market evolution [10], and investor behaviors [11]. Green bond issuances contribute to sustainable enterprise operations and enhance long-term value creation, thereby improving financing efficiency by optimizing enterprise financial structures [12]. Additionally, Zhang et al. [13] and MacAskill et al. [14] observed increased investor interest in green projects, indicating better financing conditions for issuers. Furthermore, Bel and Joseph [15] noted that green bonds attract socially responsible investors, thereby boosting financing efficiency. Therefore, this paper aims to clarify the potential reasons why green bonds promote the improvement of financing efficiency and then provide policy references for accelerating the progress of corporate financing efficiency, easing financing constraints, and achieving high-quality economic development.

Research on green bonds primarily addresses environmental and financial impacts, often overlooking their roles in enhancing financing efficiency [16]. Additionally, studies predominantly focus on developed economies, neglecting emerging markets and developing countries [17,18,19], which narrows the global perspective on the effectiveness of green bonds. A thorough assessment of social benefits, long-term effects, and financing mechanisms is crucial for fully appreciating the implications of green bonds.

In addition, business operation is a complex process, and the financing process of an enterprise and the allocation and use of funds after financing is also a process. Therefore, the traditional DEA model can only reflect the overall financing efficiency of the enterprise, while the three-stage DEA model considers this problem and refines the financing process into three sub-stages [20,21], that is, an intermediate product is added between the second and third stages, which is the output of the first and second stages and the input of the second and third stages. Current research is also limited to analyzing the overall financing efficiency using DEA or SFA methods, without considering the more detailed stages of the financing process [22,23]. The above analysis shows that, although existing studies have conducted relatively rich discussions on the impact of green bonds on the development of the green economy and the improvement of enterprises’ financial performance and its transmission mechanism, due to the lack of consideration of different stages of financing efficiency there are relatively few studies on how green bonds affect financing efficiency and its different stages, and even fewer discussions on its potential impact mechanism. Based on this, we can clarify a core hypothesis that the issuance of green bonds helps improve the financing efficiency of enterprises. This hypothesis is based on the market signaling effect, investor preference, and potential government support that green bonds may bring. Specifically, green bonds can serve as a signal to convey to the market the commitment of enterprises to environmental responsibility and attract investors who focus on sustainable investment. In addition, green bonds may benefit from incentives provided by the government, such as tax breaks or subsidies, which can help reduce the financing costs of enterprises. Moreover, issuing green bonds can motivate enterprises to become green, further improving their financing capabilities. To verify these hypotheses, we will use empirical analysis methods to evaluate the changes in financing efficiency before and after the issuance of green bonds by Chinese A-share listed companies.

The marginal contributions of this paper are as follows. Firstly, it integrates green bonds into the analytical framework for enterprise financing efficiency, establishing a theoretical linkage between green bonds and enterprise financing efficiency. Secondly, traditional DEA models only reflect an enterprise’s overall financing efficiency. Hence, this paper utilizes a sophisticated three-stage DEA model that captures the complexity of the financing process, illuminating mechanisms that boost financing efficiency and offer actionable insights for policy development. In the context of enterprise financing, this paper analyzes in detail whether green bonds can influence the efficiency of each financing stage, leading to an overall improvement in financing efficiency. The three-stage DEA model, a novel approach introduced in this study, not only helps to identify potential factors contributing to the enhancement of financing efficiency but also provides a fresh perspective on the sub-stages of the financing process. This unique perspective can serve as a valuable policy reference for accelerating the improvement of enterprises’ financing efficiency. Third, the study delves into green bonds’ effects on investor recognition, financing costs, and green transformation, shedding light on their crucial role in advancing enterprises’ financing efficiency and overcoming market challenges. The findings offer new theoretical perspectives and empirical evidence to support policymaking.

The subsequent chapters are organized as follows. Section 2 introduces the literature review and hypothesis development in detail. Section 3 presents the methods and data used. Section 4 provides an in-depth analysis of the empirical results, and Section 5 is an extended heterogeneity analysis. The last section summarizes the research and offers relevant implications for investors and policymakers.

2. Literature Review and Hypothesis Development

Green bonds, a financial instrument that meets investors’ demand for sustainable development and provides a channel for enterprises to secure low-cost capital [24], are not just dedicated to pro-environmental development projects [25] but are also a crucial tool in mitigating environmental risks, while effectively reducing uncertainty and perceived risk in the use of funds by investors [26]. By issuing green bonds, enterprises demonstrate their commitment to environmental protection and sustainable development [27], shaping a green enterprises image and enhancing public trust and recognition [28], translating into lower financing rates [29] and increased competitiveness in the financial market [30], thereby reducing financing costs. Furthermore, green bonds facilitate enterprises in broadening financing channels, diversifying funding sources, and increasing financing flexibility and diversity [31,32]. Enterprises often rely on internal funds or limited financing channels to support business development when facing financing constraints. Solving the financing constraint problem can help enterprises broaden their channels and ways of raising funds, and improve their financing capabilities. Based on the analysis, the first hypothesis is proposed:

H1:  

Green bonds can improve the financing efficiency of enterprises.

The emergence of green bonds has positively impacted investor recognition of enterprises’ financing efficiency. Enterprises issue green bonds to raise funds for environmental projects, demonstrating their commitment to environmental protection and sustainable development [26]. This enhanced environmental image helps to enhance investors’ recognition of enterprises, especially those with environmental awareness and social responsibility. More and more investors are beginning to incorporate ESG factors into their investment considerations, and they are more willing to support enterprises that actively promote environmental protection and sustainable development [33]. Therefore, enterprises that issue green bonds are likely to win higher recognition among these investors and thus obtain more financial support, thereby improving financing efficiency. The green bond market is relatively new, but, as the focus on sustainable development continues to increase, this market will gradually grow. Enterprises choosing to issue green bonds can help investors enter this emerging market and improve market liquidity for financing. More investors participating in the green bond market may lead to higher trading activities, thereby improving the efficiency of enterprises’ financing.

Green bonds can often obtain lower financing costs or higher prices than traditional ones, which means that the market has a higher degree of recognition for green bonds, and investors are willing to pay a higher price for them, thereby reducing financing costs for enterprises and improving financing efficiency [34]. The issuance of green bonds may attract investors who have long been concerned about enterprises’ sustainable development. These investors are more inclined to cooperate with enterprises with consistent long-term strategic goals, which positively impacts enterprises’ financing efficiency and long-term development. On the other hand, some investors, especially institutional investors and socially responsible investors who focus on sustainable development, prefer to invest in enterprises with green bonds. At the same time, investors’ demand for green bonds has also promoted the development and innovation of the market. This preference may lead to more investors willing to buy enterprises’ bonds, improving the efficiency of enterprises’ financing. By issuing green bonds, enterprises can attract a group of investors with environmental tendencies and provide more options for their financing activities. During the fundraising stage, investors usually evaluate an enterprise’s market potential, innovation capabilities, and management team strength. High investor recognition can reduce financing costs and speed up fundraising. The ability of the enterprise to repay debts directly affects its credit rating and future financing conditions. Based on the analysis, the second hypothesis is proposed:

H2:  

Green bonds affect financing efficiency by increasing investors’ recognition.

