Green Technology Innovation Premium: Evidence from New Energy Vehicle Industry in China

Abstract

Climate change and environmental issues have received increasing attention across the world. China’s governmental targets for carbon peak and carbon neutralization show the ambition and efforts necessary in challenging these problems. The transportation industry will be crucial in reducing carbon emissions. Based on the green patent application data in China’s new energy vehicle (NEV) industry from 2006 to 2021, this article focuses on risk premium of green technology innovation. In particular, the premium effects of the green technology innovation and the cooperative network are empirically examined. Furthermore, two channels that play a role in generating the premium are investigated, i.e., attracting market attention and reducing financing constraints. The empirical results show that the stock returns are positively correlated to the green technology innovation and the company’s central position in the cooperative network, i.e., there exist the premium effects of green technology innovation in China’s NEV industry. The positional advantage in the cooperative innovation network can further increase analyst following and reduce financing constraints. The research can provide evidence and policy implications for the government, companies and investors.

1. Introduction

With the rapid development of the global economy in recent decades, environmental issues such as climate change and air pollution have become increasingly severe. In September 2020, China put forward the timeline target for carbon peak and carbon neutralization (Dual Carbon Goal) to enhance public awareness and encourage effective practices in environmental protection activities. The transportation industry will be an important field to reduce carbon emissions, and thus promoting green technology innovation in the new energy vehicle (NEV) industry will play an important role in addressing climate change problems.

Companies are the important force in green innovation. Green innovation can help companies achieve a win-win situation of economic and environmental benefits and can effectively lessen adverse effects on the environment [1]. Concerning both environmental protection and technology advancement, green innovation typically faces high technical barriers and requires substantial investments in both human resources and financial capital. Meanwhile, it faces high technological and market risks. Due to the long return cycle, market uncertainty and high risks, continuous financial support is needed to maintain its green investment [2]. The firm performance in the capital market will affect the firm’s motivation to engage in green innovation as well as the capital sources. Whether there exists risk premium in green innovation will also affect the investors’ expectations and their investment decisions.

Green innovation involves complex and diverse technologies, requiring the integration of technology and resources from multiple stakeholders to make progress. The participation of the government and the market is necessary. A well-built network facilitates the sharing and dissemination of relevant information, thereby enhancing innovation efficiency across the companies as well as the related industries.

Besides, companies with strength in green innovation can receive more resources. In the capital market, environmentally friendly companies are more favored by investors for their positive social image, thus leading to active market transactions and higher return expectation. Moreover, as the government initiates policies promoting green development, companies actively engaged in green innovation are more likely to secure funding through green credit or green bond issuance, thereby easing financing constraints and improving the level of green technology innovation [3].

In summary, this article aims to answer the following questions: Is there risk premium for green technology innovation in the capital market? Can the innovation network also have the premium effect? What role do market attention and financing constraints play in generating the premium?

Specifically, based on green patent application data in China’s new energy vehicle industry, this article examines the existence of the innovation premium in the Chinese capital market, and focuses more on testing the premium effect of innovation network. The market attention and financing constraints are further investigated to check their role in generating innovation premium.

The research contributes to the following two aspects. First, existing research mainly explores the promoting effects of environmental regulations [4,5], credit markets [6], and government policies [7] on green innovation, while there is relatively little research on the performance of green innovation in the Chinese capital market. This article focuses more on studying the premium effect of green technology innovation network. Green technology innovation usually faces more challenges and risks for their environmental and innovative concerns, and thus the cooperative network can play an important role in resource sharing as well as risk sharing and diversification. Moreover, we further investigate the channels of analysts following and financing constraints that the innovation network functions in generating the premium. Second, the research findings have policy implications. The results can provide decision-making references for the government to promote innovation policies, encourage companies to be active in green innovation, and motivate investors to give more attention to the companies excelling in green technology innovation.

2. Literature Review and Hypothesis Development

2.1. Green Technology Innovation

Green innovation is defined as hardware and software innovation related to green products or processes, which includes technology innovation in energy saving, pollution prevention, waste recycling, green product designs, or corporate environmental management [8]. Green technology innovation can improve the utilization of resources, prevent pollutants, and reduce waste and has been applied to other environmental management practices [9]. Furthermore, when companies engage in green innovation, the associated knowledge and technological advancements are not only utilized within their organizations but also shared with other companies, leading to spillover effects.

