The Impact of Carbon Risk on Value Creation of High-Carbon-Emission Enterprises: Evidence from China

Abstract

Based on the Cost Theory and Porter’s Hypothesis, this study focuses on high-carbon-emission enterprises and systematically explores how carbon risk affects their value creation. The sample comprises listed firms in high-carbon-emission industries listed on China’s Shanghai and Shenzhen A-shares during 2012–2022. A carbon risk measurement index is constructed using industrial energy consumption data, and a two-way fixed-effects model is employed to empirically test the relationship between carbon risk and value creation of these enterprises. Further, the internal mechanisms by which debt financing costs and innovation R&D expenditures influence the impact of carbon risk on enterprise value creation are analyzed separately. Finally, differences in the inhibitory effect of carbon risk on value creation across heterogeneous enterprises are examined. The results show that carbon risk significantly reduces value creation. It raises debt financing costs and diverts resources away from innovation, weakening firms’ capacity to create value. The negative effect is stronger for small firms, non-state-owned firms, and younger firms. The findings provide evidence for policymakers to improve carbon pricing mechanisms, for financial institutions to better assess climate risk, and for firms to develop effective carbon risk management strategies. Overall, the study offers practical implications for promoting a green and low-carbon transition in the real economy.

1. Introduction

Since the Industrial Revolution, the sharp rise in carbon emissions from human activities has become the main driver of global warming. This has resulted in more frequent extreme weather events, rising sea levels, and other serious challenges, posing systemic threats to global economic and social development [1]. The Intergovernmental Panel on Climate Change Seventh Assessment Report (AR7) confirms that climate change is now affecting every region of human habitation. Extreme events such as heatwaves, heavy rainfall, and droughts occur worldwide, while existing models still struggle to capture the rapid pace of climate change. With the signing of the Paris Agreement and China’s global commitment to “peak carbon dioxide emissions before 2030 and strive for carbon neutrality before 2060” (hereinafter referred to as the “Dual Carbon” Goals), enterprises face intensifying climate-related risks and transition pressures [2]. Against this backdrop, “carbon risk” has gradually become a core topic in corporate strategy and financial performance research. It primarily refers to compliance costs, stranded asset risks, and competitive uncertainties confronting enterprises due to tighter climate policies, technological shifts, and evolving market preferences for low-carbon models [3]. Specifically, as the economy and society pursue low-carbon transition, costs stemming from corporate carbon emissions are increasingly internalized. Carbon risk not only directly raises enterprises’ operating and compliance costs [4] but also elevates their financing thresholds and capital costs, creating financial risks [5]. Under such pressures, enterprises must urgently sustain their growth and market standing through environmental management and social responsibility fulfillment to enhance corporate value. Carbon risk encompasses multiple factors that pose latent risks to corporate operations, particularly for high-carbon-emission enterprises. These enterprises must not only navigate stricter carbon emission policies but also adapt to shifting market preferences for low-carbon products and technological disruptions. Thus, carbon risk has become a pivotal factor shaping the value creation of high-carbon-emission enterprises.

Existing literature on carbon risk can be roughly divided into two perspectives: one emphasizes the compliance and operational costs it entails, arguing that carbon regulations increase corporate burdens and reduce short-term financial performance [6]; the other points out that effective carbon risk management can trigger an innovation compensation effect, enhancing enterprises’ long-term competitiveness through green technological innovation and efficiency improvement [7]. However, most existing studies focus on developed country markets or remain limited to evaluating effects at the macro-policy and industry levels. They lack systematic examination of firm-level micro-mechanisms—especially the transmission paths between carbon risk and corporate value under different ownership structures and scale characteristics [8]. Additionally, although carbon risk is a global phenomenon, its manifestations and impact mechanisms may vary significantly across economies with distinctly different institutional backgrounds. As the world’s largest carbon emitter and a representative emerging market, China’s advancement of the “Dual Carbon” goals provides a unique policy laboratory for corporate carbon risk management [9]. Focusing on Chinese high carbon-emission industries is therefore not only of local policy significance but also provides a valuable comparative case for understanding national climate governance and corporate responses.

Based on the aforementioned research gaps, this study uses listed companies in high-carbon-emission industries on China’s Shanghai and Shenzhen A-shares from 2012 to 2022 as the sample. It constructs a carbon risk measurement index based on industrial energy consumption and employs regression analysis to explore in depth the relationship between carbon risk and the value of high-carbon-emission enterprises, as well as the underlying mechanisms. It also aims to expand theories related to environmental regulation and corporate value—particularly from the perspective of the Resource-Based View [10] and Institutional Theory [11]—to reveal the “double-edged sword” effect of carbon risk in a specific institutional context.

