Li Jiawei (2024) [2] analyzed the motives and effects of BYD Co., Ltd.’s (Build Your Dream Corporation Limited) green bond financing based on an investment philosophy that comprehensively considers environmental, social, and corporate governance factors from an ESG perspective. The analysis results show that BYD has issued green bonds for two consecutive years due to environmental benefits, green brand image, and alleviation of mismatched investment and financing periods. At the environmental level, BYD has reduced the emission density of greenhouse gases through green projects, thereby reducing the harm of production activities to the ecological environment. At the level of corporate governance, BYD closely follows the green policy orientation and occupies an important position in the new energy market, effectively reducing risks in supply chain management in both supply and demand dimensions. Ye Peijie et al. (2024) [3] used Industrial Bank’s green bonds as an example to analyze the current situation of commercial bank’s green bonds. They studied the risks of Industrial Bank’s green bonds from bond interest rate risk, credit risk, liquidity risk, etc., and proposed risk management strategies for Industrial Bank’s green bonds under the background of big data, thereby guiding commercial banks to establish and improve green credit mechanisms and enhance environmental investment levels. Yao Haifeng (2024) [1] took Beijing Bank as an example and proposed a complete set of paths and prevention strategies for the impact of green bonds on bank risks. There are still many literature on similar case studies, for example, Yang Minglu et al. (2024) [4] studied the case of BAIC Green Debt; Du Limin and Lu Yang et al. (2024) [5] take the “23 TianYing GK” bond as an example.

Green financing activities in the environmental protection industry cover project investment, financing, operation, and risk management services in the fields of environmental protection, energy conservation, clean energy, green transportation, and construction. Feng Wei (2024) [6] believes that with the increasing severity of global environmental problems and the growing popularity of sustainable development concepts, the funding demand for the environmental protection industry continues to grow, especially in areas such as clean energy, green transportation, and green buildings. As the investment scale of projects continues to expand, the demand for green financing is also increasing. Liu Lijun and Shen Cuiwei (2024) [7] used a multi time difference in differences model to examine the impact of green bond issuance on the debt financing scale of real enterprises based on data from listed companies from 2018 to 2023. The research results indicate that issuing green bonds has expanded the scale of debt financing for real enterprises; In terms of mechanism, the issuance of green bonds has improved the ESG performance of enterprises, played an intermediary role in ESG, and expanded the debt financing scale of physical enterprises; The effect of issuing green bonds to expand the scale of corporate debt financing has heterogeneity in corporate property rights, which is significant in state-owned enterprises but significantly reduced in non-state-owned enterprises; The effect of issuing green bonds to expand the scale of corporate debt financing has heterogeneity in enterprise size, which is significant in large-scale enterprises but not significant in small-scale enterprises.

Zekai Şenol, et al. (2025) [8] employed a series of methods, including quantile-on-quantile return connectedness (QQC), cross-quantilogram (CQ), and Granger causality in quantiles (GC), which allow us to make inferences about the safe-haven ability of green bonds. Their results included that the connectedness at directly related quantiles is more pronounced than those of inversely related quantiles for all cases; The degree of return connectedness is highest with conventional bonds, followed by gold and stocks; Green bonds act as the net transmitter of spillover shocks to Bitcoin and gold markets, whereas it serves as the receiver of shocks from the remaining markets; Their CQ results indicate that positive returns in green bonds significantly follow negative returns in conventional bonds 5 days later, suggesting that green bonds have a one-week lagged safe-haven ability. Their GC from green bonds is bidirectional and mostly intensified in extreme quantiles with oil, cryptocurrencies, and conventional stocks, but unidirectional and visible with Islamic stocks during market downturns.

The existing literature on the impact of green bonds on corporate value is mainly based on empirical analysis of data from listed companies, and most of them have concluded that they have a positive impact. Wang Qian et al. (2021) [9] used regression models to study mixed panel data of listed companies issuing green bonds, and concluded that issuing green bonds can have a positive impact on company value, and the proportion of green bonds has a stacking effect on value enhancement; Ma Yaming and Hu Chunyang (2020) [10] concluded that issuing green bonds can dynamically enhance a company’s value based on the DID model; Wu Yuhui et al. (2020) [11] used bond issuing companies from 2009 to 2019 as a sample and found that issuing green bonds can generate good industry spillover effects; Lu Xuechun (2022) [12] constructed a model using PSM method and also concluded that issuing green bonds has a significant effect on increasing the company’s value. At the level of financing constraints, scholars pay their attention to the impact mechanism on company value through financing constraints. Ning Jinhui et al. (2021) [13] found that green bonds reduce maturity mismatches by alleviating financing constraints based on the basic principle of maturity matching; Huang Huaji et al. (2022) [14] used companies that issue green bonds and ordinary bonds as sample data to measure the operational efficiency of listed companies through the DEA-BCC model. At the same time, they analyzed the role of financing constraints using mediation effects and found that green bonds can improve the operational efficiency of companies by reducing financing constraints, promoting corporate transformation and upgrading, and reducing operating costs; Wang Qian et al. (2021) [9] also found through empirical research that green bonds have a mediating effect of increasing company value by reducing financing constraints.