Green technology innovation involves high technical barriers and substantial capital investment for its targets in environmental protection and innovation, thus facing significant technical and market risks. Risk premium is typically used to describe the excess return required by investors for assuming excess risks. Hou et al. [10] studied the R&D intensity of listed companies across 21 countries and suggested that the companies with higher R&D intensity get higher stock returns, which is more likely attributable to risk premium. Li [11] and Gu [12] suggest that innovation-intensive enterprises, bearing more risks, can obtain higher excess returns. The market prefers the companies committed to green innovation and provides them with more risk compensation, which can enhance their stock returns. Green technology innovation can attract more investors’ attention, increase the trading volume of the stocks, and have an impact on stock liquidity [13]. Based on the above analysis, this article proposes the following hypothesis:

H1 (Innovation premium).

There is a risk premium for green technology innovation.

2.2. Green Technology Innovation Network

An innovation network can be defined as a basic institutional arrangement to cope with systemic innovation, and the cooperative relationships among firms are a key linkage mechanism for network configuration [14]. The process of green technology innovation requires the participation and collaboration of multiple parties such as governments, companies, research institutions, and consumers, which can accelerate the development and optimize the industrial structure [15].

The status and evolution of green innovation networks can be studied from the perspective of joint green patent applications. Some scholars have employed joint patent data to study cross-border cooperation in green technology innovation [16,17]. Cao et al. [18] constructed a cooperative technology innovation network by using patent application data about new energy vehicles and further used simulation to analyze the characteristics of the network.

From the viewpoint of companies, the position in the innovation network will affect their ability to obtain information and funds. The companies occupying a central position in the network can access more external resources and opportunities [19]. The central positions in the network can provide the companies with more information and resources, thereby enhancing their firm performance [20]. The central companies will become information hubs and can obtain more cooperation opportunities. Meanwhile, companies need to incur the costs to keep their central network position, such as maintaining the investment intensity of R&D activities and engaging in communication and knowledge sharing with other companies in the industry. In addition to the maintenance costs, these activities may have the risk of technology leakage. Based on the above analysis, this article proposes the following hypothesis:

H2 (Innovation network premium).

There is a premium effect for the central position in green innovation network.

2.3. Innovation Network and Market Attention

Higher R&D investment serves as a signal of active innovation and indicates more potential to launch new products or services, which attracts attention from the market. Financial analysts encourage firms to make more efficient investments related to innovation [21]. The companies engaging in green innovation are favored by socially responsible investors or green investors [13] and are more likely to receive relatively optimistic evaluations from the market, ultimately enhancing the company’s market value and stock return [10].

The companies more central in the network generally possess more resources and stronger innovation capabilities. Their activities in green innovation send signals to the market and can enhance their environmental reputation. The stocks of these central companies usually have good liquidity in the capital market [13]. Therefore, companies with network advantage have an impact on their stock returns and generate the premium by the channel of market attention. Accordingly, the following hypothesis is proposed:

H3. 

The companies with network advantage receive more market attention.

2.4. Innovation Network and Financing Constraints

Financing constraints are attributed to imperfections in the capital market, such as information asymmetry and incentive problems [22]. Unstable sources of capital investments directly affect the company’s financial status and impede the development in the long run. When the companies face higher financing constraints, their green innovation activities will be impaired [23]. Cao et al. [24] found that financing constraints have a significant and negative effect on innovation activities while social networks can alleviate financial constraints.

The companies with network advantage can prioritize information and resources from partner companies and improve their innovation capabilities through cooperation. Meanwhile, they have more opportunities to convey information to the market, reduce information asymmetry, and thus ease external financing. The company’s central position in the network can alleviate its financing constraints [25]. The central position in the innovation network can help improve corporate reputation and obtain governmental support, such as tax incentives, fiscal subsidies, and procurement preferences, thereby bringing more sales profits and improving operating performance. When engaging in more green innovation, the companies can benefit from lower equity financing costs [26], and possibly obtain green credit or issue green bonds, thus reducing bond financing costs, easing financing constraints, and increasing green innovation [27]. Accordingly, the following hypothesis is proposed:

H4. 

The companies with network advantage can alleviate their financing constraints.

3. Green Technology Innovation and Network in NEV Industry

3.1. Green Technology Innovation

The data on patents is sourced from the Patsnap database (Patsnap is a global patent and innovation database that provides users with a comprehensive and user-friendly platform for conducting patent searches. https://www.patsnap.com). According to the IPC Green Inventory (IPC Green Inventory, launched by the World Intellectual Property Organization in September 2010, is a tool designed to facilitate public access to patent information related to environmentally sound technologies. https://www.wipo.int/classifications/ipc/green-inventory/home), as shown in

Table 1. The list for the IPC Green Inventory.