The potential contributions of this study are reflected in three aspects: First, it uncovers the dual mechanisms through which carbon risk influences firm value at the micro level, addressing the gap in prior research that emphasizes macro perspectives while overlooking micro-level heterogeneity. Second, it analyzes differences between state-owned and private firms, as well as between large and small firms in China, providing evidence to support the design of more targeted climate policies. Third, it identifies the value relevance of carbon risk and offers empirical evidence for firms to refine low-carbon strategies, for investors to strengthen ESG evaluation, and for policymakers to improve carbon market mechanisms. Overall, the research findings not only help understand how high-carbon-emission enterprises can achieve sustainable value creation under the “Dual Carbon” Goals but also provide insights for emerging economies worldwide to balance economic growth and climate governance.

2. Literature Review

2.1. Definition, Measurement and Economic Impact of Carbon Risk

Research on carbon risk originated from the practical need to address climate change. Early efforts were led by international organizations such as CDP, and academic studies gradually developed a systematic framework.

2.1.1. Definition and Classification System of Carbon Risk

The definition of carbon risk has evolved from macro-level climate risks to firm-level micro risks. The IPCC Fourth Assessment Report (2007) identified six categories of climate-related risks—physical, regulatory, litigation, competitive, production, and reputational risks—which provided the foundation for subsequent corporate-level studies [12]. Hoffmann and Busch (2008) further defined carbon risk as the uncertainty arising from climate change and fossil fuel use [13]. Labatt and White (2007) classified risks into five categories using a firm–industry perspective [14]. Later, the widely used “physical risk–transition risk” framework was introduced [15], with transition risk including policy, technology, and market dimensions [16]. In the Chinese context, Huang et al. (2021) added “institutional adaptability risk”, highlighting the dynamic evolution of carbon risk conceptualization [17].

2.1.2. Carbon Risk Measurement Methods

Currently, no consensus exists within academia regarding carbon risk measurement methodologies. Existing studies primarily adopt two approaches: (1) direct measurement using total carbon emissions or constructed carbon risk indices, and (2) proxy variable methods. For instance, Bose et al. (2021) demonstrated that high-emission firms face greater regulatory penalties and operational risks [18]. However, due to limited disclosure of carbon emission data among Chinese listed companies, domestic scholars frequently employ proxy variables. These include estimating CO₂ emissions based on industry energy consumption [19], or establishing dummy variables for emission violations [16]. Alternative methodologies have emerged to address data limitations. Some researchers utilize quasi-natural experiments, such as Wang Jiaxin et al. (2022) who employed difference-in-differences analysis with the Paris Agreement as a policy shock to examine corporate risk perception [20].

Other innovative approaches include fuzzy comprehensive evaluation models [21], machine learning-based SVM models for risk early-warning systems [22], and the use of text analysis to quantify the climate change news index [23]. These methodological advancements have significantly expanded the toolkit for carbon risk assessment.

2.1.3. The Impact of Carbon Risk

Sorting out the existing literature, carbon risk acts on enterprises through the following channels:

Capital market channel: with the increasing awareness of low-carbon environmental protection, carbon risk will change investors’ judgement and investment choices, thus affecting the performance of enterprises in the capital market [24]. Carbon risk exposure leads to a “carbon premium” in stock prices [25], while stocks of high-carbon emitting firms tend to earn higher returns [3] and ESG preferences reinforce this effect, i.e., an increase in carbon risk prompts investors to reduce their holdings of stocks of high-emitting firms and focus instead on low-emitting firms. emitting firms’ stocks in favour of low-carbon investment opportunities [26].

Business decision-making channel: Servaes and Tamayo (2013) state that corporate executives take into account the environmental aspirations of shareholders and other stakeholders in their decision-making [27]. By doing so, they can increase employee buy-in, build a favourable image of the company and improve its social credibility, thus attracting more high-quality resources. These include innovation incentives [28], asset stranding [29] real surplus management behaviour [30] and blocked cross-border mergers and acquisitions [31]. Notably, there is heterogeneity in the direction of impact: high-carbon firms face downward valuation in the short term [32] but may gain compensatory benefits through green innovation in the long term [28].

2.2. Factors Affecting Enterprise Value Creation

The ultimate goal of a company in its business operations is to create value to the greatest extent possible. Value creation by a company refers to the process where, through coordination and cooperation with upstream and downstream enterprises in the supply chain, a series of production and business activities are carried out to maximize the utilization of corporate resources, ultimately providing users with corresponding products and services, thereby achieving value addition for the company. In the face of constantly changing internal and external environments, scholars have recognized that enhancing value creation capabilities plays a positive role in helping companies gain sustainable competitive advantages and long-term stable economic benefits. However, how to further improve the level of value creation remains a challenge for companies today. Therefore, exploring various factors influencing corporate value creation has become particularly important. This mainly includes two points:

1. Technological Innovation. In the digital economy era, the ability to innovate technologically has gradually become the core competitiveness of enterprises. Enterprises must continuously innovate to ensure they can consistently create value for both themselves and their users. Wang Ping et al. (2022) argue that technological innovation is a crucial factor in corporate value creation [33]. The essence of value creation lies in the company’s ability to transform its current resources or control into future profits. The ultimate goal of technological innovation is to help companies achieve high-quality technological outcomes, thereby enabling them to create greater value using existing resources [33]. Cai Yanzhe et al. (2022) believe that digital technology can fundamentally change the original innovation mode of companies, make innovation plans consistent with market trends in a timely manner, reduce the cost of trial-and-error innovation, help enterprises activate the “multiplier effect” of technological innovation, and then realize the improvement of enterprise value creation ability [34].