From the perspective of financing constraints and pecking order financing theory, green bond issuance can often obtain more favorable financing conditions. As investors favor environmental protection and sustainable development projects, they may be willing to purchase green bonds at a lower interest rate, thereby reducing the financing costs of enterprises [35]. This low interest rate tendency reflects the recognition of green bonds in investors’ minds, which is conducive to enterprises reducing financing costs and improving financing efficiency. In addition, with the improvement of environmental awareness, more and more investors are paying attention to investment opportunities in environmental protection and sustainable development. Therefore, the issuance of green bonds by enterprises may attract more investors to participate, increase the demand for bond subscriptions, and thus reduce financing costs. If investors’ demand for environmental protection and sustainable development projects increases, the demand for green bonds will also increase. Enterprises’ issuance of green bonds can meet this demand and increase its attractiveness among investors. As investors’ demand for green bonds increases, bond prices may increase, thereby reducing the financing costs of enterprises.

Green bonds may receive higher prices or better liquidity in the market, which reflects the market’s increased recognition of them. Investors are willing to pay higher prices for green bonds or sell them at a lower premium, thereby reducing the financing costs of enterprises and improving financing efficiency. The last point is the recognition of long-term investment. Investors holding green bonds are usually more concerned about enterprises’ long-term development and sustainability, and they may be more willing to establish long-term cooperative relationships with enterprises [36]. This long-term investment recognition helps enterprises reduce financing costs and improve financing efficiency. Initial financing costs are usually high because the high risk increases the cost of funds. The effective use of funds can reduce financing costs because successful project implementation improves enterprises’ credit. Based on the analysis, the third hypothesis is proposed:

H3:  

Green bonds affect financing efficiency by reducing financing costs.

From the perspective of information asymmetry theory, stakeholder theory, and reputation theory, investors consider economic benefits when allocating funds, as well as enterprises’ social responsibility and environmental performance. By issuing green bonds, enterprises convey positive signals of environmental protection and sustainable development to investors, which meets the needs of investors for socially responsible investment, thereby improving the financing efficiency of enterprises [31]. As a financial tool for sustainable development, green bonds help enterprises incorporate environmental protection and sustainable development into strategic planning, requiring enterprises to use the raised funds for environmental protection and sustainable development projects, which prompts enterprises to accelerate green transformation. By promoting and implementing environmental protection projects and sustainable development strategies, enterprises can improve resource utilization efficiency (UE), and obtain better operating benefits and competitive advantages, achieving win–win economic and social benefits, thereby improving financing efficiency.

The issuance of green bonds provides enterprises with a financing channel, making it easier for them to raise funds for environmental protection projects [24]. These projects can include energy efficiency improvement, clean energy development, waste management, etc. Enterprises can implement more environmental protection measures through green transformation and promote their financing efficiency. Finally, participating in the green bond market enables enterprises to seize market opportunities in sustainable development [31]. As the green bond market matures, investors’ demand for environmental protection projects continues to increase, which provides enterprises with more market opportunities [27]. By issuing green bonds, enterprises can achieve green transformation and find new business opportunities in sustainable development. Although the initial cost is high, implementing green transformation in operations, such as adopting energy-saving technologies or sustainable materials, can reduce operating costs in the long run [29]. In addition, timely debt repayment can reduce interest expenses, thereby reducing overall financing costs. Based on the analysis, the fourth hypothesis is proposed:

H4:  

Green bonds affect financing efficiency by facilitating green transformation.

3. Research Design

3.1. Data and Variables

The data source for green bond issuance is mainly the Wind database, supplemented by the CSMAR database, and the data from the China Financial Information Network (CFIN) are also compared to ensure accuracy and reliability. The CSMAR database is a large-scale financial database developed by Shenzhen Guotai’an Information Technology Co., Ltd., Shenzhen, China. It provides detailed data on China’s financial markets, including stock markets, bond markets, fund markets, derivative markets, macroeconomics, industry data, company financial data, corporate governance data, market transaction data, etc. In terms of representativeness in the field of green bonds in China, the CSMAR database contains green bond issuance information, including the issuance date, maturity date, coupon rate, issuance scale, issuer, bond rating, and use of raised funds of green bonds. Market transaction data: information such as the transaction price, trading volume, and yield of green bonds in the secondary market, which helps to analyze the liquidity and market performance of green bonds. Company financial data: financial statement data of companies issuing green bonds, such as income statements, balance sheets, and cash flow statements, can be used to evaluate the financial status and debt repayment capacity of companies issuing green bonds. The CSMAR database provides comprehensive data covering China’s financial markets, including various financial instruments and related economic indicators, including green bonds. All control variables at the enterprise level are also taken from the CSMAR database. Although the CSMAR database has released data for 2023, we found that many listed companies’ data are missing after downloading it. When calculating investment efficiency, we must ensure the data are strongly balanced panel data. If there are missing values, the calculation will be wrong. Therefore, we did not cover the data for 2023 with more missing values. Several studies have also used the CSMAR database to discuss topics related to green bonds and financing efficiency [3,20,37]. All listed enterprises in China’s A-share market from 2000 to 2022 were selected as the initial research sample, and “financial institutions”, “ST”, and “ST*” samples were excluded. The data were eventually deflated at the 1% and 99% levels to mitigate the effects of potential outliers.

3.1.1. Financing Efficiency Measurement

The three-stage DEA model considers the effects of the external environment and stochastic factors [38].