First Level Classification	Second Level Classification	International Patent Classification (IPC)
Transportation	hybrid electric vehicles, electric vehicle charging stations, hydrogen fuel engines, etc.	B60K6/00, B60K6/20, H02K29/08, H02K49/10,
		B60L7/10-7/22, B60L8/00, B60L9/00, F02B43/00
		F02M21/0, F02M27/02, B60K16/00, H02J7/00

, the patent application data in China’s new energy vehicle industry can be collected. Figure 1 shows the temporal change in the number of (joint) green patent applications from 2006 to 2021.

As shown in Figure 1, the number of green patent applications has increased continuously, from 33 in 2006 to 14,332 in 2021. This reflects the industry’s growing attention to green technology innovation and increasing R&D investment. Meanwhile, the number of joint patent applications has also increased, reaching 924 in 2021, accounting for 6.45% of the total number of green patent applications in this year. The smaller number of joint patent applications shows the need for more institutions to engage in cooperative innovation activities to promote the relevant knowledge spillover and thus enhance the efficiency of the innovation system. The proportion of joint patents is still at a low level and has fallen from 8.08% in 2015 to 4.99% in 2019, then rising slightly to 6.45% in 2021.

3.2. Green Technology Innovation Network

Based on the joint patent application data in China’s NEV industry, which include the information about the patent applicants (companies, research institutions, individuals, etc.), the collaboration relationships between applicants can be described and thus a cooperative innovation network can be constructed, called a green technology cooperative innovation network and referred to briefly as a green innovation network or innovation network hereafter.

This sample covers green patent application data for three stages, publication, substantive examination, and authorization. From 2006 to 2021, 87,226 green patents in NEV industry are collected and the data are processed as follows: exclude patents with only one applicant; exclude patents jointly applied for with foreign, Hong Kong, Macao, and Taiwan patentees; for data about the same patent application at the different stage, prioritize publication data; treat the headquarters and branches as separate applicants.

After the above screening and processing, 4137 green patent applications are obtained, involving 3000 innovative entities such as 2659 companies, 159 universities, 131 research institutes, etc.

In 2012, the State Council of China issued the Development Plan for Energy Conservation and New Energy Vehicle Industry (2012–2020), which had a positive impact on the NEV industry. Accordingly, the development in China’s NEV industry can be divided into two stages.

Table 2. The proportion of different types of green patent applicants.

Development Stage	Company	Research Institute	University	Other
The first stage (2006–2012)	0.74	0.11	0.08	0.07
The second stage (2013–2021)	0.84	0.06	0.06	0.05

gives the proportion of different types of the green patent applicants.

Companies are the main body in joint green innovation and they play an important role in the NEV industry. Based on the data in the sample, 1217 joint applications involving listed companies were screened out to construct a cooperative innovation network.

Figure 2 shows the innovation network for the periods of 2006–2012 and 2013–2021.

Table 3. The structural properties of the green innovation network.

Development Stage	Company Node	Joint Application	Density	Average Path Length	Clustering Coefficient
The first stage (2006–2012)	32	59	0.12	1.23	0.75
The second stage (2013–2021)	423	855	0.01	2.49	0.74

gives the structural properties of the networks in Figure 2. It can be shown that China’s innovation network has experienced rapid development in recent years. The number of companies involved in the cooperative green innovation network has increased from 32 to 423, and the number of joint applications has increased from 59 to 855. The listed companies tend to collaborate more with other companies. The density and average path length are decreased as the nodes (companies) are largely increased. The similar values of clustering coefficients indicate the nearly stable local clustering feature (cooperation) in the network.

4. Models and Results

This section empirically examines if there exists a premium effect of green innovation and the cooperative network. After baseline regression and robustness checks for Hypothesis 1 and 2, we further investigate the channels of market attention (Hypothesis 3) and financing constraints (Hypothesis 4) in generating the premium.