2. Corporate Social Responsibility. Existing research on the relationship between corporate social responsibility and value creation presents differing perspectives. In terms of positive effects, companies must properly manage social interests among all parties while actively practicing their social responsibilities, ensuring that shareholders’ investment returns are adequately protected [35]. This will effectively enhance their brand image and social reputation, thereby using this as capital to drive the company to create greater value. Wang Qi et al. (2023) propose a constructive view: companies can refer to the practices of their peers, integrating social responsibility into their own development strategies, thus making positive contributions to the enhancement of corporate value creation capabilities [36]. Particular attention should be paid to the fact that the value assessment of high-carbon emission companies should not be limited to traditional indicators. In addition to financial performance and market competitiveness, non-financial factors such as environmental, social, and governance (ESG) considerations must also be given significant weight, as these factors are becoming increasingly important. Carbon risk, as a crucial environmental factor, can profoundly impact the value creation of high-carbon emission companies by influencing their cost structure, revenue sources, and capital costs.

2.3. The Relationship Between Carbon Risk and Corporate Value Creation

Research on value creation has traditionally focused on areas such as technological innovation [33] and corporate social responsibility [35]. However, in the context of low-carbon transition, new trends have emerged.

2.3.1. Value Contribution from Non-Financial Factors

Environmental, Social, and Governance (ESG) performance has become a critical dimension of value creation [37]. For high-carbon firms, in particular, carbon management capabilities can enhance long-term competitiveness by reducing capital costs or increasing product premiums. Elliot et al. (2014) demonstrated that superior corporate social responsibility (CSR) performance not only improves corporate performance but also enhances reputation [38]. In today’s business environment, a strong reputation serves as vital intangible capital, strengthening core competitiveness, fostering positive consumer relationships, and enabling greater market share capture [39].

2.3.2. Double-Edged Sword Effect of Environmental Regulation

Facilitation Theory: Drawing on Porter’s hypothesis of the “innovation-compensation effect,” scholars argue that well-designed environmental regulations can stimulate corporate innovation and enhance overall development [7]. Recent studies grounded in stakeholder theory [40] suggest that superior CSR performance not only improves corporate reputation but also strengthens financial performance and stakeholder engagement. For instance, investors increasingly incorporate ESG metrics into investment appraisals, motivating firms to prioritize CSR commitments. Gregory (2016) found a positive correlation between CSR performance and stock value [41], while other studies confirm that CSR fulfillment positively influences firm value [37]. Moreover, reasonable carbon constraints can trigger innovation compensation [38] and improve total factor productivity [34].

Inhibition Theory: Rooted in cost theory, this perspective posits that carbon risks and stringent environmental regulations impose significant energy-saving and emission-reduction pressures on firms, increasing environmental governance costs and straining resource availability [42]. Additionally, aggressive policies may lead to production contraction [43] and amplify reputational losses via the “Lucifer effect” [44]. This divergence underscores the critical importance of policy design and corporate adaptability. Furthermore, based on signaling theory, corporate characteristics and behaviors convey critical information to external stakeholders, influencing assessments and decisions [45]. Institutional investors, acting as information intermediaries, may exacerbate the Lucifer effect, damaging corporate reputation. This mechanism can intensify carbon risks and ultimately diminish firm value. For example, Liu et al. (2021) observed a decline in market value among carbon-intensive firms following decarbonization policies in China [32]. Similarly, Choi et al. (2021) found a negative relationship between carbon emissions and market value across 28 countries, even after controlling for voluntary disclosure factors [46].

In summary, the relationship between environmental protection and economic growth—often reflected in corporate value creation at the micro level—remains a contentious topic among scholars. While consensus is lacking, most studies agree that carbon risks have significant economic consequences for firms. Under China’s “dual-carbon” policy, corporate carbon risk has emerged as a pressing issue requiring resolution. Although existing literature predominantly examines macro-level policy effects, few studies explore the internal transmission mechanisms of carbon risk or compare state-owned enterprises (SOEs) with private enterprises (POEs) in China’s institutional context. Achieving low-carbon transformation demands increased investments in energy-saving and emission-reduction technologies, reshaping corporate cost–benefit structures and ultimately value creation. Thus, an in-depth investigation into carbon risk and firm value holds significant theoretical and practical implications.

