Green Credit Policy and Enterprise Green M&As: An Empirical Test from China

Abstract

Green credit is an important financial tool to coordinate the relationship between economic development and environmental protection. The Green Credit Guidelines (GCGs) issued in 2012 comprise the first formal, dedicated green credit policy. To test the effectiveness of the GCGs in green governance, in this study, we use the differences-in-differences (DID) method to test the impact of the implementation of the GCGs on enterprise green mergers and acquisitions (M&As) and further examine the performance of green M&As. The results show that the implementation of the GCGs have significantly promoted the green M&A activities of heavily polluting enterprises, and the promotion effect is more significant in enterprises with poor green innovation ability and enterprises with low financial marketization levels. Further research reveals that green M&As can improve the green innovation performance of enterprises. From the perspective of green M&As, in this paper, we expand the research on the effect of green credit policy, providing a decision-making reference for the promotion and improvement of subsequent green credit policy.

1. Introduction

As an important financial tool with which to coordinate the relationship between economic development and environmental protection, green credit has potential to guide enterprises to eliminate the traditional development path characterized by high pollution and high energy consumption to achieve green transformation and thus realize green, low-carbon and sustainable economic development. In February 2012, the China Banking Regulatory Commission (CBRC) issued the Green Credit Guidelines (GCGs), which marked the standardization of China’s green credit policy. The GCGs require banking financial institutions to effectively identify, measure, monitor and control environmental and social risks in credit business activities and implement differentiated credit policies based on environmental and social risk ratings. On the one hand, credit must be restricted to customers whose environmental and social performance is not in compliance, with increased support for the green economy, low-carbon economy and circular economy. On the other the GCGs introduced changes in terms of loan terms and interest rates. In contrast to administrative measures such as environmental regulation, the GCGs aim to guide the flow of funds to the sustainable development field by adjusting the allocation of credit resources, thereby curbing the blind expansion of heavily polluting enterprises from the source and promoting the achievement of green transformation [1].

Green technology innovation is the core of green transformation. The green transformation path of enterprises can be divided into two categories according to the source of green technology: an endogenous path, such as by increasing R&D investment to develop clean technology and green products, and an exogenous path, such as when enterprises introduce external green technology resources. For example, the green mergers and acquisitions (M&As) studied in this paper represent an important method of exogenous transformation. Green M&As refer to the M&A activities carried out by enterprises to obtain and expand their green competitive advantages driven by the goal of realizing environmental protection and sustainable development [2,3]. Introducing high-quality green assets, green technology and green management experience through green M&As is an important breakthrough for enterprises to achieve green transformation, especially for heavily polluting enterprises.

The existing research on the implementation effect of green credit policy mostly focuses on enterprise-independent innovation. Previous studies have tested the effectiveness of green credit policy in promoting the green transformation of enterprises from the perspectives of environmental protection investment [4] and technological innovation [5,6,7]. However, there is still a lack of evaluation results from the perspective of green M&As, which provides a starting point for this study. Compared with independent innovation with a long R&D cycle and high risk, green M&As can help companies quickly acquire the green resources of target companies and have a significant speed advantage in promoting the green transformation of companies. Moreover, the debt financing constraints of heavily polluting enterprises, especially the long-term financing constraints [8,9] caused by the green credit policy, will weaken the motivation of enterprises to invest in high-risk fields [10], which is not conducive to sustainable green innovation. Under such circumstances, will heavily polluting enterprises seek transformation and upgrading through green M&As under the pressure brought about by the green credit policy, and what will be the final effect? In this paper, we aim to answer the above questions, first of all by exploring the relationship between green credit policy and enterprise green M&As as a whole to make up for the lack of relevant research and, secondly, by carrying out a systematic analysis along the chain of what kind of relationship exists and the heterogeneity of effects and green M&A performance in an attempt to deconstruct the internal logic of green credit policy and green M&As, providing a reference to improve green credit policy and accelerate the green transformation of enterprises. To this end, we use the release of the GCGs to construct a quasi-natural experiment using the differences-in-differences (DID) method to deeply explore whether green credit affects enterprises’ green M&A behavior and further test the performance of green M&As based on changes in the level of green technology innovation.

The possible research contributions of this paper are as follows. First, from the perspective of enterprise green M&As, we expands the research perspective of the microeffects of green credit policy, providing further empirical evidence of the implementation effect of green credit policy and enriching the research on the influencing factors of enterprise green M&As. Second, by studying the direct effects and economic consequences of green credit policy on green M&As of enterprises, this study will help increase the attention of the outside world to the green M&A activities of enterprises and provide ideas and references for enterprises to achieve innovation and transformation through green M&As. Third, the goal of carbon peaking and carbon neutralization introduces requirements for the development of green finance, in which green credit plays a very important role. In this context, by focusing on the economic effects of green credit policies, this paper can provide inspiration for the improvement of the international green financial system and the realization of the goal of carbon peak and carbon neutralization.

The rest of this paper is arranged as follows. In Section 2, we introduce the results of related research and the process of formulating research hypotheses. In Section 3, we introduced the empirical design, including the model setting of this study, data source variable definitions, etc. In Section 4, we present the empirical results and an analysis of the results. In Section 5, we present a discussion, in which we analyze our research findings and compare them with existing research conclusions. In the last section, we summarize the conclusions and policy implications, analyze the shortcomings of this paper and propose future research prospects.

2. Literature Review and Research Hypotheses

2.1. Literature Review

2.1.1. Implementation Effect of Green Credit Policy

Green credit belongs to the category of green finance, also known as environmental financing and sustainable financing. As early as the 1990s, the United Nations Environment Programme issued the Declaration on the Environment and Sustainability of the Financial Industry, pointing out the importance of environmental factors in risk assessment and encouraging the private sector to invest in technologies and services that are beneficial to the environment. The Equator Principles jointly proposed by many banks around the world in 2002 first expounded the concept of green credit. In the “Opinions on Implementing Environmental Protection Policies and Regulations to Prevent Credit Risk” issued in July 2007, China formally proposed the concept of “green credit” for the first time. The document pointed out that cooperation and linkage between environmental protection and financial supervision departments should be bolstered to strengthen environmental supervision, promote credit security and support environmental protection with strict credit management. After that, the Green Credit Guidelines (GCGs) were issued in 2012, which introduced clear requirements for financial institutions implement green credit and further clarified the evaluation index system of green credit implementation. Since then, the scale of green-credit-related business of banks has expanded rapidly.