4.1. Empirical Models

The following models are used to examine the premium effect of green technology innovation and the innovation network.

$${Premium}_{i,t}=\mathsf{\alpha }+\beta {GreenInnov}_{i,t}+\sum {\gamma }_k{Control}_{i,k,t-1}+{\mu }_i+{\lambda }_t+{\epsilon }_{i,t}$$

(1)

$${Premium}_{i,t}=\mathsf{\alpha }+\beta {InnovNet}_{i,t}+\sum {\gamma }_k{Control}_{i,k,t-1}+{\mu }_i+{\lambda }_t+{\epsilon }_{i,t}$$

(2)

${Premium}_{i,t}$ is the proxy variable for risk premium of green innovation. ${GreenInnov}_{i,t}$ is the number of green technology patents application for the firm i in the time t while ${InnovNet}_{i,t}$ is the centrality measure of the firm i in the innovation network. ${Control}_{i,k,t-1}$ is the k-th control variable. ${\mu }_i$ controls the firm fixed effect and ${\lambda }_t$ controls the time fixed effect. ${\epsilon }_{i,t}$ is the error term.

4.2. Data and Variables

Based on the green technology patent data from the Patsnap database, the green patent applications for China’s NEV industry can be collected from 2006–2021. The data on stock prices and characteristics of the listed companies are collected from CSMAR (China Stock Market & Accounting Research Database (CSMAR) provides comprehensive financial and accounting data for research and analysis in the Chinese stock market. https://www.csmar.com). The observations about specially treated companies or those with missing values are removed from the sample.

Table 4. The list of variables and explanations.

	Symbol	Explanation
Dependent variable	Premium	Annual accumulative excess return
Predictor variable	GreenInnov	The natural logarithm of one plus the number of green patents applied by the company in the year
	Degree	Degree centrality of the company in the green innovation network
	Closeness	Closeness centrality of the company in the green innovation network (natural logarithm)
	Betweenness	Betweenness centrality of the company in the green innovation network
	Eigenvector	Eigenvector centrality of the company in the green innovation network
Control variable	Volatility	The standard deviation of the weekly returns for the company’s stock in the year
	ReturnAvg	The average of the weekly returns for the company’s stock in the year
	Size	The total asset of the company (natural logarithm)
	Lev	The total debts divided by the total asset
	Independent	The ratio of independent directors on the board
	Dual	If the CEO is concurrently holding the position of the board chairman, the value is one; otherwise, zero
	LargestShareholding	The holding percentage of the largest shareholder
	Shareholding	The aggregate holding percentage of the shareholders from the second largest to the tenth
	TobinQ	Company value, proxied by market value divided by the total asset
	SOE	If the company is state-owned, the value is one; otherwise, zero
	AgeListed	The operating years since the company was listed in the stock exchange
	Intangible	The ratio of intangible assets to total assets

gives the list of the variables used in the models. Premium is proxied by the excess return, calculated as follows:

$${premium}_{i,t}={\prod }_{t=1}^n(r_{i,t}+1)-{\prod }_{t=1}^n(r_{m,t}+1)$$

(3)

$r_{i,t}$ is the weekly return for the stock of the company i while $r_{m,t}$ is the weekly return for the market (here, the weighted average return based on floating A-share market values in Shanghai and Shenzhen Stock Exchange is used as the market return).

$GreenInnov$is the number of green technology patent applications, representing the green innovation capability of the company. $InnovNet$ is further used to measure the position of the company in the green innovation network. Specifically, centrality measures such as degree, closeness, betweenness, and eigenvector are used to gauge the extent about how central the company is in the network, referring to Newman [28] for detailed definition (omitted here for brevity). These measures can represent the central position of the company from different perspectives.

Control variables are also explained in Table 4. For example, Size controls the scale of the company; Independent and Dual are the variables to control the corporate governance; and SOE controls the ownership type.

Table 5. Descriptive statistics and correlation coefficients.

1. Descriptive Statistics
Variables		Mean	SD	Min		Max
Premium (%)		12.031	64.641	−147.425		630.36
GreenInnov		2.001	1.692	0		7.32
Degree		0.357	0.964	0		9.00
Closeness		−0.759	1.6	−4.99		0.00
Betweenness		0.227	1.594	0		25.00
Eigenvector		−3.332	10.751	−41.154		0.00
Volatility		3.75	2.574	0.914		53.57
Return_Avg (%)		3.909	1.594	1.301		14.83
Size		23.003	1.902	19.713		28.64
Lev		0.45	0.17	0.036		0.98
Independent		0.388	0.07	0.25		0.80
Dual		0.306	0.461	0		1.00
LargestShareholding		0.374	0.169	0.087		0.86
Shareholding		0.247	0.127	0.015		0.62
TobinQ		1.876	1.139	0.784		12.53
SOE		0.405	0.491	0		1.00
AgeListed		2.047	0.792	0		3.37
Intangible		0.035	0.023	0		0.16
2. Correlation Coefficients for Variables
		(1)	(2)	(3)	(4)	(5)	(6)
Premium	(1)
GreenInnov	(2)	0.023
Degree	(3)	0.02	0.183 ***
Closeness	(4)	0.060 **	−0.158 ***	−0.657 ***
Betweenness	(5)	0.023	0.142 ***	0.705 ***	−0.194 ***
Eigenvector	(6)	0.044	−0.096 ***	−0.487 ***	0.659 ***	−0.194 ***