3. Theoretical Analysis and Research Hypothesis

Carbon risk, a prominent corporate risk dimension under climate change, refers to compliance costs, stranded asset risks, and uncertainties in the operating environment arising from stricter climate policies, technological change, and shifting market preferences [3]. Prior research shows that carbon risk affects not only financial performance but also value creation through multiple channels, particularly in high carbon-emission industries. Yet, its overall impact remains debated. Some studies emphasize rising costs and reduced investment [6], whereas others suggest potential innovation compensation effects that strengthen long-term competitiveness [7]. Based on this, this study constructs a dual-path theoretical framework (Figure 1) to systematically explain how carbon risk influences corporate value creation through the “cost effect” and “innovation compensation effect,” and proposes competing hypotheses.

First, carbon risk constrains enterprises’ value creation capabilities by increasing debt financing costs. Based on agency theory, there exists a conflict of interest between enterprises and creditors amid carbon risk: enterprises may tend to invest in high-carbon projects to pursue short-term profits, while creditors focus on the threat of long-term environmental risks to debt-servicing capacity [47]. This agency conflict leads creditors to demand a higher risk premium from high-carbon enterprises. Specifically, first, according to the Cost Theory, carbon risk raises enterprises’ debt financing costs. Against the backdrop of strengthened carbon regulations (e.g., carbon taxes, carbon emission trading systems), creditors pay greater attention to enterprises’ carbon exposure levels. Studies show that the loan spreads of carbon-intensive enterprises increase significantly as climate policies become stricter. Particularly in China, green credit policies require financial institutions to incorporate carbon information into credit decisions, and the quality of carbon information disclosure is significantly negatively correlated with debt financing costs. If enterprises fail to manage carbon risk effectively, they will face exacerbated financing constraints and rising interest rates. Further, high corporate debt financing costs may hinder value creation. Debt financing elevates a company’s financing costs owing to fixed interest payments. Substantial interest costs constrain corporate cash flow, curtailing reinvestment capacity and consequently suppressing value creation. Additionally, excessive borrowing costs generate investor uncertainty regarding debt repayment capacity, triggering stock price declines and eroding market confidence. This vicious cycle exacerbates financing costs while continuously suppressing value creation.

Second, carbon risk may constrain corporate value creation through increased R&D investment requirements in innovation. Carbon risks originate primarily from exogenous factors, particularly regulatory and policy changes. These external pressures compel firms to allocate additional R&D investment toward innovation. Governments impose carbon emission costs via taxation or trading schemes, mandating firms to pursue emission reduction through technological innovation while complying with stringent standards that necessitate process improvements or low-carbon technology development [7]. The implementation of carbon reduction policies directly increases compliance costs for high-emission companies. Carbon risks necessitate strategic re-evaluation of development models, thereby stimulating internal R&D expenditures. The development of low-carbon technologies enables firms to diminish reliance on high-emission operations while hedging against policy change risks. Low-carbon technologies potentially reduce energy consumption and emission costs, thus decreasing long-term operational expenditures. Consequently, firms must prioritize investments in energy-efficient technologies, carbon allowances, tax obligations, and emission-related R&D to bolster innovation capabilities. Furthermore, resource-constrained firms face crowding-out effects as carbon-reduction investments compete with production, operations, and other innovation funding, ultimately impairing operational efficiency. At the same time, the inherent uncertainty of R&D outcomes may result in non-productive investments, undermining value creation. Additionally, environmental equipment installation incurs learning costs that reallocate human resources from production and R&D activities [48]. Therefore, we propose Hypothesis H1a:

H1a. 

Carbon risk exerts a significant negative impact on the value creation of high-carbon-emission enterprises.

According to Porter’s hypothesis, environmental regulation generates three key effects—first-mover advantage, innovation compensation, and learning effects—all of which contribute to increased enterprise value [7]. For high-carbon-emitting enterprises, carbon risk, as a form of environmental regulatory pressure, can enhance value creation through three distinct pathways:

First, the innovative compensation path. Against the backdrop of the gradual strengthening of carbon constraint policies, the innovation compensation effect emerges as the primary approach for high-carbon-emitting enterprises to manage carbon risk. When policy pressure exhibits a progressive nature, firms can achieve value enhancement through systematic innovation activities. Specifically, companies will augment their R&D investment in low-carbon technologies, encompassing the development of breakthrough technologies like carbon capture and storage (CCUS) technology and the hydrogen smelting process, as well as incremental innovations such as waste heat recovery and energy efficiency enhancement. Such technological innovations not only yield direct cost savings but, more importantly, create green technology barriers that are difficult to replicate. It is noteworthy that substantial industry heterogeneity exists in the innovation compensation effect, and this path holds particular significance in technology-intensive industries. Enterprises in these industries establish a green intellectual property system through patent portfolios, thereby securing new types of value sources, such as technology licensing revenues.