Stage 1: use an input-oriented BCC−DEA model. $x_{ik}$ denotes the i th input of the k th DMU. $y_{ik}$ represents the r th output of the k th DMU. $x_{ik}$ and $y_{ik}$ are non-negative. Under the assumption of variable returns to scale, the BCC−DEA for evaluating the ${\mathrm{DMU}}_0$ can be described as:

$$\begin{array}{cc}\hfill & min{\theta }_0-ϵ\left(\sum _i^m{s}_{io}^{-}+\sum _r^s{s}_{ro}^{+}\right)\hfill \\ \hfill & \mathrm{s}.\mathrm{t}.\left\{\begin{array}{c}\sum _{k=1}^n{\lambda }_k{x}_{ik}={\theta }_0{x}_{io}-s_{io}^{-},i=1,\cdots ,m\\ \sum _{k=1}^n{\lambda }_k{y}_{ik}=y_{io}+s_{ro}^{+},r=1,\cdots ,s\\ \sum _{k=1}^n{\lambda }_k=1,k=1,\cdots ,n\\ {\lambda }_k>0,s_{io}^{-}>0,s_{ro}^{+}>0\end{array}\right.\hfill \end{array}$$

(1)

$ϵ$ is a minimal positive value used to avoid unbounded linear programming solutions [39]. ${\lambda }_k$ represents the weights of ${\mathrm{DMU}}_{\mathrm{k}}$. $s_{io}^{-}$ and $s_{ro}^{+}$ denote the input and output relaxation of ${\mathrm{DMU}}_0$, respectively. ${\theta }_0$ is the efficiency value of ${\mathrm{DMU}}_0$. If ${\theta }_0=1$ and $s_{io}^{-}=s_{ro}^{+}=0$, the ${\mathrm{DMU}}_0$ is fully efficient. If ${\theta }_0=1$ and $s_{io}^{-}\ne 0$ or $s_{ro}^{+}\ne 0$, then ${\mathrm{DMU}}_0$ is weakly efficient. If ${\theta }_0$ < 1, then ${\mathrm{DMU}}_0$ is inefficient.

Stage 2: the SFA model is applied to adjust the initial inputs, with the input slack calculated in Stage 1 as the explanatory variable $\left(S_{ik}\right)$ and the external environmental factors as the explanatory variables, as explained below:

$$S_{ik}=f\left(Z_k,{\beta }^i\right)+V_{ik}+U_{ik}$$

(2)

$S_{ik}$ is the i th input relaxation time of ${\mathrm{DMU}}_{\mathrm{k}}.Z_k$ is the external environmental factor and ${\beta }^i$ is the coefficient value. $V_{ik}$ and $U_{ik}$ are the random error term and the management inefficiency, respectively. The inputs are then adjusted using the regression coefficients of the SFA model:

$$x_{ik}^a=x_{ik}-\left[max\left\{Z_k{\widehat{\beta }}^i\right\}-Z_k{\widehat{\beta }}^i\right]+\left[max\left(\widehat{V_{ik}}\right)-\widehat{V_{ik}}\right]$$

(3)

$x_{ik}^a$ is the adjusted input. $\left(\widehat{V_{ik}}\right)$ represents the estimate of the random error. max $\left\{Z_k{\widehat{\beta }}^i\right\}$ represents the least efficient DMU. $max\left\{Z_k{\widehat{\beta }}^i\right\}-Z_k{\widehat{\beta }}^i$ is the adjustment to the external environmental factors. $max\left(\widehat{V_{ik}}\right)-\widehat{V_{ik}}$ is the adjustment for random error.

Stage 3: game cross-efficiency DEA identifies each DMU as a participant to reach the Nash equilibrium. $h_{rz}^{*}$ and $w_{rz}^{*}$ are the optimal input and output weights. Therefore, the cross-efficiency value of ${\mathrm{DMU}}_{\mathrm{k}}$, concerning ${\mathrm{DMU}}_{\mathrm{z}}$, can be calculated as:

$$E_{zk}={\displaystyle \frac{{\sum }_{r=1}^s{h}_{rz}^{*}{y}_{ik}}{{\sum }_{i=1}^m{w}_{rz}^{*}{x}_{ik}}}$$

(4)

However, the optimal weights are usually not unique, so cross-efficiency introduces a game perspective. If the initial cross-efficiency of ${\mathrm{DMU}}_{\mathrm{z}}$ is ${\alpha }_z$, the other DMUs try to maximize the efficiency without decreasing ${\alpha }_z$. The linear programming of the game cross-efficiency is:

$$\begin{array}{cc}\hfill & max\sum _{r=1}^s{h}_{rz}^{*}{y}_{ik}\hfill \\ \hfill & \mathrm{s}.\mathrm{t}.\left\{\begin{array}{c}\sum _{i=1}^m{w}_{rz}^{*}{x}_{ik}-\sum _{r=1}^s{h}_{rz}^{*}{y}_{ik}\ge 0\\ \sum _{i=1}^m{w}_{rz}^{*}{x}_{ik}=1\\ {\alpha }_z\sum _{i=1}^m{w}_{rz}^{*}{x}_{ik}-\sum _{r=1}^s{h}_{rz}^{*}{y}_{ik}\le 0\\ w_{rz}^{*}>0,h_{rz}^{*}>0\end{array}\right.\hfill \end{array}$$

(5)

$h_{rz}^{z*}\left({\alpha }_z\right)$ is defined as the optimal solution.

The optimal game cross-efficiency of ${\mathrm{DMU}}_{\mathrm{k}}$ concerning all ${\mathrm{DMU}}_{\mathrm{z}}$ can be obtained by solving the model $\mathrm{n}$ times. Hence, the average game cross-efficiency value of ${\mathrm{DMU}}_{\mathrm{k}}$ is:

$${\alpha }_k={\displaystyle \frac{1}{n}}\sum _{z=1}^n\sum _{r=1}^s{h}_{rz}^{z*}\left({\alpha }_z\right)y_{ik}.$$

(6)

Referring to the method of measuring financing efficiency by Yin et al. [23], the total efficiency (TE) of China’s green bonds is obtained by multiplying the three stages and the indicators shown in

Table 1. Financing efficiency indicator system and measurement of each stage.

	Input Indicators	Output Indicators
Initial Financing Efficiency (IFE)	Finance costs: operating cost	Total financing: financial leasing assets
	Human resource costs: employee salary	Enterprise value growth: total market value of individual stocks
	Consumption of other resources: sales + management expenses
Utilization Efficiency (UE)	Total financing: financial leasing assets	Revenue growth: total operating income
	Financing costs: employee salary + sales + management expenses	Profit growth: operating profit
	Management costs: business and management expenses	Asset growth: total assets growth rate
		Market share: Herfindahl index
		RD output: R&D investment amount
		Operational efficiency: working capital turnover rate
Repayment Efficiency (RE)	Repayment of principal: cash paid to repay debt	Debt service capacity: EBIT/interest expense
	Repayment of interest: debt interest expense	Change in credit rating: credit impairment loss
	Financial leverage: debt-to-asset ratio	Financial stability: financial leverage
		Capital structure optimization: long-term debt-to-working capital ratio
		Long-term solvency: long-term debt ratio

.