In terms of information transparency, Qi Huaijin (2021) [15] used “green” labeled bonds issued within China from 2016 to 2019 as the research object to examine the impact mechanism of green premiums, and found that information disclosure is an important channel affecting green premiums; Li Hongwei (2017) [16] used the mediating effect to test the transmission law of media supervision affecting company value through information disclosure level, revealing the transmission path of “media supervision environmental information disclosure company value”. Lu Ahui et al. (2021) [17] also found a positive relationship between information transparency and company value.

Against the backdrop of global efforts to address climate change and environmental challenges, green finance, as an important tool for promoting sustainable economic development, is increasingly receiving attention from governments and financial institutions around the world. Since the reform and opening up, the industrialization process has accelerated, and people have begun to realize that there is an undeniable contradiction between ecological protection and economic development, making environmental pollution a focus of public attention. Nowadays, the automotive industry urgently needs to invest in the research and development of new energy technologies. In its development process, green bond financing has gradually gained attention and favor from the industry. Due to the fact that China only officially issued green bonds in 2016, the research time on green bonds in China is relatively short, and the quantity and content are far behind those related to ordinary bonds and stocks. Existing literature mainly focuses on empirical analysis and financial institutions, lacking case studies on non-financial institutions, especially clean energy companies, issuing green bonds. Secondly, it has been fully demonstrated that issuing green bonds can increase a company’s value, but most companies measure their value based on stock prices. Considering the dual role of the capital market and corporate debt, Tobin’s Q is more suitable for measuring company value to eliminate the impact of stock market fluctuations; In addition, there is a lack of research on the combination of corporate value and its impact mechanism, and the relationship between green bond intensity and corporate value is relatively scarce. On the basis of the above literature, in this paper, we firstly used Tobin’s Q to measure company value, and uses panel data to explore the impact of issuing green bonds on company value through propensity score matching method, as well as whether the issuance of green bonds and the increase in intensity have a significant effect on company value; Secondly, by analyzing the mediating effect on the impact mechanism of company value, efforts are made to discover the potential of green bonds, that is, to what extent they can enhance company value by alleviating financing constraints and improving information transparency.

In this paper, we choose the propensity score matching model to do the test of whether to issue green bonds and its intensity on the company value. Firstly, we screen the companies listed in A-share and issued green bonds from 2014 to 2023, totaling 272 companies. The selected data are treated as follows: (A) Excluding listed companies in the financial industry; (B) Excluding companies containing missing variables; (C) Excluding delisted companies. Finally, 82 companies remain. At the same time, A-share listed companies with the same year and industry as the above 82 companies but without issuing green bonds are selected, and 1208 companies are left after being sorted out according to the above treatment.

All the selected company financial variables are summarized in Table 1 :

The results of descriptive statistics for all variables in the treatment group are shown in Table 2 .

As shown in Table 2 , the mean and median of the financial data of all the selected companies are controlled within a reasonable range, and the outliers have less influence on the data. This indicates that most of the companies issuing green bonds are operating well and meet the study’s requirements.

The balancing assumption and the common support assumption are the two basic assumptions of PSM, which need to be tested to ensure rational analysis. Due to the fact that propensity score matching can achieve balance between the treatment group and the control group on multiple covariates through the matching process, thereby improving the accuracy and reliability of statistical estimation, it is chosen for this section.

As shown in Table 3 , the result of the PSM balance test. None of the T-tests for the means of the control variables after matching are significant, which solves the problem of significant deviation. The result of the PSM balance test indicates the null hypothesis is accepted which indicates the differences between the groups are not significant and the variables are balanced in the treatment and control groups after matching, in line with the balancing assumption of the PSM.

Before matching, the probability of a firm’s decision to issue green debt, i.e., the propensity score of issuance of green debt, needs to be estimated. Whether a company issues green bonds is a logistic variable, so this paper uses the Logit model to estimate the propensity score of issuance of green bonds and intensity of green bond issuance, and the estimated model is shown below:

$p_i=p\left(Y=1|X_i\right)$, $Y=Gre/GST$ (3.1)

where $p_i$ is the propensity score of issuing green debt; $X_i$ is the control variable that influences both the company’s issuance of green debt and company value; $Gre$ is whether to issue green bonds or not, and $GST$ is the intensity of green debt issuance.