Second, the market premium path. As consumers’ environmental awareness awakens and green consumption preferences take shape, the path of transforming carbon risks into market opportunities has gained increasing significance. This path primarily creates value through three mechanisms: the green brand premium effect, the green supply chain access advantage, and the green financing convenience. It should be emphasized that export-oriented enterprises derive more substantial benefits from this path. Moreover, the effectiveness of the market premium path hinges on enterprises’ signalling capabilities. Therefore, it is of utmost importance to enhance carbon information disclosure and third-party verification.

Third, the asset replacement path. When the structural transformation of energy prices surpasses the tipping point, the asset replacement path serves as the core channel for the transformation of high-carbon enterprises. This path manifests itself in two levels of strategic adjustment: with regard to stock assets, enterprises proactively eliminate high-carbon-locked assets; with respect to incremental investment, enterprises in heavy-asset industries actively deploy new energy businesses. The key to asset replacement lies in timing the technological generational change appropriately. Acting too early may expose enterprises to the risk of technological immaturity during the transition, while acting too late may result in the loss of market opportunities. By eliminating outdated production capacity and deploying new businesses, enterprises can accomplish the value chain leap. This path plays a decisive role in asset-heavy industries.

The aforementioned three paths are not mutually exclusive but exhibit significant synergies. Innovation compensation offers technical support for market premiums, asset replacement liberates resource space for innovation activities, and collectively, these three paths form a dynamic system of carbon risk-driven enterprise value creation. Enterprises at various stages of development may select different combinations of paths: technology-leading enterprises are inclined to favor the innovation compensation-led model, brand-advantaged enterprises concentrate on the market premium path, and traditional asset-heavy enterprises depend more on asset replacement to achieve transformation. This diversity of path choices precisely embodies the rich practical manifestations of Porter’s hypothesis within a carbon-constrained environment. Therefore, we propose hypothesis H1b:

H1b. 

Carbon risk exerts a significant positive impact on the value creation of high-carbon-emitting firms.

4. Research Design

4.1. Sample Selection and Data Sources

This study selects A-share listed companies in high-carbon-emission industries traded on the Shanghai and Shenzhen stock exchanges during 2012–2022 as the research sample. The sample selection follows the CSRC’s Guidelines for Industry Classification of Listed Companies (2012 Edition), covering eight high-carbon industries: Pulp and Paper (C22), Petroleum Refining, Coking and Nuclear Fuel (C25), Chemical Materials and Products (C26), Non-metallic Mineral Products (C30), Ferrous Metal Smelting (C31), Non-ferrous Metal Smelting (C32), Electric Power and Heat Supply (D44), and Gas Supply (D45). The data processing procedure involved: (1) excluding ST and *ST firms; (2) removing observations with missing key variables; (3) winsorizing continuous variables (except dummy variables) at the 1% and 99% percentiles to control for outliers. The final sample comprises 809 unique firms with 5685 firm-year observations. Industry-level main business cost data (for carbon risk calculation) were sourced from the China Industrial Economic Statistical Yearbook, whereas energy consumption data were obtained from the China Energy Statistical Yearbook. CO₂ emissions were estimated using the conversion factor of 2.493 tons CO₂ per ton of standard coal, as specified by the Xiamen Energy Conservation Center. All other variables were collected from the CSMAR database.

4.2. Variable Selection and Definition

1. Dependent Variable: Tobin’s Q (TobinQ). This study employs Tobin’s Q as the measure of corporate value creation, primarily due to its strong theoretical foundation and widespread acceptance in academic research. This indicator comprehensively reflects market expectations regarding a firm’s future growth and profitability [49]. Compared to accounting-based measures such as ROA or ROE, Tobin’s Q adopts a forward-looking market perspective. It incorporates not only current operating performance but also investors’ valuation of the firm’s future cash flows, growth potential, and intangible resources, thereby effectively overcoming the short-term focus and manipulability inherent in accounting metrics [50]. It should be specifically noted that under the “Dual Carbon” policy context, market valuations of high-carbon firms may be subject to policy-induced systematic bias. Therefore, our empirical design controls for year and industry fixed effects to mitigate the influence of general temporal fluctuations and industry heterogeneity. Following the approach of Lu et al. [51], we use Tobin’s Q to capture the market’s overall assessment of the long-term value creation capacity of high-carbon enterprises under the “Dual Carbon” goals, while acknowledging both the unique significance and limitations of this indicator within this specific policy context.