3.1.2. Other Variable Measurements

Green bonds: according to Tang and Zhang [12], the independent variable (GB) represents whether enterprise i issued green bonds in period t, taking a value of 1 if yes and 0 if vice versa. The dataset encompasses 1905 bond types, including enterprise, financial bonds, asset-backed securities, and government-related bonds. It excludes non-RMB-denominated floating-rate bonds issued by international institutions, green asset-backed securities, and any private bonds that do not publicly disclose information to ensure the availability of environmental information data and financial data related to bond issuers.

Intermediary variables: the credit rating in annual reports is measured as investor recognition (IR). According to the People’s Bank of China’s “Credit Rating Elements, Symbols and Meanings” and major domestic rating agencies’ credit rating classification methods, the rating range is C− to AAA, a total of 25 levels. This article uses a quantitative method to assign values in sequence, with the lowest level, C−, assigned as 1 and the rating scores increasing in sequence to 25. Financing costs refer to the fees and costs an enterprise must pay when obtaining funds. The following are some commonly used indicators to measure financing costs. The interest rate is one of the most commonly used indicators to measure financing costs. Given the availability of data and the above analysis, this paper uses the sum of interest expenses, fees, and other items in the income statement of listed enterprises divided by the total liabilities as a proxy variable for financing costs (FC) for mechanism testing. The ratio of green patent applications to total annual patent submissions is an indicator of green transformation (GT) for the mechanism test.

Control variables: following Chen et al. [40], Jamil and Khan [41], Meng et al. [42], and Wang et al. [43], select nine variables. The age of an enterprise (Age) is measured by the natural logarithm of its founding date, that is, the period from the date of the enterprise’s founding to the present. The natural logarithm of the operating income of the enterprise measures the operating income of an enterprise (Sale). The book-to-market ratio (PB) is calculated by comparing the enterprise’s book value (i.e., net assets) with its market value. The leverage ratio (Lev) is calculated by dividing the total liabilities by assets. The return on assets (ROA) is measured by dividing the enterprise’s net profit by its total assets, reflecting its relative efficiency in profit generation. The fixed asset turnover rate (Fixed) is measured by dividing the enterprise’s sales revenue by its net fixed assets or the original value of fixed assets. The proportion of institutional investors’ holdings (Inst) is measured by calculating the proportion of the total number of shares held by institutions to the enterprise’s total share capital. The proportion of the largest shareholder’s holdings (Top1) is measured by calculating the proportion of the number of shares held by the largest shareholder to the enterprise’s total share capital. Tobin’s Q (TobinQ) is measured by the ratio of a enterprise’s market value to its total assets. Descriptive statistics are provided in

Table 2. Descriptive statistics.

Variables	Obs	Mean	Std. Dev.	Min	Max
IFE	132,342	0.1601	0.2622	0.0034	0.7689
UE	132,342	0.1792	0.3153	0.0014	0.9116
RE	132,342	0.1913	0.3131	0.0013	0.9105
TE	132,342	0.2091	0.3578	0.0087	1
GB	132,342	0.0063	0.0825	0	1
IR	132,342	13.5805	7.2928	1	25
FC	132,342	0.1433	0.2649	0.0021	0.9893
GT	132,342	0.0611	0.7484	0.0352	66.5001
Age	132,342	0.8810	1.1082	0	4.4598
Sale	132,342	8.7819	10.2041	2.1942	22.0487
PB	132,342	0.2272	0.3184	0.0425	0.8152
Lev	132,342	0.1767	0.3676	0.0174	1.9064
ROA	132,342	0.0272	0.0426	−0.0267	0.2144
Fixed	132,342	1.0064	1.9933	0.0214	7.8265
Inst	132,342	0.8417	2.2837	0.0147	7.3904
Top1	132,342	0.1685	0.1717	0.0197	0.5422
TobinQ	132,342	0.4005	0.6982	0.1224	2.5774

.

3.2. Methodology

To explore the relationship between green bonds and enterprises’ financing efficiency, we construct the two-way fixed effects model:

$$Y_{it}={\alpha }_0+\beta G{B}_{i,t-1}+\gamma Z_{i,t-1}+{\mu }_i+{\lambda }_t+{\epsilon }_{it}$$

(7)

where $Y_{it}$ represents the financing efficiency of enterprise i in each sub-stage and total stage in period t. $G{B}_{i,t-1}$ represents the core explanatory variable (green bond issuance) in period $t-1$. $Z_{i,t-1}$ represents the control variables introduced above in period $t-1$. ${\mu }_i$ denotes the individual fixed effect, ${\lambda }_t$ denotes the time fixed effect, and ${\epsilon }_{it}$ is the random error term.

4. Empirical Results

4.1. Benchmark Regression

Correlation analysis can help us understand the degree of correlation between different variables, thus providing a basis for regression analysis. If the correlation between two variables is strong, they may be suitable for regression analysis. In addition, correlation analysis helps to detect whether there is a high correlation between multiple independent variables, that is, multicollinearity. If multicollinearity exists, regression analysis results may be unreliable, so correlation analysis is needed before regression analysis to eliminate this problem.

Table 3. Correlation analysis.

	IFE	UE	RE	TE	GB	Age	Sale	PB	Lev	ROA	Fixed	Inst	Top1	TobinQ
IFE	1
RE	0.006	0.023	1
TE	0.029 *	0.070 *	0.996 *	1
GB	0.025	0.034 *	0.014	0.014	1
Age	−0.010	−0.022	−0.103 *	−0.108 *	−0.038 *	1
Sale	0.010	0.021	0.050 *	0.048 *	−0.052 *	0.798 *	1
PB	−0.001	0.001	0.049 *	0.048 *	−0.044 *	0.742 *	0.892 *	1
Lev	0.050 *	−0.060 *	0.073 *	0.070 *	−0.046 *	0.697 *	0.899 *	0.810 *	1
ROA	0.010	−0.050 *	0.121 *	0.117 *	−0.042 *	0.539 *	0.789 *	0.659 *	0.842 *	1
Fixed	0.013	0.022	0.008	0.010	−0.046 *	0.611*	0.794 *	0.697 *	0.728 *	0.718 *	1
Inst	−0.015	−0.010	−0.004	−0.006	−0.029 *	0.142 *	0.462 *	0.377 *	0.494 *	0.446 *	0.447 *	1
Top1	−0.003	−0.036 *	−0.04 *	−0.052 *	−0.044 *	0.711 *	0.874 *	0.855 *	0.797 *	0.789 *	0.704 *	0.391 *	1
TobinQ	0.007	−0.001	−0.03 *	−0.030 *	−0.052 *	0.748 *	0.910 *	0.774 *	0.808 *	0.726 *	0.737 *	0.475 *	0.812 *	1

Note: * p < 0.01.

reports the correlation between variables. The results show a specific correlation between most variables, indicating that they should be included in the regression model.