*** and ** represent significant level at 1% and 5%, respectively.

-1 gives the descriptive statistics. For example, premium reports a mean of 12.031% with a standard deviation of 64.641%. Degree centrality shows a mean of 0.357 with a standard deviation of 0.964. Table 5-2 only presents the correlations for dependent and predictor variables for brevity.

4.3. Baseline Regression Results

Table 6. Baseline regression: green innovation (network) premium.

	Dependent Variable
	Premium
	(1)	(2)	(3)	(4)	(5)
GreenInnov	0.0333 ***
	(2.74)
Degree		3.2914 *
		(−1.6814)
Closeness			1.603
			(1.40)
Betweenness				1.9708 **
				(1.99)
Eigenvector					0.169
					(0.97)
_cons	6.9626 ***	280.6917 ***	607.7599 ***	282.1100 ***	607.3805 ***
	(2.74)	(2.86)	(5.41)	(2.88)	(5.39)
Control variables	YES	YES	YES	YES	YES
Firm Fixed	YES	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES	YES
Observation	3374	988	988	988	988
adj. R2	0.105	0.367	0.125	0.368	0.124

***, ** and * represent significant level at 1%, 5% and 10%, respectively.

gives the baseline regression results. Column 1 shows the significant and positive effect of green innovation on risk premium. Consistent with Gu [12] and Hou et al. [10], higher R&D intensity can lead to higher stock return. The green technology innovation can generate the premium in China’s NEV industry, i.e., Hypothesis 1 holds. Columns 2–5 show the effect of the innovation network. Degree and betweenness centrality show significant and positive effects on the premium while the effects of closeness and eigenvector are positive but insignificant. These results indicate that the connections (degree) and mediation (betweenness) in the innovation network play a more important role in generating the premium, consistent with Hypothesis 2, which further emphasizes the network effect. Degree and between centrality show their significance in affecting the premium, indicating that cooperative innovation activities such as connecting with more partners or acting as an intermediary can have positive effects.

4.4. Robustness Checks

Table 7. Robustness checks by replacing dependent and independent variables.

	premium_moni,t		premiumi,t
	(1)	(2)	(3)	(4)
Degreei,t	3.6554 *
	(1.96)
Betweennessi,t		2.3291 **
		(2.47)
Degreei,t-1			5.0711 **
			(2.16)
Betweennessi,t−1				10.4843 *
				(1.67)
_cons	363.3479 ***	366.5366 ***	333.5526 ***	334.5825 ***
	(3.894)	(3.940)	(3.458)	(3.447)
Control variables	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES
Firm Fixed	YES	YES	YES	YES
Observation	988	988	967	967
adj. R2	0.357	0.358	0.38	0.379

***, ** and * represent significant level at 1%, 5% and 10%, respectively.

gives the robustness checks by replacing dependent and predictor variables. Columns 1 and 2 recalculate the risk premium by using monthly returns instead of weekly returns in Equation (3). Columns 3 and 4 use the lagged centrality in the innovation network. The results are consistent with the baseline regression, indicating the robust effect of the innovation network on risk premium.

Considering the possible impact of the previous premium on the current one,

Table 8. Regression results from dynamic panel.

	Premiumi,t
	(1)	(2)
Premiumi,t−1	−0.431 ***	−0.466 **
	(−2.682)	(−2.187)
Degreei,t	3.879 **
	(2.45)
Betweennessi,t		2.055 ***
		(3.13)
AR(1)	0.02	0.08
AR(2)	0.37	0.11
Sargan	0.87	0.90
Control variables	YES	YES
Year Fixed	YES	YES
Firm Fixed	YES	YES
Observation	988	988

*** and ** represent significant level at 1% and 5%, respectively.

employs the dynamic panel model to include the lagged dependent variable. The results show the significant and positive effects of the network, consistent with the previous ones.