2. Explanatory Variable: Carbon Risk (CRISK). Prior international studies typically measure carbon risk using the carbon-emission-to-revenue ratio. Given data limitations in China’s Carbon Disclosure Project (CDP) and insufficient emission disclosures by listed firms, we implement Shen and Huang’s (2019) [52] methodology, apply estimates of industry-level energy use based on business operating costs: Corporate carbon emissions = Industry total energy consumption × 2.493 (CO₂ conversion coefficient) × (Firm operating cost/Industry operating cost). Carbon risk is then quantified as the natural logarithm of the emissions-to-revenue ratio: CRISK = Ln (Carbon Emissions/Operating Revenue). Higher values indicate greater carbon risk exposure.

3. Control Variables. Consistent with established literature, we control for: firm size (Size), return on equity (ROE), leverage ratio (Lev), book-to-market ratio (BM), operating cash flow (CashFlow), independent director ratio (Indep), listing duration (ListAge), and Big Four auditor dummy (Big4). Furthermore, industry (Ind) and year (Year) fixed effects are included, with detailed definitions provided in

Table 1. Variable definition table.

Type of Variable	Variable Name	Symbol	Definition
explained variable	Value creation for enterprises	TobinQ	Tobin’s Q value
explanatory variable	Carbon risk	CRISK	Ln (enterprise carbon emissions/enterprise operating income)
controlled variable	company size	Size	The natural logarithm of the total assets of an enterprise at the end of the period
	asset-liability ratio	Lev	The total liabilities of the enterprise at the end of the year divided by the total assets at the end of the year
	Return on equity	ROE	The net profit of the enterprise in the current year divided by the shareholders’ equity at the end of the period
	Book market value ratio	BM	The total assets of the enterprise in the current year divided by the market value of the enterprise
	Cash flow from operations	CashFlow	Net cash flow from operating activities/net of total assets
	Proportion of independent directors	Indep	The proportion of independent directors on the board of directors
	Listing period	ListAge	Year of year-year of listing
	Whether there are four major audits	Big4	Enter 1 for the four major audits, otherwise enter 0
	trade	Ind	Virtual variables control for industry fixed effects
	a particular year	Year	Virtual variables to control for fixed effects of years

.

4.3. Model Construction

To empirically test the research hypotheses regarding carbon risk’s impact on high-carbon-emission firms’ value creation, we establish the following fixed-effects model:

TobinQ_i,t = α₀ + α₁CRISK_i,t + αn∑Controls_i,t + Ind + Year + ε_i,t

where TobinQ_i,t Value creation for i enterprises at t period; CRISK_i,t denotes the carbon risk level; Controls_i,t represent the control variables; Ind and Year indicate industry and year fixed effects, respectively; and ε_i,t is the error term. A significantly negative α₁ coefficient would support H1a, while a positive coefficient would validate H1b.

4.4. Descriptive Statistics

Table 2. Descriptive statistics.

Variable	N	Mean	p50	SD	Min	Max
TobinQ	5685	1.815	1.461	1.095	0.818	7.160
CRISK	5685	0.202	0.263	0.434	−1.286	0.924
Size	5685	22.444	22.232	1.384	19.982	26.314
Lev	5685	0.446	0.446	0.209	0.066	0.960
ROE	5685	0.058	0.068	0.154	−0.838	0.524
BM	5685	0.359	0.340	0.166	0.014	0.809
CashFlow	5685	0.058	0.057	0.066	−0.137	0.243
Indep	5685	0.372	0.333	0.050	0.333	0.556
ListAge	5685	2.197	2.485	0.911	0.000	3.367
Big4	5685	0.062	0.000	0.241	0.000	1.000

reports the descriptive statistics for all variables. As evidenced in Table 2, TobinQ exhibits substantial variation, ranging from 0.818 to 7.160. With a mean (median) of 1.815 (1.461) and a standard deviation of 1.095, these metrics reflect pronounced heterogeneity in value creation across high-carbon-emission firms. The core variable CRISK has a mean of 0.202 (SD = 0.434), demonstrating considerable cross-sectional variation in carbon risk exposure. All control variables exhibit values within theoretically expected ranges.

5. Empirical Results and Analysis

5.1. Analysis of the Results of Benchmark Regression

Table 3. Results of the benchmark regression.

	(1)	(2)
VARIABLES	TobinQ	TobinQ
CRISK	−0.688 ***	−0.098 **
	(−8.88)	(−2.50)
Size		−0.124 ***
		(−13.52)
Lev		−3.463 ***
		(−51.65)
ROE		0.326 ***
		(5.73)
BM		−5.766 ***
		(−79.92)
CashFlow		−0.497 ***
		(−3.72)
Indep		0.596 ***
		(3.74)
ListAge		0.131 ***
		(12.67)
Big4		0.155 ***
		(4.39)
Constant	1.216 ***	7.590 ***
	(18.09)	(41.25)
Ind	Yes	Yes
Year	Yes	Yes
Observations	5685	5685
R-squared	0.158	0.705
r2_a	0.155	0.704
F	68.96	521.2