Based on the original intention of the policy, scholars have conducted valuable tests on the implementation effect of the green credit policy. These studies can be roughly divided into macro and micro levels. Studies conducted at the macro level have confirmed the role of green credit policy in promoting industrial structure upgrading [11,12,13], green economic development [14] and environmental quality improvement [13,15]. The research at the micro level can be divided into two parts. The first part is the impact of the green credit policy on corporate investment and financing activities. The results of this part show that the green credit policy has a financing penalty effect and an investment inhibitory effect on heavily polluting enterprises [8,9]. On the one hand, the green credit policy has raised the threshold for highly polluting enterprises to obtain credit financing, which has reduced their financing scale, especially long-term financing [16]. On the other hand, the green credit policy also considerably increases the debt financing cost of enterprises [17,18], resulting in an obvious financing punishment effect and even causing enterprises to face financial difficulties [19]. Furthermore, in the case of limited financing, many enterprises tend to “cut back” by reducing capital investment [20]. The second part is the green effect of the green credit policy. Some scholars believe that the green credit policy has played a positive role to a certain extent and can guide enterprises to actively participate in environmental governance, promote green technology innovation [5,6,21,22] and provide support for green transformation [23]. Furthermore, many scholars have questioned the effect of green credit policy, believing that the credit constraint effect and cost effect [7] caused by green credit policy will lead to a reduction in R&D investment by highly polluting enterprises [4,24], which is not conducive to technological R&D, transformation and upgrading of enterprises.

2.1.2. Influence Factors and Economic Consequences of Green M&As

The current research on green M&As mainly focuses on their influencing factors and their consequences. According to research on green M&As, there are two types of motivation for enterprises to implement green M&As. First, there is a regulatory motivation to avoid environmental punishment and maintain the corporate image, which is mainly manifested in obedience to environmental laws and regulations, as well as social supervision. Industries that are most affected by environmental regulations are more inclined to implement green M&As [25]. In the face of media pressure and social attention, companies also tend to implement green M&As to build a positive reputation to overcome the crisis of public opinion [26]. The second is the profit motive to pursuing sustainable competitive advantage. In this case, enterprises, starting from their own development strategy, take the initiative to acquire enterprises with technological resource advantages in order to achieve synergies in technology and product development [27].

Scholars have also conducted research on the performance of green M&As at the theoretical and empirical levels. First of all, the positive role of green M&As in improving enterprise economic performance has been affirmed. Green M&As can help polluting enterprises to obtain additional resources, ease financing constraints, reduce tax burdens [28], improve enterprise operating efficiency [29], increase the ROA of the main merging party and promote value creation [3,30]. Owing to its “green” attribute, scholars not only consider the changes in financial indicators before and after an M&A but also attach importance to the changes in its green indicators. However, scholars have not reached an agreement in this regard. Huang and Yuan (2022) [31], Liang et al. (2022) [2] and Lu (2021) [32] all concluded that green M&As can promote green innovation in enterprises based on empirical tests. Pan et al. (2019) [26], on the other hand, came to the opposite conclusion that enterprises may implement green M&As out of instrumental motivation, which is not conducive to green innovation.

2.2. Research Hypotheses

2.2.1. Green Credit Policy and Enterprise Green M&As

Enterprises are embedded in various external environments; thus, their behavioral decisions are inevitably affected by the external environment. The institutional environment is also an important factor affecting enterprise behavioral decisions [33]. As a policy tool to solve environmental problems by relying on market mechanisms, the GCGs change the behavioral choices of microsubjects by guiding the allocation of funds. Specifically, the green credit policy guides the green transformation behavior of enterprises through the following two channels.

First, the green credit policy creates compliance pressure on enterprises through the credit constraint effect, forcing heavily polluting companies to carry out green transformation. Specifically, by setting credit environment access thresholds and implementing differentiated credit policies, i.e., providing preferential loans to green environmental protection enterprises or institutions and imposing loan limits and punitively high interest rates on polluting enterprises and projects, the green credit policy directly links the environmental performance of enterprises with credit financing, which internalizes the negative externalities generated by corporate pollution emissions [9]. Heavy polluting enterprises face increased financing thresholds and financing costs, forcing them to pay attention to their environmental problems and seek green transformation. Second, the green credit policy stimulates the green development awareness of heavily polluting enterprises through the information transmission effect to promote the green transformation of enterprises. The issuance of the GCGs conveys to the market that the government attaches considerable importance to environmental protection and restricts the blind development of heavily polluting enterprises. The guidelines also serve as an early warning to enterprises to accelerate green transformation. Under the influence of these signals, enterprises will also face environmental protection pressure from stakeholders and thus actively seek “greening” approaches.

Green M&As are a superior way for heavily polluting companies to weaken the resource-limiting effect and achieve green transformation. On the one hand, green enterprises have obvious advantages in obtaining credit resources, so heavily polluting enterprises can obtain financing synergies by implementing M&As with them, reducing the resource constraints caused by policies [34,35]. The high market attention of M&As can also play a role in information transmission and release “green signals” to the outside world, thus enhancing investors’ confidence in the future development of enterprises and guiding the flow of scarce resources to the company. On the other hand, introducing high-quality green assets, green technology and green management experience through green M&As is an important breakthrough for heavily polluting enterprises to achieve green transformation. Enterprises can quickly obtain clean energy and green technology to improve production processes, improve production efficiency and reduce pollution emissions. The integration of heterogeneous technologies, talent and management elements of both sides can also produce technological synergy, effectively compensating for the lack of an endogenous driving force of the acquirer, driving the acquirer’s subsequent green technology R&D [31] and ultimately promoting their innovation and transformation.