The results of the benchmark regression are presented in

Table 4. Regression results of green bonds and total stage financing efficiency.

	(1)	(2)	(3)
	TE
GB	1.241 *	1.329 **	0.6391 ***
	(0.5221)	(0.6312)	(0.2672)
Age		−0.3274 ***	−0.3982 ***
		(0.0342)	(0.0138)
Sale		0.2938 ***	0.3289 ***
		(0.0252)	(0.0593)
PB		0.3279 ***	0.2229 **
		(0.0277)	(0.0962)
Lev		−0.0731 ***	−0.0287 **
		(0.0124)	(0.0083)
ROA		0.0145	−0.0792 ***
		(0.0358)	(0.0192)
Fixed		0.0012	0.0023 ***
		(0.0011)	(0.0007)
Inst		−0.0299	−0.0021
		(0.0390)	(0.0327)
Top1		−0.0074 ***	−0.0131 ***
		(0.0020)	(0.0018)
TobinQ		−0.0523 **	0.0257
		(0.0197)	(0.0329)
Mean VIF			1.24
Year fixed effects	NO	NO	YES
Individual fixed effect	NO	NO	YES
Obs	126,588	126,588	126,588
$R^2$	0.1673	0.7830	0.9338

Note: * p < 0.10, ** p < 0.05, *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

, where column (1) presents the results on aggregate stage efficiency only, column (2) presents the results with the addition of the control variables, column (3) presents the results with the addition of the year fixed effects and the enterprises’ individual fixed effects. In the regression model, if there is a high correlation between variables (that is, multicollinearity), it may lead to inaccurate estimation of model parameters and increase the standard error, thereby affecting the model’s explanatory power and predictive power. Through the collinearity test, multicollinearity problems can be identified and handled. The results are shown in Table 4. The mean VIF is still less than 5, indicating that all variables selected in this article do not have serious multicollinearity problems.

The estimated coefficient of 0.6391 in column (3) of Table 4, significant at the 1% level, demonstrates that green bonds dramatically augment finance efficiency throughout the total stage. Green bonds attract investors who prefer pro-environmental investments, thereby broadening the investors. Furthermore, green bond enterprises, like government departments and financial institutions, often benefit from preferential policies, and so may offer advantages such as preferential interest rates, reduction or waiver of handling charges, or provision of additional loan guarantees and financing support that encourage participation in green projects [44], thus improving financing efficiency. Therefore, Hypothesis 1 is verified.

The magnitude of the estimated coefficients of the control variables is also approximately the same as expected in previous studies. As a company ages, it may face debt burden, performance, and asset quality and market share challenges, leading to decreased financing efficiency. The higher the operating income is, the more assets and cash flows a company usually has, which can be provided to financial institutions and investors as collateral, thereby increasing the creditworthiness and reliability of the enterprise’s financing. Moreover, enterprises usually need to bear a larger business scale and market demand. In order to meet these needs, enterprises need more financial support, thereby increasing the enterprise’s financing scale and financing needs and improving the enterprise’s financing efficiency level. The effect of the book-to-market ratio is not significant, but the sign of the coefficient is positive. High leverage may lead to increased financial pressure, such as the pressure to repay debt interest and principal, thereby reducing the efficiency of a company’s financing repayment phase. The effect of return on assets is not significant, but the sign of the coefficient is positive. A high Tobin’s Q may reflect the market’s overly optimistic expectations for its future profitability, making it easier for enterprises to fall into high valuation difficulties when financing, and financing costs may be high. It may also mean that enterprises are overinvesting in specific projects or areas, resulting in less efficient financing.

Table 5. Regression results of green bonds and total stage financing efficiency.

	(1)	(2)	(3)
	IFE	UE	RE
GB	2.1790 ***	−0.7328	1.3946 ***
	(0.4038)	(0.5762)	(0.3220)
Control variables	YES	YES	YES
Year fixed effects	YES	YES	YES
Individual fixed effect	YES	YES	YES
Obs	126,588	126,588	126,588
$R^2$	0.4279	0.4782	0.5517

Note: *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

reports that the estimated coefficient of GB on IFE is 1.0063 at the 10% significance level, indicating that green bonds can significantly improve the efficiency level of enterprises’ financing in the financing stage. The emergence of the green bond market provides an additional financing channel for enterprises. Traditionally, businesses have raised capital through bank loans, stock issuances, etc., but these methods may have limitations or high costs. The green bond market enables enterprises to raise funds through the issuance of bonds, thus broadening financing channels and increasing financing flexibility.

The estimated coefficient of GB on UE does not pass the significance test, and the estimated coefficient is negative, indicating that green bonds cannot improve the efficiency level of enterprises in the financing use stage. Funds raised by green bonds are often used for specific green projects, but the specific flow and use of funds can be challenging to track and monitor [45]. Enterprises lack transparency and accountability mechanisms during the financing use stage, resulting in funds that may be misused or not effectively used for environmental protection purposes. Enterprises face complex issues when selecting and screening green projects, such as project risk assessment, investment return forecast, etc. Suppose enterprises have biases in project selection or inaccurate assessments. In that case, it may result in the inability to achieve the expected benefit improvement during the financing and use stage of green bonds.

The estimated coefficient of GB on RE is 0.5836 at the 1% significance level, indicating that green bonds can significantly improve the efficiency level of the enterprises’ financing repayment stage. Issuing green bonds can bring in new groups of investors, especially those who pay more attention to social responsibility and environmental issues. These investors are generally more inclined to support environmentally friendly and sustainable development projects and, therefore, have a higher interest in green bondholders. As investor demand increases, the success rate of enterprises issuing green bonds will also increase. In this way, enterprises can obtain more funds, thereby improving the efficiency level of financing repayment. In addition, issuing green bonds can promote enterprises’ strategic transformation and business innovation. In order to meet the issuance criteria of green bonds, enterprises need to carry out environmentally friendly and sustainable development projects. These projects may include innovations in energy conservation, emission reduction, clean energy, circular economy, etc. [37]. Implementing these projects helps enterprises improve their environmental protection levels and promotes their strategic transformation and business innovation, thereby improving the efficiency of financing repayment.