Notes: ***, ** denote statistical significance at the 1%, 5% levels, respectively; t-statistics are reported in parentheses.

presents the regression analysis of the relationship between value creation and carbon risk in high-carbon-emission firms. Column (1) reports the fixed-effects specification, whereas Column (2) incorporates control variables while maintaining the fixed-effects framework. Consistent across specifications, the CRISK coefficient remains significantly negative (p < 0.01), indicating robust evidence that carbon risk adversely affects value creation in high-carbon-emission firms. Economically, a one-standard-deviation increase in CRISK is associated with a 0.098 unit decline in TobinQ, suggesting that firms with greater carbon risk exposure exhibit systematically lower valuation. These findings strongly support H1a, confirming carbon risk as an impediment to value creation among high-carbon-emission firms. Economically, a one-standard-deviation increase in CRISK is associated with a 0.098 unit decline in TobinQ, suggesting that firms with greater carbon risk exposure exhibit systematically lower valuation. These findings strongly support H1a, confirming carbon risk as an impediment to value creation among high-carbon-emission firms. Carbon risk suppresses value creation for high-carbon emission firms mainly due to increasingly stringent global carbon reduction policies and regulations, which increase operating costs. Additionally, the markets’ preference for low-carbon products and companies is shifting, leading to pressure on high-carbon emission firms to face declining market share and rising financing costs. Furthermore, substantial investments in technological transformation and supply chain adjustments, along with the potential risk of “asset stranded,” further weaken these firms’ profitability and long-term competitiveness, thereby suppressing their value creation.

5.2. Robustness Tests

5.2.1. Instrumental Variable (IV) Regression

To address potential endogeneity issues, this study employs the two-stage least squares (2SLS) method to validate the regression model. Following the strategy of Zhu and Zhao [53], we first select the one-period lagged carbon risk as the instrumental variable. The lagged variable is uncorrelated with the current-period error term, satisfying the exogeneity requirement; meanwhile, it is highly correlated with the current-period carbon risk, meeting the relevance condition. Results in Column (1) of

Table 4. Robustness test: instrumental variable method.

	(1)	(2)	(3)	(4)	(5)	(6)
	First Stage	Second Stage	First Stage	Second Stage	First Stage	Second Stage
VARIABLES	CRISK	TobinQ	CRISK	TobinQ	CRISK	TobinQ
IV	0.740 ***		0.335 ***		0.087 ***
	(81.22)		(28.72)		(21.29)
CRISK		−0.153 ***		−0.234 **		−0.229 *
		(−2.63)		(−2.13)		(−1.66)
Size	−0.011 ***	−0.122 ***	−0.035 ***	−0.130 ***	−0.120 ***	−0.099 ***
	(−5.20)	(−11.99)	(−12.28)	(−12.78)	(−25.27)	(−9.16)
Lev	0.131 ***	−3.498 ***	0.397 ***	−3.400 ***	0.428 ***	−3.510 ***
	(8.50)	(−46.54)	(19.28)	(−41.30)	(19.22)	(−37.65)
ROE	−0.169 ***	0.236 ***	−0.161 ***	0.302 ***	−0.253 ***	0.275 ***
	(−13.14)	(3.82)	(−8.97)	(5.06)	(−11.88)	(3.93)
BM	0.164 ***	−5.823 ***	0.360 ***	−5.708 ***	0.380 ***	−5.684 ***
	(9.89)	(−72.57)	(16.10)	(−67.90)	(16.17)	(−62.30)
CashFlow	−0.252 ***	−0.377 **	−0.394 ***	−0.556 ***	−0.578 ***	−0.441 ***
	(−8.03)	(−2.52)	(−9.38)	(−3.95)	(−12.39)	(−3.05)
Indep	−0.017	0.697 ***	−0.072	0.590 ***	−0.018	0.597 ***
	(−0.46)	(4.01)	(−1.42)	(3.71)	(−0.33)	(3.82)
ListAge	0.003	0.140 ***	0.045 ***	0.138 ***	0.057 ***	0.125 ***
	(1.08)	(10.37)	(13.94)	(11.84)	(16.74)	(10.25)
Big4	0.010	0.145 ***	0.016	0.155 ***	−0.016	0.114 ***
	(1.26)	(3.80)	(1.47)	(4.40)	(−1.43)	(3.40)
Constant	0.052	7.627 ***	0.093	7.616 ***	2.692 ***	6.992 ***
	(1.20)	(37.01)	(1.60)	(41.21)	(24.25)	(37.73)
Ind	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Observations	4779	4779	5685	5685	5250	5250
R-squared	0.915	0.709	0.812	0.705	0.785	0.718
Cragg-Donald	6596.735		824.676		453.206

Notes: ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively; t-statistics are reported in parentheses.

show that the current-period and lagged carbon risk are significantly positively correlated at the 1% significance level. Additionally, the Cragg-Donald Wald F statistic is 6596.735, which is much higher than the critical value of 16.38 for the Stock-Yogo test at the 10% significance level, rejecting the weak instrumental variable hypothesis. In the regression of Column (2), the impact of carbon risk (CRISK) on corporate value creation (TobinQ) is significantly negative at the 1% significance level. This indicates that after accounting for endogeneity, carbon risk still exerts an inhibitory effect on corporate value creation.