In summary, the green credit policy forces and motivates enterprises to conduct green M&As through the credit constraint effect and the information transmission effect. Based on the above analysis, the following research hypothesis is proposed:

Hypothesis 1 (H1).

Compared with non-heavily polluting enterprises, the implementation of a green credit policy significantly promotes the green M&As of heavily polluting enterprises.

2.2.2. The Heterogeneity of Green Technology Innovation Capability

Green technology is the driving force and support for the green transformation of enterprises; however, the development of green technology is associate with considerable risks and is highly dependent on the innovation foundation of enterprises. Therefore, the choice of the manner in which to obtain green technology may be constrained by an enterprise’s own green innovation capability. For enterprises with poor green technology innovation ability and a relatively weak innovation foundation, relying on independent R&D, on the one hand, is associated with technical barriers, increasing the risk of R&D failure; on the other hand, even if R&D is successful, enterprises will experience longer periods of R&D. It is difficult to keep up with the speed of technological upgrading [36], and these barriers may even delay the optimal time for enterprise development. Therefore, such enterprises may be more inclined to acquire green technologies from outside through green M&As and further break through organizational boundaries through open innovation models to obtain richer external knowledge to reshape their own technological creation and boost their innovation capabilities [37,38]. Based on the above analysis, we propose the following hypothesis:

Hypothesis 2 (H2).

The promotion effect of the green credit policy on the green M&As of heavily polluting enterprises is more significant in enterprises with poor green innovation ability.

2.2.3. Heterogeneity of the Level of Marketization in the Location of the Enterprise

Green credit policy mainly guides the green behavior of heavily polluting enterprises by imposing credit constraints on them. Corporate credit availability differs in regions with varying levels of marketization; therefore, differences in behavior may occur when faced with restrictions on green credit policy. Heavily polluting enterprises located in regions with a high degree of regional marketization have broader financing channels. They can alleviate credit constraints caused by green credit policies through various financing channels, which may weaken the regulatory role of the guidance to a certain extent and result in a lack of motivation among enterprises to carry out green M&As. Based on the above analysis, Hypothesis 3 is proposed as follows:

Hypothesis 3 (H3).

Compared with regions with a high level of marketization, the implementation of green credit policy has a more significant effect on promoting green M&As of heavily polluting companies in regions with a low level of marketization.

3. Empirical Design

3.1. Data Sources

The research sample in this study consists of the M&A events of Chinese A-share listed companies from 2007 to 2020. With reference to relevant studies [26,31], we processed the samples as follows:

• Kept the sample of the listed company as the acquirer;
• Removed the samples whose transactions failed;
• Excluded the samples whose M&A type is divestiture, asset replacement, debt restructuring or share repurchase;
• Excluded samples whose acquisition target is not equity;
• Excluded samples with an equity acquisition amount less than CNY 1 million or an equity acquisition proportion less than 30%;
• Excluded samples that held more than 30% of the equity of the target before the acquisition; and
• For multiple M&A transactions initiated by the same enterprise in the same year, those with the same subject matter were merged, whereas those with different subject matter only retained the transaction sample with the largest transaction amount.

The M&A data used in this study were drawn from the Wind and M&A databases of CSMAR. Other enterprise-level financial data were also obtained from CSMAR. In addition, to reduce the impact of extreme values, we trimmed continuous variables at the 1% and 99% levels.

3.2. Variable Definition

3.2.1. Definition and Judgment Method of Green M&As

Drawing on the research of Salvi et al. (2018) [3] and Huang et al. (2022) [31], in this paper, green M&As are defined as M&As carried out by enterprises to acquire and expand green elements, such as green technology, green management experience, green equipment and clean energy, as well as M&As for transformation to clean production and the green environmental protection industry.

Based on the above definition, we followed the steps listed below to judge whether an M&A sample is a green M&A. (1) If the target enterprise obtained a green patent within three years before the M&A, then it is regarded as a green M&A. Enterprise green patent data were obtained from the CNRDS green patent research database. (2) With reference to the “Green Industry Guidance Catalogue” and the relevant chapters of green industry segmentation products and services in the “Guidance Catalog of Key Products and Services for Strategic Emerging Industries”, the relevant information published in the M&A announcement, such as background and purpose, the business scope of both parties and the impact on the merger, was comprehensively analyzed to determine whether the M&A meets the definition of green M&A. If it meets this definition, it is considered a green M&A.

3.2.2. Definition of Control Variables

We selected these variables based on the following three aspects. The characteristics of enterprises include property rights nature (Soe), return on equity (Roe), enterprise scale (Size), financial leverage (Lev), cash flow ratio (Cashflow), Tobin’s Q ratio (Tobinq), and employee size (Employee); corporate governance characteristics include duality (Dual), ownership concentration (Top1), board size (Board) and the proportion of independent directors (Indep); and M&A transaction characteristics include transaction consideration (Price), the proportion of acquired equity (Ratio) and payment method (Paymethod). The specific definitions of the variables are shown in

Table 1. Definition of main variables.

Variable	Meaning	Variable Definition
Greenma	Green M&A	Green M&A is recorded as 1; otherwise, it is 0.
Post	Time dummy variable	Post is recorded as 0 for 2007–2011, and 1 for 2012 and later.
Treat	Group dummy variable	Treat is recorded as 1; otherwise, it is 0.
Treat $\times$ Post	Policy effect of the GCGs	Multiplicative term of Treat and Post.
Size	Enterprise size	The natural logarithm of the total assets of the enterprise.
Lev	Financial leverage	End-of-period liabilities/end-of-period total assets.
Roe	Return on equity	Net profit/average balance of shareholder equity.
Cashflow	Cash flow ratio	Net cash flow from operating activities/total assets.
Board	Board size	The logarithm of the number of directors after adding 1.
Indep	The proportion of independent directors	Number of independent directors/number of all directors.
Dual	Combined title of Board Chair and CEO	If the chairman and the general manager positions are held by the same individual, it is equal to 1; otherwise, it is equal to 0.
Top1	Ownership concentration	Shareholding proportion of the largest shareholder
Tobinq	Tobin’s Q ratio	(Market value of tradable shares + face value of non-tradable shares)/total assets.
Soe	Property rights nature	1 for state-owned enterprises and 0 for others.
Employee	Employee size	The logarithm of the number of employees after adding 1.
Price	Transaction consideration	The natural logarithm of the M&A transaction price.
Ratio	Proportion of acquired equity	Proportion of acquired equity.
Paymethod	Payment method	Equal to 1 when the payment method is share-based payment; otherwise, it is equal to 0.