Note: *, **, and *** indicate significance at the 10%, 5%, and 1% significance levels, respectively. The same applies to all subsequent tables.

As shown in Table 4 , the control variables exhibit statistical significance at specific levels regarding the probability of issuing green bonds and the intensity of green bond issuance, which indicates the logit model can explain the propensity scores of whether to issue green bonds and the intensity of issuance well.

Figure 1 and Figure 2 represented the histograms of propensity scores of the treatment and control groups before and after matching, respectively. As shown in Figure 1 and Figure 2 , before matching, there is a significant difference of the propensity scores distribution between the treatment and control groups. However, after matching, there is almost no difference between the two groups, which indicates the samples of the treatment group are matched in the control group, in line with the common support assumption of PSM.

Based on the Logit model and balancing assumption test, a 1:1 matching method is employed to match companies in the control group with those in the treatment group that have similar propensity scores. The matching results are listed as follows:

As shown in Table 5 , ATT is greater than 0 but not significant before matching. ATT is 0.0269 after matching, and the issuance of green bonds leads to an increase in Tobin’s Q growth rate by 0.0269 under the same propensity score, which indicates the issuance of green bonds increases the company value.

As shown in Table 6 , ATT is greater than 0 but not significant before matching. ATT is 0.0487 after matching, and Tobin’s Q growth rate of companies with high green bond issuance intensity is higher than that with low issuance intensity by 0.0487 under the same propensity score, and thus an increase in issuance intensity contributes to the value of the firm.

We constructed a basic linear regression model by using the data obtained after matching based on whether issuing green bonds or not:

$Q\text{\_}in{c}_{it}={\beta }_0+{\beta }_{1}Gr{e}_{it}+{\beta }_2{X}_{it}+T_t+In{d}_i+{\epsilon }_{it}$ (3.2)

where $Q\text{\_}in{c}_{it}$ is the growth rate of Tobin’s Q. $Gr{e}_{it}$ is the core explanatory variables, $X_{it}$ are a set of control variables, the basic linear regression model controls for industry ( $Ind$) and year ( $T$).

As shown in Table 7 , Gre’s coefficient is 0.0588 and significant at the 5% level under the fixed year and other variables being controlled, indicating that there is a significant positive correlation between the issuance of green bonds and company value, in line with the results of the previous PSM analysis.

To study whether issuing green bonds increases company value by alleviating financing constraints, an SA index is constructed to measure the extent of financing constraints.

$SA=-0.737Size+0.043Siz{e}^2-0.04Ag{e}^2$ (3.3)

The SA index is negative, and the smaller the SA index, the more severe the company’s financing constraints are.

Then, we constructed the mediating effects of financing constraints:

$S{A}_{it}={\alpha }_0+{\alpha }_{1}Gr{e}_{it}+{\alpha }_2{X}_{it}+T_i+In{d}_i+{\epsilon }_{it}$ (3.4)

$Q\text{\_}in{c}_{it}={\beta }_0+{\beta }_{1}Gr{e}_{it}+{\beta }_{2}S{A}_{it}+{\beta }_3{X}_{it}+T_t+In{d}_i+{\epsilon }_{it}$ (3.5)

As Table 8 shown, Column A presents the result of the regression equation examining the impact of the company’s green bonds issuance on financing constraints, and the coefficient showing significantly positive, indicating that green

bonds issuance effectively alleviates corporate financing constraints. Column B is the regression equation containing the mediating variable financing constraints, and the coefficients are significantly positive, indicating that issuing green bonds can increase company value effectively by alleviating the company’s financing constraints. Observing the bootstrap confidence interval for the mediating variable SA, the interval does not contain zero and the mediating effect is significant.

Constructed the mediating effects of information transparency as follows:

$Tr{a}_{it}={\alpha }_0+{\alpha }_{1}Gr{e}_{it}+{\alpha }_2{X}_{it}+T_t+In{d}_i+{\epsilon }_{it}$ (3.6)

$Q\text{\_}in{c}_{it}={\beta }_0+{\beta }_{1}Gr{e}_{it}+{\beta }_{2}Tr{a}_{it}+{\beta }_3{X}_{it}+T_t+In{d}_i+{\epsilon }_{it}$ (3.7)

As shown in Table 9 , column A presents the result of the regression equation examining the impact of the company’s green bonds issuance on information transparency, and the coefficient is positive, indicating that green bonds issuance effectively enhances the information transparency of the company column B is the result of the regression equation containing mediating variable information transparency, and the coefficients are significantly positive, indicating that issuing green bonds can effectively improve information transparency and thus enhance company value. By observing the bootstrap confidence interval for the mediating variable Tra, we found that the interval does not contain zero and the mediating effect is significant.