Table 9. Robustness checks by using instrument variables.

	First-Stage		Second-Stage
Dependent Variable	Degree	Betweenness	Premium
	(1)	(2)	(3)	(4)
Degree			3.189 *
			(1.75)
Betweenness				3.327 ***
				(3.14)
Degree_city	7.069 ***
	(3.07)
Degree_industry_mean	−77.997 ***
	(−124.369)
Betweenness_city		0.623 ***
		(3.58)
Betweenness_industry_mean		−50.883 ***
		(−62.398)
Control variables	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES
Firm Fixed	YES	YES	YES	YES
F statistics	522.7 ***	25.71 ***
Over Identification Test			2.07	2.56
			(p = 0.1500)	(p = 0.1106)
Weak Instrument Test			4.07	7.31
			(p = 0.0256)	(p = 0.0022)
Observation	988	988	988	988
adj. R2	0.757	0.842	0.462	0.464

*** and * represent significant level at 1% and 10%, respectively.

further uses the two-step least squares (2SLS) method to mitigate endogeneity concerns. The city network centrality and industry-year average centrality are combined and used as the instrument variables. The tests for instrument variables are passed and the coefficients of degree and betweenness are significant and positive, indicating the consistent effects of the innovation network on risk premium.

4.5. Channels of Analysts Following and Financing Constraints

Table 10. Analysts following and financing constraints.

	Analyst		SA
	(1)	(2)	(3)	(4)
Degree	0.054 *		−0.012 ***
	(1.79)		(−3.505)
Betweenness		0.159 **		−0.039 ***
		(1.96)		(−4.182)
Constant	−9.159 ***	−9.050 ***	−5.015 ***	−4.981 ***
	(−6.024)	(−5.933)	(−28.579)	(−28.366)
Control variables	YES	YES	YES	YES
Year Fixed	YES	YES	YES	YES
Firm Fixed	YES	YES	YES	YES
Observation	988	988	988	988
adj. R2	0.232	0.232	0.753	0.754

***, ** and * represent significant level at 1%, 5% and 10%, respectively.

reports the results for examining the influencing channels of analysts following and financing constraints. Analysts following (Analyst) is proxied by the natural logarithm of one plus the number of analysts that follow the company [21]. SA index is used as the proxy for financing constraints [29], defined as follows:

$$SA=-0.737\times Size+0.043\times {Size}^2-0.04\times Age$$

(4)

Columns 1 and 2 show the significant and positive effects of the innovation network on analysts following. Consistent with Hypothesis 3, higher centrality in the green innovation network can increase the number of the analysts that follow the company, indicating more attention from the market. Similarly, Columns 3 and 4 show the significant and negative effect on financing constraints. Consistent with Peng et al. [25], higher centrality in the network can decrease financing constraints that the company faces in its operation. As such, Hypothesis 4 holds.

5. Conclusions

Based on the green patent application data in China’s new energy vehicle industry from 2006 to 2021, this article studies the risk premium of green technology innovation, and empirically examines if there exists the premium in the capital market, focusing on the premium effect of green innovation network. Market attention and financing constraints are further investigated for their role in generating innovation premium.

The research has the following findings: green technology innovation has a significantly positive effect on the company’s excess return, indicating the existence of risk premium on green technology innovation; the companies more central in the green innovation network can achieve higher excess return, i.e., the network advantage has a premium effect in the capital market; the companies with network advantage can receive more market attention and alleviate financing constraints. Meanwhile, these results are from the green innovation data sample in China’s new energy vehicle industry. Future research can be conducted to make comparisons across the industry as well as in other countries to understand the role of green innovation in the capital market.

Policy implications are as follows. First, the government can further introduce policies to encourage companies to intensify their efforts in green technology innovation, including green innovation funds, R&D subsidies, tax incentives, etc. In particular, the government can incentivize the innovative companies to actively engage in cooperative innovation, thereby expanding the network scale and enhancing its efficiency. Second, companies should continue to invest in green technology innovation and set up alliances or cooperative R&D laboratories. Companies can actively seek external partners and strengthen cooperation with universities and research institutions. Third, investors should insist on green investments, paying more attention to companies excelling in green technology innovation, and thus gain the premium from green innovation as the economic transition toward sustainable development models occurs.