.

3.3. Econometric Model

In this study, we use a logit-based differences-in-differences model to investigate the impact of the GCGs on the green M&A decisions of heavily polluting enterprises. In order to determine whether a fixed-effects model or a random-effects model should be adopted, the Hausman test was conducted; the result revealed a p value = 0.0000, indicating that the fixed-effects model should be used. Accordingly, the model was established as follows:

Greenma_it = β₀ + β₁Treat_i × Post_t + β₂Treat_it + β₃Post_it + ρControls_it + μ_j + ω_t + ε_it

(1)

where Greenma_it refers to whether enterprise i had green M&As in year t; it is equal to 1 if enterprise i had a green M&A in year t and 0 otherwise. Treat_i is a grouped dummy variable, which is equal to 1 in the treatment group and 0 in the control group. In this study, the heavily polluting enterprises that are considerably affected by the guidelines are regarded as the “treatment group”, and other enterprises are regarded as the “control group”. According to the “Guidelines for Industry Classification of Listed Companies” revised by the China Securities Regulatory Commission (CSRC) in 2012, the “List of Listed Companies’ Environmental Protection Inspection Industry Classification Management List” and “Guidelines for Environmental Information Disclosure of Listed Companies” formulated by the Ministry of Environmental Protection, 15 industries, namely, thermal power, steel, cement, electrolytic aluminum, coal, metallurgy, the chemical industry, the petrochemical industry, building materials, paper making, brewing, the pharmaceutical industry, fermentation, textile, leather making and mining, are selected as heavily polluting industries. If the listed company belongs to one of the abovementioned industries, then Treat_i is equal to 1; otherwise, Treat_i is equal to 0. Post_t is a time-based dummy variable that is equal to 1 for years since 2012 and equal to 0 before 2012, as the GCGs were implemented in February 2012. The key explanatory variable in the model is Treat_i × Post_t, i.e., the multiplicative term of treat and post, which represents the policy effect.

In this study, we use the differences-in-differences method for research. The premise of this method is that the treatment group and the control group must meet the parallel trend assumption, that is, if there is no policy impact, the change trend of the result variable in the two groups of samples should be the same. In this regard, we use the event study method to test. The results of the parallel trend test show that there was no significant difference in the level of green M&As between the two groups of samples before the release of the GCGs, and the parallel trend hypothesis is established.

We also performed correlation analysis and a multicollinearity test. The results are shown in Appendix A and Appendix B. The correlation coefficient matrix presented in Appendix A shows that there is a strong correlation between the dependent variable (Greenma) and the independent variable, providing preliminary support for subsequent analysis. The control variables are significantly related to the dependent variable, indicating that the regression analysis needs to control them. We used the variance inflation factor (VIF) to check whether the model is subject to a multicollinearity problem so as to avoid distortion of model estimation results due to potential high correlation between explanatory variables. According to Appendix B, the VIF values of all explanatory variables are less than 10, which is within the acceptable range, indicating that the collinearity problem of the model is weak.

Owing to the large sample size and the fact that we directly adopted the heteroscedasticity robust standard error in regression, we did not test heteroscedasticity.

3.4. Descriptive Statistics

Table 2. Variable descriptive statistics.

	Variable	Observation	Mean	Max	Min	Std. Dev
Treatment group	Greenma	1122	0.290	1	0	0.454
	Post	1122	0.742	1	0	0.438
	Size	1122	22.58	25.89	19.53	1.424
	Soe	1122	0.481	1	0	0.500
	Cashflow	1122	0.088	0.761	−0.415	0.231
	Dual	1122	0.215	1	0	0.411
	Lev	1122	0.480	0.925	0.0591	0.198
	Board	1122	9.798	17	5	2.398
	Top1	1122	37.14	73.01	8.975	15.40
	Roe	1122	0.082	0.383	−0.358	0.109
	Indep	1122	0.368	0.600	0.250	0.059
	Tobinq	1122	1.776	8.954	0.888	1.032
	Employee	1122	7.999	10.86	4.127	1.304
	Price	1122	19.01	23.15	14.56	1.883
	Ratio	1122	80.56	100	30	22.97
	Paymethod	1122	0.183	1	0	0.387
Control group	Greenma	3805	0.182	1	0	0.386
	Post	3805	0.803	1	0	0.398
	Size	3805	22.07	25.89	19.53	1.166
	Soe	3805	0.315	1	0	0.465
	Cashflow	3805	0.226	0.761	−0.415	0.223
	Dual	3805	0.302	1	0	0.459
	Lev	3805	0.429	0.925	0.059	0.198
	Board	3805	9.234	17	5	2.180
	Top1	3805	34.72	73.01	8.975	14.83
	Roe	3805	0.090	0.383	−0.358	0.094
	Indep	3805	0.380	0.600	0.250	0.065
	Tobinq	3805	2.191	8.954	0.888	1.405
	Employee	3805	7.595	10.86	4.127	1.216
	Price	3805	18.87	23.15	14.56	1.805
	Ratio	3805	79.66	100	30	23.29
	Paymethod	3805	0.235	1	0	0.424

This outcome shows that heavily polluting enterprises are more inclined to implement green M&As; however, more powerful evidence obtained through empirical research is required to determine whether this phenomenon is related to the implementation of green credit policy.

shows the results of descriptive statistics of variables that distinguish the treatment group from the control group. The total sample size is 4927, including 1122 samples in the treatment group and 3805 samples in the control group. Among all the samples, the proportion of green M&A samples is 0.207; however, there is a large gap between the treatment group and the control group. The proportion of green M&A samples in the treatment group is 0.29, which far exceeds the 0.182 found for the control group.