4.2. Endogeneity Issues

The endogeneity problem may challenge the conclusion that green bonds positively affect enterprises’ financing efficiency. The sources of endogeneity in this paper may be reverse causality. Enterprises themselves have higher (lower) financing efficiency and are more (less) likely to obtain green bond issuance qualifications or tend to issue (not issue) green bonds. That is, the level of financing efficiency determines the decision of enterprises whether to issue green bonds or not, improving financing efficiency. We use the instrumental variable (IV) method and explain the reasons for selecting IV to alleviate the problem of reverse causality. We construct an instrumental variable for green bond issuance by other enterprises ($GBOT$), a dummy variable for whether other bond issuers in the same industry choose to issue green bonds after one enterprise issues green bonds. Regarding operations, enterprises in the same industry are relatively similar regarding business scope, product structure, and daily operations [46]. Therefore, the profit model of enterprises in the same industry is consistent, which determines that the changes in the credit risk of enterprises in the same industry will show robust synchronization [47]. The above behaviors lead to synchronizing performance returns and risk changes among enterprises in the same industry, providing an objective and realistic basis for spreading green signals. Such similar behaviors will lead investors to infer that enterprises that have not issued green bonds will have the same or similar behaviors as enterprises in the same industry that have issued green bonds. When enterprises issue green bonds and send green signals to the market, other enterprises in the same industry may make more strategic decisions conducive to environmental protection. Therefore, $GBOT$ is related to enterprises’ issuance of green bonds, but it is unlikely to affect the financing efficiency. It only affects the financing efficiency through the similarity of other enterprises in the same industry issuing green bonds, which meets the requirements of correlation and homogeneity of instrumental variables. Columns (1) and (2) in

Table 6. Endogeneity test.

	(1)	(2)	(3)	(4)	(5)
	GB	TE	IFE	UE	RE
	IV Stage 1	IV Stage 2
GBOT	7.9324 ***
	(0.3989)
GB		2.1248 ***	4.3294 ***	−4.2527	5.9483 ***
		(0.0324)	(0.7893)	(2.1874)	(0.2366)
Control variables	YES	YES	YES	YES	YES
First stage F value	86.79
Unidentification identification test		0.017	0.029	0.081	0.004
Weak identification test		86.788	85.987	86.469	87.219
Year fixed effects	YES	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES	YES
Observations	132,332	132,332	132,332	132,332	132,332
R-squared	0.2965	0.6783	0.8826	0.7380	0.8471

Note: *** p < 0.01. Cluster robust standard errors in parentheses. The unidentification identification test reports the p value, and the weak identification test reports the F value. Both the dependent variable and the control variables are first-order lag terms.

report the regression results of the two-stage least squares method (2SLS). Column (1) shows that $GBOT$ passes the joint test of excluding instruments and supports its correlation with GB (first stage F value > 10). Column (2)–(5) show that the F value of the weak identification test is greater than the critical value of 16.38 of 10% bias in the Stock–Yogo weak ID test critical values. This result shows that the instrumental variables selected in this paper do not have the problem of unidentification and weak identification. The regression result of GB is still significantly positive, consistent with the benchmark result.

4.3. Robustness Tests

A robustness test is performed to verify the accuracy of the results above. We need to make several modifications to perform the robustness tests presented in

Table 7. Robustness tests.

	(1)	(2)	(3)	(4)	(5)	(6)
	TE	FS	TE	TE	TE	TE
GBD	0.0261 ***
	(0.0028)
GB		0.2874 ***	0.1288 ***	0.7219 ***	1.3627 ***	0.9294 *
		(0.4181)	(0.1623)	(0.5859)	(0.2339)	(0.4260)
Control variables	YES	YES	YES	YES	YES	YES
Year fixed effects	YES	YES	YES	NO	NO	YES
Individual fixed effect	YES	YES	YES	NO	YES	NO
Year–individual fixed effect	NO	NO	NO	YES	NO	NO
Obs	126,588	126,588	126,588	126,588	126,588	126,588
$R^2$	0.1980	0.5293	0.2071	0.2539	0.1674	0.2018

Note: * p < 0.10, *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

. First, green bond underwriting intensity (GBD) replaces the independent variable green bond issuance (column 1); enterprises’ financing size (FS) replaces the dependent variable financing efficiency (column 2); industry clustering of green bond issuance is mitigated by the exclusion of the construction and electricity, and heat industry samples (column 3); and fixed effects are substituted with fixed effects of year–individual interactions, capturing the characteristics of individuals over time, thus mitigating different enterprises’ potential policy impacts at different times (column 4). We then tested the regression results that included individual fixed effects and time effects (columns (5) and (6)). The results consistently show that the coefficient of green bonds on aggregate stage efficiency is significantly positive and robust.

4.4. Impact Mechanism Testing

Table 8. Investor recognition mechanism test.

	(1)	(2)	(3)	(4)	(5)
	IR	TE	IFE	UE	RE
IR		0.6238 ***	−0.0166	−0.3290	5.0275 ***
		(0.1208)	(0.0284)	(0.2769)	(1.0381)
GB	0.0583 ***	0.6394 ***	1.3482 **	−1.6349	5.6348 ***
	(0.0213)	(0.1830)	(0.4927)	(1.3902)	(0.9347)
Control variables	YES	YES	YES	YES	YES
Year fixed effects	YES	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES	YES
Obs	126,588	126,588	126,588	126,588	126,588
$R^2$	0.1893	0.2388	0.4582	0.5297	0.3329

Note: ** p < 0.05, *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

shows the impact of green bonds on enterprises’ investor recognition. Column (1) shows that green bonds can significantly improve the investor recognition of enterprises. Columns (2)–(5) show that green bonds can positively promote overall and repayment stage financing efficiency by improving the investor recognition of enterprises. The impact on other stages is not significant. The issuance of green bonds needs to meet specific environmental standards and disclosure requirements, which makes it easier for investors to understand the use of funds and environmental benefits. Investors who are concerned about the environment and sustainable development are more willing to invest funds in projects with environmentally friendly attributes, thereby improving the company’s investor recognition. The green bond market has attracted investors who are specifically concerned about the environment and sustainable development, such as environmental funds, socially responsible investment funds, etc. [48]. These investors may differ from investors in the traditional bond market in that they are more willing to support and invest in environmental protection projects [26]. Therefore, enterprises can attract more investor groups by issuing green bonds, expanding financing channels, and improving financing efficiency. Hypothesis 2 is verified. However, the impact of green bonds on other stages may be relatively small. The reason may be project selection restrictions. Green bond funds must be used for projects that meet environmental and sustainable development standards, which may impose certain restrictions on the company’s project selection range. Not all projects meet the requirements of green bonds, so non-environmentally friendly projects in other stages may not be able to enjoy the financing advantages of green bonds. Furthermore, there may be a mismatch between supply and investor demand in the green bond market. Although the green bond market has developed rapidly in recent years, it still faces an imbalance between investor demand and available green projects for investment. This may result in some enterprises being unable to take full advantage of what the green bond market offers when financing at other stages.

Table 9. Mechanism tests for financing costs.

	(1)	(2)	(3)	(4)	(5)
	FC	TE	IFE	UE	RE
FC		−0.2371 ***	0.0429	−0.2184	−2.0078 ***
		(0.0343)	(0.2042)	(0.2361)	(0.3789)
GB	−0.0218 ***	0.7324 ***	2.3469 **	−2.2234	6.6749 ***
	(0.0028)	(0.2344)	(0.4368)	(1.6843)	(1.6389)
Control variables	YES	YES	YES	YES	YES
Year fixed effects	YES	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES	YES
Obs	126,588	126,588	126,588	126,588	126,588
$R^2$	0.2381	0.2327	0.2579	0.8743	0.3257

Note: ** p < 0.05, *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

shows the impact of green bonds on enterprises’ financing costs. The results in column (1) show that green bonds can significantly reduce the financing costs of enterprises. Columns (2)–(5) show that green bonds can positively promote the financing efficiency of the total stage and repayment stage by reducing the financing costs of enterprises. The effect on other stages is not significant. Issuing green bonds can reduce financing costs for enterprises because the green bond market usually has a specific investor base, and these investors are more willing to support environmental protection and sustainable development projects. Since investors have higher recognition of these projects, they may be willing to provide financing to enterprises at lower interest rates, thus reducing the financing costs of enterprises [35]. Hypothesis 3 is verified. However, the funds raised by green bonds must be used for projects that meet environmental and sustainable development standards, which may impose certain restrictions on enterprises’ project choices. At other stages, enterprises may need financing for non-environmental projects and are, therefore, unable to take full advantage of the financing cost advantages brought by green bonds. Although the green bond market has developed rapidly in recent years, it still faces an imbalance between investor demand and available green projects for investment. This may result in some enterprises being unable to take full advantage of the financing cost advantages provided by the green bond market when financing at other stages.

Table 10. Mechanism tests for green transformation.

	(1)	(2)	(3)	(4)	(5)
	GT	TE	IFE	UE	RE
GT		2.1421 **	−0.3252	1.3255 ***	1.7345 ***
		(0.2145)	(0.3658)	(0.2674)	(0.1283)
GB	1.3523 ***	0.1453 ***	1.7345	0.2054 ***	0.3783 ***
	(0.2551)	(0.0457)	(1.5431)	(0.0924)	(0.1200)
Control variables	YES	YES	YES	YES	YES
Year fixed effects	YES	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES	YES
Obs	126,588	126,588	126,588	126,588	126,588
$R^2$	0.6372	0.3255	0.7945	0.7047	0.7844

Note: ** p < 0.05, *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

examines the impact of green bonds on enterprises’ green transformation, by which green bonds can positively contribute to financing efficiency in phases additional to the financing phase. By implementing clean technologies and adopting energy-saving and emission-reduction measures, enterprises can reduce environmental pollution and resource consumption, lower environmental and legal regulatory risks, and promote a green transformation that makes them more stable and sustainable, thus increasing investors’ confidence and improving financing efficiency [31]. Hypothesis 4 is verified.

The small scale of the green bond market in developing countries and traditional high-carbon sectors results in limited market demand and investor recognition in these sectors, thereby shrinking enterprises’ ability to raise capital through green bond issuance. Given the project-specific nature of green bond financing, enterprises must ensure that the green projects they select are feasible and profitable enough to attract investor interest and trust. Consequently, enterprises that deviate significantly from traditional business models may encounter challenges in project selection, potentially restricting the ability to leverage green bond financing effectively.

5. Heterogeneity Analysis

According to the difference between state-owned and private enterprises regarding property rights, the impact of green bonds issued by enterprises on their financing efficiency may differ. Therefore, this paper conducts interactive regression based on the unique characteristics of listed enterprises.

Table 11. Heterogeneity analysis: business ownership.

	(1)	(2)	(3)	(4)
	TE	IFE	UE	RE
$GB\times SOE$	0.7347 ***	0.1582 **	0.8321 ***	5.6372 ***
	(0.2378)	(0.0393)	(0.0439)	(1.2810)
SOE	0.7941	−0.0774	0.2248 *	3.0875
	(0.7805)	(0.1535)	(0.1348)	(2.5419)
GB	0.7205 *	0.1589 *	0.7328 *	1.0811
	(0.3239)	(0.0701)	(0.3881)	(1.4376)
Control variables	YES	YES	YES	YES
Year fixed effects	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES
Obs	126,588	126,588	126,588	126,588
$R^2$	0.3255	0.7382	0.2389	0.8326

Note: * p < 0.10, ** p < 0.05, *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

shows that the coefficients on $GB\times SOE$ are significantly positive at the 5% confidence level, suggesting that SOEs contribute more considerably to financing efficiency at any stage than private enterprises. The promotion effect of state-owned enterprises is more significant than that of private enterprises. The following is the most likely explanation. Firstly, state-owned enterprises are usually controlled or directly managed by the government and enjoy government support and endorsement. This gives investors greater confidence and reliability, allowing state-owned enterprises to raise funds at lower interest rates [49]. As the endorser, the government promises to be responsible for the supervision and risk management of the debts of state-owned enterprises. Secondly, state-owned enterprises have a monopoly or essential market share in specific industries or fields and possess abundant resources and assets, which allow state-owned enterprises to provide more collateral or guarantees and improve the asset quality of bonds, thereby reducing financing costs. Finally, state-owned enterprises usually have long-term and stable development plans and a higher sense of responsibility for environmental protection and sustainable development. This makes it easier for state-owned enterprises to meet green bond issuance requirements, such as raising funds for environmentally friendly projects. At the same time, investors are more willing to support enterprises with clear environmental strategies, increasing the attractiveness of state-owned enterprise green bonds. In addition, the estimated coefficients on the financing efficiency in the financing, use, and repayment stages gradually increase. As society pays more attention to environmental issues, investors’ demand for green projects may increase. If state-owned enterprises can meet this demand, the possibility of issuing green bonds will also increase. In this case, the market recognition of green bonds increases, and investors’ preference for them increases, thereby reducing financing costs and improving financing efficiency in the financing stage. If state-owned enterprises can demonstrate superior environmental performance during project use and repayment phases, investors’ trust in them may increase. This will help reduce financing costs in subsequent stages and improve financing efficiency in the use and repayment stages.

The green bonds issued by high-polluting and non-high-polluting enterprises may differ, leading to different impacts on financing efficiency. Because of this, this article conducts interactive regression, as shown in

Table 12. Heterogeneity analysis: nature of polluting enterprises.