4. Empirical Results

4.1. The Impact of Green Credit Policy on Enterprise Green M&A

The estimation results of Model 1 are shown in

Table 3. The impact of green credit policy on enterprise green M&As.

Variable	Greenma
	(1)	(2)
Treat $\times$ Post	0.785 *** (0.233)	0.708 *** (0.238)
Post	0.827 ** (0.324)	0.990 *** (0.336)
Treat	−0.007 (0.217)	0.031 (0.223)
Size		0.114 * (0.062)
Soe		0.050 (0.098)
Cashflow		0.556 ** (0.257)
Dual		−0.063 (−0.090)
Lev		1.094 *** (0.322)
Board		−0.001 (−0.018)
Top1		−0.001 (−0.003)
Roe		0.842 * (0.449)
Indep		−1.168 * (0.624)
Tobinq		−0.116 *** (0.042)
Employee		−0.241 *** (0.054)
Price		−0.072 *** (0.027)
Ratio		−0.003 * (0.002)
Paymethod		0.331 *** (0.119)
Constant	−3.366 *** (0.585)	−2.473 ** (1.234)
Industry FE	Y	Y
Year FE	Y	Y
Pseudo R2	0.1288	0.1414
LR	646.463 (p = 0.000)	709.855 (p = 0.0000)
N	4927	4927

Note: standard errors in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

. All regressions control for industry fixed effects and year fixed effects. Column (1) shows the regression results when the control variable is not added. The influence coefficient of GCGs on green M&As is 0.785, which is significant at the 1% confidence level. According to column (2), after controlling the variables at the three levels of enterprise characteristics, corporate governance and M&A transaction characteristics, the estimated coefficient is 0.708, which is still significant at the 1% confidence level. The estimation results show that the implementation of the GCGs has promoted the green M&A activities of heavily polluting enterprises; thus, Hypothesis 1 is verified.

4.2. Robustness Test

4.2.1. Parallel Trend Test

In this study, the treatment group and the control group are required to have the same trend of implementing green M&As before the issuance of the GCGs. Therefore, we conducted tests conducted using the event study method. The specific mean was generated by generating dummy variables for each year from 2007 to 2011 before the implementation of the GCGs and multiplying the newly generated dummy variables for each year by the group dummy variables (Treat). Then, all these multiplication terms (Treat $\times$ Post₂₀₀₇, Treat $\times$ Post₂₀₀₈, Treat $\times$ Post₂₀₀₉, Treat $\times$ Post₂₀₁₀ and Treat $\times$ Post₂₀₁₁) were incorporated into Model (1) for estimation. If these multiplier coefficients are not significant, it means that there is no significant difference in the level of green M&As between the two groups of samples before the release of the GCGs, and the parallel trend hypothesis is established. As shown in

Table 4. Parallel trend test.

Variable	Greenma
	(1)	(2)
Treat $\times$ Post2007	−0.315 (0.681)
Treat $\times$ Post2008	−0.696 (0.532)
Treat $\times$ Post2009	0.373 (0.473)
Treat $\times$ Post2010	0.015 (0.427)
Treat $\times$ Post2011	0.391 (0.411)
Treat $\times$ Post		0.742 *** (0.114)
Controls	Y
Year FE	Y
Industry FE	Y
Pseudo R2	0.1422
N	4927

Note: standard errors in parentheses; *** p < 0.01

and Figure 1 the estimated coefficients of these multiplication terms are not significant, indicating that there is no significant difference in green M&As between heavily polluting enterprises and non-heavily polluting enterprises before the implementation of the policy. Therefore, the study sample passed the parallel trend test. According to the result of the dynamic effect (Figure 1), the driving effect of GCGs on enterprise green M&As was significant in most years after the policy was introduced, with a long-term stable effect.

4.2.2. PSM-DID

Owing to the large differences in the size and financial status of sample enterprises, the regression results may be subject to the problem of sample selection error. To solve this problem, we conducted a PSM-DID test. We selected the aforementioned control variables as characteristic variables, used the logit model to estimate the probability of each sample being selected in the experimental group and then used the radius-matching method to match the treatment group with a reasonable control group.

Table 5. Balance test of PSM.

Variable	Unmatched Matched	Mean		Bias		t-Test
		Treatment Group	Control Group	%bias	Reduction (%)	t Value	p > |t|
Size	Unmatched	22.58	22.07	39.4	87.7	12.27	0.000
	Matched	22.45	22.38	4.9		1.140	0.256
Soe	Unmatched	0.481	0.317	33.9	94.5	10.19	0.000
	Matched	0.452	0.443	1.9		0.420	0.678
Cashflow	Unmatched	0.0881	0.225	−60.0	97.7	−17.83	0.000
	Matched	0.106	0.109	−1.4		−0.320	0.748
Dual	Unmatched	0.215	0.301	−19.7	87.1	−5.630	0.000
	Matched	0.224	0.235	−2.5		−0.600	0.546
Lev	Unmatched	0.480	0.430	25.2	76.9	7.430	0.000
	Matched	0.470	0.459	5.8		1.370	0.171
Board	Unmatched	9.798	9.239	24.4	80.7	7.370	0.000
	Matched	9.679	9.787	−4.7		−1.040	0.298
Top1	Unmatched	37.14	34.74	15.9	95.3	4.720	0.000
	Matched	36.42	36.53	−0.7		−0.170	0.867
Roe	Unmatched	0.0822	0.0901	−7.8	74.8	−2.400	0.0170
	Matched	0.0843	0.0823	2.0		0.460	0.644
Indep	Unmatched	0.369	0.380	−18.8	90.1	−5.400	0.000
	Matched	0.370	0.369	1.9		0.440	0.659
Tobinq	Unmatched	1.776	2.191	−33.7	88.0	−9.210	0.000
	Matched	1.815	1.865	−4.1		−1.070	0.285
Employee	Unmatched	7.999	7.598	31.8	85.4	9.550	0.000
	Matched	7.911	7.852	4.6		1.080	0.278
Price	Unmatched	19.01	18.87	7.4	77.2	2.200	0.028
	Matched	18.95	18.92	1.7		0.380	0.701
Ratio	Unmatched	80.56	79.65	3.9	77.9	1.150	0.251
	Matched	80.34	80.55	−0.9		−0.200	0.843
Paymethod	Unmatched	0.183	0.234	−12.7	98.4	−3.650	0.000
	Matched	0.193	0.194	−0.2		−0.050	0.963

and Figure 2 show the balance test results of PSM and the standardized deviations of covariates before and after matching. There is no significant difference in each characteristic variable between the treatment group and the control group after matching, indicating that the matching was effective.