	(1)	(2)	(3)	(4)
	TE	IFE	UE	RE
$GB\times Pol$	0.9327 ***	1.2863 ***	0.8376 **	1.3829 **
	(0.1773)	(0.2013)	(0.2760)	(0.2748)
Pol	−0.0869 *	−0.0392 *	0.0644	0.3705
	(0.0359)	(0.0154)	(0.0549)	(0.6312)
GB	0.7153 *	0.7636 **	0.6420 ***	1.5560 **
	(0.2916)	(0.2685)	(0.1258)	(0.5492)
Control variables	YES	YES	YES	YES
Year fixed effects	YES	YES	YES	YES
Individual fixed effect	YES	YES	YES	YES
Obs	58,300	68,288	65,978	60,610
$R^2$	0.6372	0.6919	0.8027	0.8956

Note: * p < 0.10, ** p < 0.05, *** p < 0.01. Cluster robust standard errors in parentheses. Both the dependent variable and the control variables are first-order lag terms.

. The $GB\times Pol$ coefficient of highly polluting enterprises is significantly positive at the 10% confidence level. The results show that, compared with non-polluting enterprises, the financing efficiency of highly polluting enterprises increases significantly after issuing green bonds. The possible reasons are as follows. Firstly, highly polluting enterprises improve their image and social reputation by issuing green bonds [43]. These businesses may face public and stakeholder concerns and criticism about their environmental impacts. By investing in environmentally friendly projects and demonstrating their commitment to environmental sustainability to the public, these businesses can create a more positive image and win public approval. This improved image helps attract investors’ trust and improves financing efficiency. Secondly, highly polluting enterprises require significant capital investment for environmental improvements. Issuing green bonds can provide these enterprises with an additional source of funding to support their environmental projects and sustainable development plans [50]. In this way, highly polluting enterprises can better meet regulations and industry standards, and gradually achieve sustainable operations. This commitment to sustainability helps improve financing efficiency.

6. Conclusions and Implications

6.1. Conclusions

Green development and enterprises’ financing efficiency are critical issues in sustainable economic development. China has also introduced a series of green financing support policies and green financial innovation measures to provide policy support and institutional guarantees for enterprises’ financing. China faces severe environmental challenges as one of the countries with the world’s most significant carbon emissions and primary energy consumption. China has adopted a series of sustainable policy measures, including promoting the development of green finance and introducing green bonds. In the current context, green bonds are one of the critical tools of green finance. This paper explores the influence of green bonds on the financing efficiency of China’s A-share listed enterprises. Analyzing data from 2000 to 2022 using a three-stage DEA model, it assesses efficiency across different financing phases. We find a significant positive correlation between the issuance of green bonds and improved corporate financing efficiency. This finding is consistent with the conclusions of some previous studies [12,13]. However, our study further reveals how green bonds affect financing efficiency by reducing debt costs and improving corporate reputation in the specific context of China’s A-share market. The empirical analysis results of this study provide new evidence on the role of green bonds in Chinese corporate financing, especially in the current context of growing global attention to green finance. As confirmed by robustness tests, the findings reveal that green bonds significantly boost financing efficiency, particularly during the total and repayment stages. Green bonds improve financing by affecting investor recognition, financing costs, and the green transformation of enterprises. Specifically, they enhance total stage financing efficiency through investor recognition, and positively improve the total and repayment stages through reduced financing costs. Furthermore, except in the fundraising phase, green bonds facilitate efficiency in other stages through green transformation. Moreover, interactive regression analysis of state-owned and high-polluting enterprises shows that green bonds significantly enhance their financing efficiency, likely due to their resources, market dominance, and environmental commitments. These insights suggest policy directions for China’s green bond financial policies and enterprises’ investment strategies.

6.2. Policy and Managerial Implications

Based on the above research conclusions, the following policy implications can be drawn:

(1) The government should continue to support the development of green finance and green bond markets, including increasing policy support, improving market transparency and standardization, and encouraging innovation and international cooperation. Moreover, it needs to further improve and strengthen policies related to the green bond market, including preferential tax policies, fiscal subsidies, and incentives, to promote the development and growth of the green bond market. The government should also strengthen green financial innovation. Green bonds should receive more policy support and encouragement as an essential financing tool and a mechanism to promote sustainable development. The government needs to further increase its support for green finance, provide more policy and tax incentives, encourage enterprises to issue green bonds, and promote the healthy development of the green financial market.

(2) In the start-up stage, enterprises should pay more attention to the display of innovation capabilities and team background. They should focus on improving financial performance and market expansion capabilities in the growth and maturity stages. It is recommended that enterprises reduce capital costs through government subsidies, angel investment, etc., in the early stages of growth, optimize capital structure, and reduce financing costs by issuing bonds and listing in the mature stage. In addition, enterprises should seek support from the government and non-governmental organizations in the early stages, and use green subsidies and technical support. In the later stage, cost advantages should be achieved by improving operational efficiency and investing in green technologies, formulating effective debt management strategies, and prioritizing repaying high-interest debts.

(3) The government should enhance investors’ recognition of green bonds. Investor recognition is one of the essential mechanisms affecting green bond financing efficiency. Investors should pay attention to the development of the green bond market and understand the characteristics and investment risks of green bonds to allocate their assets more effectively and contribute to sustainable development. The government needs to increase awareness and understanding of green bonds through publicity and training, improve investors’ acceptance and trust, and attract more funds to flow into the green finance field. The government should reduce financing costs and promote green transformation. Research shows that green bonds can reduce enterprises’ financing costs and promote enterprises’ green transformation. The government needs to optimize financial policies, guide banks and financial institutions to provide more green financial products and services, reduce enterprises’ financing costs, encourage enterprises to increase green investment, and promote sustainable development. At the same time, to address the issue of green bonds having an insignificant effect on resource use efficiency, the government and regulatory authorities are able to strengthen the supervision and transparency of enterprises’ fund use and establish a more stringent accountability mechanism.

(4) The government should also focus on supporting state-owned and high-polluting enterprises to issue green bonds. Research results show that the issuance of green bonds by state-owned enterprises and highly polluting enterprises significantly improves financing efficiency. The government needs to formulate corresponding policies to encourage and support these enterprises in issuing green bonds, guiding them in promoting environmental protection and green transformation, and promoting sustainable economic development. Relevant enterprises should strengthen environmental management and technological innovation, improve their sustainable development capabilities, and actively explore the potential of green bonds as a financing tool to improve financing efficiency.

6.3. Limitations and Suggestions for Future Research

A possible limitation of this paper is that the study may be limited to Chinese A-share listed companies, which may limit the generalizability of the results. The Chinese A-share market may have unique characteristics, such as regulatory environment, investor behavior, and market structure, which may differ from other global markets. In addition, the availability and quality of data may affect the study’s accuracy. Future research can improve the robustness of the study by expanding the sample range, considering longer time series data, or adopting different models and methods.