Table 6. Robustness test: PSM-DID.

Variable	Greenma
Treat $\times$ Post	0.732 *** (0.247)
Post	1.049 *** (0.344)
Treat	0.038 (0.230)
constant	−2.590 ** (1.264)
Controls	Y
Industry FE	Y
Year FE	Y
N	4747
Pseudo R2	0.1416

Note: standard errors in parentheses; *** p < 0.01, ** p < 0.05.

shows the estimation results with the matched samples; the policy effect variable Treat $\times$ Post coefficient is still significantly positive, which further confirms the robustness of the conclusions of this study.

4.2.3. Change in the Industry Definition Standard

We redefined heavily polluting industries according to the industry’s pollution emission intensity and redivided the treatment group and the control group accordingly. First, four types of pollutant emissions were selected, namely, industrial sulfur dioxide emissions, industrial smoke (powder) emissions, industrial wastewater emissions and industrial solid waste, and the proportion of these four types of pollutant emissions in each industry in 2011 was calculated. If the emission ratio was greater than the median value, it was assigned a value of 1; otherwise, it was assigned a value of 0. Finally, industries whose sum was greater than 2 were defined as heavily polluting industries. Treat₂ represents the new group variable, and Treat $\times$ Post₂ represents the multiplication term of Treat₂ and Post.

Table 7. Robustness test: change in the industry definition standard.

Variable	Greenma
Treat $\times$ Post2	0.529 ** (0.227)
Post	0.974 *** (0.341)
Treat2	−0.216 (0.220)
Constant	−2.829 ** (1.224)
Controls	Y
Industry FE	Y
Year FE	Y
N	4927
Pseudo R2	0.1355

Note: standard errors in parentheses; *** p < 0.01, ** p < 0.05

shows the results of the estimation according to the changed groups. The estimation coefficient of Treat $\times$ Post₂ is significantly positive at the 5% level, which supports the previous conclusion.

4.2.4. Excluding the Interference of Other Policies

In this study, we excluded the interference of similar policies from two aspects: (a) given the potential impact of the promulgation and implementation of the new “Environmental Protection Law” in 2015 on the estimation results, we generated a dummy variable (EPL) representing the implementation of the new “Environmental Protection Law” and incorporated it into Model 1 for re-estimation. The results are shown in column (1) of

Table 8. Robustness test: excluding the interference of other policies.

Variable	Greenma
	(1)	(2)
Treat $\times$ Post	0.683 *** (0.236)	0.562 ** (0.276)
Post	−0.021 (0.188)	1.097 *** (0.385)
Treat	0.044 (0.222)	0.075 (0.258)
EPL	−0.428 *** (0.155)
constant	−222.166 *** (60.832)	−2.222 (1.565)
Controls	Y	Y
Industry FE	Y	Y
Year FE	Y	Y
N	4927	3324
Pseudo R2	0.1375	0.1485

Note: standard errors in parentheses; *** p < 0.01, ** p < 0.05.

. The previous conclusion is still tenable. (b) Since 2016, China has established green finance reform and innovation pilot zones in six provinces (autonomous regions); thus, the development trend of green finance in the pilot zones must be differ from that in other regions. To exclude the impact of this policy, we removed the samples from the six provinces, and the regression was performed again. Column (2) in Table 8 shows that the Treat $\times$ Post coefficient is still significantly positive.

4.2.5. Control over Regional and Industry Characteristics

Certain characteristics at the industry and regional levels may also influence corporate green M&A decisions. To control the impact of industry-level factors, we added the Herfindahl–Hirschman index (HHI), the average return on equity (IndROE), the operating income growth rate (IndGrowth) and Tobin’s Q (IndTQ) of the industry to which the merger belongs for more complete control over industry-level information. To control the influence of regional-level factors, the per capita GDP of the province (ProvinceGDP) where the enterprise is located was added to reflect the regional economic level, and the provincial fixed effect was further controlled. Columns (1) and (2) in

Table 9. Robustness test: further control over regional and industry characteristics.

Variable	Greenma
	(1)	(2)
Treat $\times$ Post	0.687 *** (0.242)	0.611 ** (0.243)
Post	0.800 ** (0.348)	1.749 *** (0.477)
Treat	−0.029 (0.231)	0.111 (0.231)
HHI	−0.135 (0.222)
IndROE	0.414 (0.646)
IndGrowth	−0.056 * (0.033)
IndTQ	−0.119 ** (0.047)
ProvinceGDP		−0.00001 ** (5.43 × 10−6)
Constant	−1.965 (1.262)	−2.319 (1.410)
Controls	Y	Y
Province FE	N	Y
Industry FE	Y	Y
Year FE	Y	Y
N	4835	4927
Pseudo R2	0.1417	0.1520

Note: standard errors in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

show the regression results of further controlling industry and province characteristics, respectively; the conclusions remain unchanged.

4.3. Heterogeneity Analysis

According to the analysis presented above, the green credit policy has a significant promoting effect on the green M&As of heavily polluting enterprises at the whole-sample level. However, for enterprises with differing levels of green innovation and financial development, does the above conclusion still hold? Answering this question will help to understand the microeffects of green credit policy in different situations.

4.3.1. Analysis of the Heterogeneity of Green Technology Innovation Capability

In this study, the samples were divided into two groups, and regressions were conducted according to whether green patents were applied for in the year before the M&A as the standard for classifying the green innovation capabilities of enterprises.

The regression results are shown in

Table 10. Heterogeneity analysis: green technology innovation capability.

Variable	Greenma
	Strong Green Innovation Capability	Weak Green Innovation Capability
	(1)	(2)
Treat $\times$ Post	−0.015 (0.393)	1.142 *** (0.313)
Post	0.889 (0.677)	0.607 (0.448)
Treat	0.684 * (0.380)	−0.249 (0.286)
Constant	−2.719 (1.797)	−0.819 (1.871)
Controls	Y	Y
Industry FE	Y	Y
Year FE	Y	Y
N	2162	2756
Pseudo R2	0.1343	0.1511
Fisher’s Permutation test	0.010

Note: standard errors in parentheses; *** p < 0.01, * p < 0.1.

. Columns (1) and (2) show that the regression coefficient of the variable Treat $\times$ Post, which reflects the policy effect, is only significant in the group with weak green technology innovation ability. To ensure the robustness of the grouped regression results, we also conducted a Fisher’s permutation test. The results show that the coefficient difference between groups is significant at the 5% level, which indicates that the promotion effect of green credit on the green M&As of heavily polluting enterprises is more obvious than that of enterprises with weak green innovation ability.

4.3.2. Analysis of Heterogeneity of Marketization Level

According to the “Financial Marketization Degree Index” reported in the “China Marketization Index—2011 Report on the Relative Progress of Marketization in Various Regions”, the sample enterprises in the top five provinces in 2011 were classified as high marketization level groups, and the remaining enterprises were classified as low marketization level groups.

The regression results are listed in

Table 11. Heterogeneity analysis: marketization level.

Variable	Greenma
	High Degree of Marketization	Low Degree of Marketization
	(1)	(2)
Treat $\times$ Post	0.167 (0.379)	1.058 *** (0.308)
Post	1.062 * (0.583)	1.094 *** (0.419)
Treat	0.689 * (0.355)	−0.323 (0.293)
Constant	−1.880 (1.755)	−2.431 (1.819)
Controls	Y	Y
Industry FE	Y	Y
Year FE	Y	Y
N	2732	2184
Pseudo R2	0.1603	0.1374
Fisher’s Permutation test	0.035

Note: standard errors in parentheses; *** p < 0.01,* p < 0.1.

. The policy effect is significant in the low marketization level group but not in the high marketization level group. Fisher’s permutation test results show that the differences in coefficients between groups were also significant at the 5% level. This regression result shows that the green credit policy promotes the green M&As of heavily polluting enterprises in areas with a low level of marketization.

4.4. Further Analysis

Green Innovation Performance Test for Green M&As

Theoretically, absorbing external technological innovation elements, especially advanced green technology, through M&As can drive the green technology R&D of enterprises. However, scholars still have doubts about the actual green effect. Through their research, scholars have found that on the one hand, green M&As may be an important starting point for enterprises to change their development strategies, improve their production methods, upgrade their technological level and promote their transformation and upgrading. However, it may also be a strategic behavior to seek policy benefits or divert the focus of public opinion [26,39]. To explore whether green M&As driven by green credit policies are effective, we used the mediation effect model to test the green innovation performance of enterprises after green M&As.

Considering the high risk and time-consuming development of innovation, we used the difference (∆GPatent) between the number of green patent applications one to two years after the merger and the data from the year before the merger to measure the level of green innovation. Because green M&As are a categorical variable, whereas green technological innovation is a continuous variable, their regression coefficients are not on the same scale and thus cannot be judged only by step-by-step testing of the regression coefficients. Therefore, we used the Sobel method proposed by Iacobucci (2012) [40] and the distribution-of-the-product method proposed by MacKinnon and Cox (2012) [41]. The models are as follows:

ΔGPatent_it = η₀ + η₁Treat_i × Post_t + η₂Treat_i + η₃Post_t + η₄controls_it +μ_j + ω_t + ε_it

(2)

ΔGPatent_it = φ₀ + φ₁Treat_i × Post_t + φ₂Treat_i + φ₃Post_t + φ₄controls_it +μ_j + ω_t + ε_it

(3)

Model (2) considers the impact of green credit policy on corporate green technology innovation, and Model (3) adds the variable Greenma to (2) to examine whether it can play an intermediary role in the relationship between green credit and corporate green technology.

Table 12. Green performance of green M&As driven by green credit.

Variable	t + 1		t + 2
	∆GPatent1		∆GPatent2
	(1)	(2)	(3)	(4)
Treat × Post	2.011 *	1.713	2.474 **	2.080 *
	(1.112)	(1.135)	(1.123)	(1.149)
Greenma		2.135 ***		2.897 ***
		(0.733)		(1.093)
Post	−5.438 ***	−5.640 ***	−5.500 ***	−5.806 ***
	(0.884)	(0.852)	(0.952)	(0.995)
Treat	−2.072 **	−2.045 **	−3.334 ***	−3.264 ***
	(1.042)	(1.041)	(1.111)	(1.106)
Constant	−40.580 ***	−40.926 ***	−48.917 ***	−49.280 ***
	(7.828)	(7.837)	(13.713)	(13.772)
Controls	Y	Y	Y	Y
Year FE	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y
R2	0.0592	0.0626	0.0612	0.0651
N	4426	4426	4062	4062
The proportion of the mediation effect	0.1477738		0.15937114
Sobel-Z	0.29710182 ***		0.39435868 ***
95%CI	0.30334−3.19622		0.33608−4.47078

Note: standard errors in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the regression results.

As shown in columns (1) and (3), the coefficients of Treat×Post are significantly positive at the 10% and 5% levels, indicating that the GCGs have promoted green technology innovation in heavily polluting enterprises. Further considering the mediating effect of green M&As, as shown in columns (2) and (4), the Greenma coefficient and the Sobel-Z statistic of the mediation factor test are both significant, and the 95% confidence interval computed by the RMediation package based on the distribution of products does not contain 0, which supports the intermediary effect of green M&As. The above results show that green M&As driven by the GCGs can have green effects on enterprises and promote their innovation and upgrading.

5. Discussion

In this study, we investigated the impact of green credit policy on green M&As and further explored the performance of green M&As. The main findings are as follows:

First, Table 3 shows the empirical results of Hypothesis 1. The results show that highly polluting enterprises affected by green credit policies are more inclined to implement green M&As, which reflects the effectiveness of green credit policies in green governance. From this perspective, our research results are consistent with those reported by Wang (2022) [5], Chen (2022) [21], etc. However, unlike previous studies that focused on independent innovation, we turned our attention to the open innovation approach of enterprise green M&As and identified a new path for green credit policy to play its role. Faced with the constraint effect of credit resources caused by green credit policies, the willingness of enterprises to engage in green transformation was strengthened, with enterprises seeking transformation and upgrading through the implementation of green M&As. Moreover, according to the results presented in Table 10, enterprises with a weak green technology innovation foundation are more inclined to implement green M&As to accumulate transformation strength by acquiring the green technology and resources of the acquiree. From the unique perspective of green M&As, this paper provides new evidence of the effectiveness of green credit policy. Huang (2021) [25] believes that green industry policy is an important external driver for enterprises to implement green M&As, and the correlation between environmental regulation and green M&A activities has been confirmed. Our research provides supplementary evidence that green credit policies promote the implementation of green M&As by enterprises.

Furthermore, we used Greenma as an intermediary variable to explore whether enterprises can obtain green benefits by implementing green M&As. The results presented in Table 12 shows that the answer is yes. Compared with before the implementation of green M&As, enterprise green innovation activities are more active, and the number of green patent applications has increased. Furthermore, the performance in the second year after the M&A is improved relative to that in the first year. The above results show that green M&As driven by green credit policy can promote green technology innovation in enterprises, and this promotion is not short-term. This is consistent with the conclusion of Huang and Yuan (2022) [31] and Liang et al. (2022) [2], but compared with previous research on green M&As, which does not distinguish between the implementation background and regards it as a whole, our research is unique in that it situates the discussion on the performance of green M&As in a specific context, which confirms that under the pressure of green credit policy, enterprise green M&As are still effective and play a role in promoting green innovation. Furthermore, according to the view of motive consequentialism, the motivation of the actor plays a decisive role in the consequences of the behavior [25], so an enterprise’s green innovation performance can reflect the real motivation of green M&As to a certain extent. Our test results on the performance of green innovation imply that under the pressure of green credit policy, enterprises implement green M&As more for the purpose of seeking their own green transformation than for the strategic motivation of obtaining policy benefits or circumventing policy restrictions.

6. Conclusions and Policy Implications

6.1. Conclusions

Under the background of the dual carbon goal, the green transformation and development of enterprises is an inevitable trend and requirement. The green credit policy, which guides the green development of enterprises through the allocation of credit resources, is an important policy innovation that is used to promote the transformation and upgrading of enterprises. Based on the M&A data of A-share listed companies in Shanghai and Shenzhen from 2007 to 2020, in this study, we focused on the behavior of green M&As, taking the introduction of the “Green Credit Guidelines” as a quasi-natural experiment and empirically testing the impact of green credit policy on green M&As of enterprises using the differences-in-differences method.

Our study results show that the implementation of a green credit policy will promote the green M&As of heavily polluting enterprises; this conclusion still holds after a series of robustness tests. The results of the heterogeneity analysis show that for enterprises with poor green innovation foundations and enterprises with a low level of financial marketization, the effect of green credit policy on green M&As is more significant. Further tests show that green M&As driven by the GCGs promote the green innovation of enterprises.

6.2. Policy Implications

Based on the above research conclusions, we suggest the following.

First, from the perspective of the government, it is necessary to continuously improve the green credit system to stimulate the impetus for enterprise transformation. We found that green credit policy encourages and forces heavily polluting enterprises to improve their environmental behavior by guiding the green allocation of credit funds and ultimately improves the green performance of enterprises, thereby demonstrating the effectiveness of green governance. Therefore, on the one hand, relevant state departments should actively improve the relevant systems of green credit, ensure the implementation of green credit policy and promote the rapid development of China’s green credit business. On the other hand, it is also necessary to strengthen the monitoring of the implementation, establish an evaluation system and incorporate it into the green credit policy design and implementation process and continue to revise and improve the implementation rules and supporting policies based on feedback.

Second, from the perspective of banks, in the process of lending, it is necessary to do a good job in the role of credit approval and strengthen the assessment and management of the environmental and social risks of enterprises. In the “postloan” stage, information disclosure criteria should be actively improved, and the supervision of green credit funds should be strengthened with the help of financial technology to ensure that funds are used for environmentally friendly projects. It is also necessary to regularly evaluate and improve the development of green credit businesses to continue to play the role of a green credit policy in guiding the green transformation of enterprises.

Finally, the follow-up research conducted in this study revealed that green M&As driven by green credit policy can improve the green innovation performance of enterprises. Therefore, enterprises, especially those with poor green innovation capabilities, can consider implementing green M&As to gain first-mover advantages in transformation. On the one hand, enterprises should abandon short-sighted thinking, conform to the trend of green transformation and development and truly take green M&As as an important starting point for transformation of production methods and realization of transformation and upgrading. On the other hand, it is necessary to further strengthen integration management after M&As, realize the substantive integration of technologies and resources of both parties and promote the sustainable development of enterprises.

In this paper, we preliminarily discussed the impact of green credit policy on corporate green M&A decisions and further examined the green effects of green M&As. However, the research design of this study is subject to some limitations and needs to be further expanded and improved. The specific manifestations are as follows. The first is the basis for grouping. When grouping, we drew on the mainstream treatment methods used in the current research to take heavily polluting enterprises as the treatment group and other enterprises as the control group. When identifying the treatment group and the control group, two methods were adopted based on the division results of existing research, with enterprises grouped according to the pollution emission intensity of their parent industry. However, it cannot be ruled out that there may be enterprises in the control group that are considerably affected by the green credit policy. The second manifestation consists of the interference of environmental policies during the study period. In the context of ecological civilization construction, the state has intensively issued a series of environmentally related policies, and the impact of these policies and those of green credit policy may overlap. Although we adopted a variety of methods in the robustness test, it is still impossible to completely rule out the impact of related policies during the same period. Therefore, the evaluation of the implementation effect of the green credit policy needs to be improved and perfected with respect to the mining and identification strategies of relevant microdata.