Nevertheless, we fully acknowledge that relying on a single instrumental variable may pose robustness challenges. To enhance the reliability of our conclusions, we systematically introduce two additional instrumental variables for further testing. First, following Safiullah et al. (2021) [54], we use the average carbon risk of firms in the same region and year as an instrumental variable. This variable captures regional common exposure to environmental regulatory pressures while remaining unrelated to individual firms’ value creation. Second, drawing on Wang and Yu (2023) [55], we employ firms’ payable pollution discharge fees as another instrumental variable. This measure reflects corporate environmental compliance costs while maintaining exogeneity due to its government-mandated standardized calculation method. After re-estimating using 2SLS, both instrumental variables pass relevance and exogeneity tests. As shown in Columns (4) and (6) of Table 4, the coefficients for carbon risk (CRISK) remain significantly negative, consistent with our baseline findings and further supporting the robustness of our results.

In conclusion, after employing multiple instrumental variables and conducting a series of rigorous identification tests, the core conclusion of this study—that carbon risk significantly inhibits corporate value creation—remains robust.

5.2.2. Pandemic Period Exclusion

To isolate the pandemic’s confounding effects, we re-estimate the model after excluding 2020 observations. Column (1) of

Table 5. Robustness test.

	(1)	(2)	(3)
VARIABLES	TobinQ	TobinQ	ROA
CRISK	−0.096 **		−0.026 ***
	(−2.29)		(−10.12)
Carbon		−0.076 ***
		(−3.28)
Size	−0.140 ***	−0.116 ***	0.007 ***
	(−14.43)	(−12.77)	(8.42)
Lev	−3.402 ***	−3.525 ***	−0.107 ***
	(−47.68)	(−54.51)	(−14.06)
ROE	0.359 ***	0.342 ***	0.225 ***
	(6.03)	(6.06)	(21.19)
BM	−5.751 ***	−5.801 ***	−0.039 ***
	(−73.73)	(−82.60)	(−9.57)
CashFlow	−0.549 ***	−0.420 ***	0.153 ***
	(−3.92)	(−3.16)	(13.26)
Indep	0.581 ***	0.589 ***	−0.009
	(3.44)	(3.70)	(−1.00)
ListAge	0.139 ***	0.123 ***	−0.007 ***
	(12.58)	(12.07)	(−11.89)
Big4	0.172 ***	0.154 ***	−0.001
	(4.57)	(4.37)	(−0.51)
Constant	7.911 ***	7.581 ***	−0.075 ***
	(40.58)	(41.24)	(−5.21)
Ind	Yes	Yes	Yes
Year	Yes	Yes	Yes
Observations	5088	5685	5685
R-squared	0.706	0.706	0.704
r2_a	0.704	0.704	0.703
F	485.3	521.8	257.2

Notes: ***, ** denote statistical significance at the 1%, 5% levels, respectively; t-statistics are reported in parentheses.

confirms the persistence of our core findings (coefficient = −0.092, p < 0.01), attesting to result robustness.

5.2.3. Alternative Carbon Risk Measure

To address potential measurement bias, we adopt an alternative carbon risk metric (Carbon) following Wang et al. (2022) [56]. This comprehensive measure accounts for: (1) combustion and fugitive emissions, (2) production process emissions, (3) waste emissions, and (4) land-use change emissions (forest-to-industrial conversion). Data are sourced from corporate annual reports, sustainability disclosures, and other public filings. Column (2) demonstrates Carbon’s significantly negative coefficient (β = −0.117, p < 0.01), corroborating our primary results.

5.2.4. Alternative Performance Measure

Given potential market volatility effects on TobinQ, we substitute our dependent variable with return on assets (ROA) following Ji and Huang (2022) [57]. This alternative specification tests hypothesis H1a using accounting-based performance metrics. Column (3) reveals a significantly negative CRISK-ROA relationship (β = −0.008, p < 0.05), thereby satisfying all robustness checks.

5.3. Mechanism Test

The empirical results of this study indicate that carbon risk exerts a significant inhibitory effect on the value creation of high-carbon-emission enterprises. To further uncover the underlying mechanism at the micro level, this study selects debt financing costs and R&D investment as mediating variables based on theoretical foundations and existing literature, and explores in depth the specific pathways through which carbon risk affects corporate value creation. Theoretical analysis shows that carbon risk may indirectly constrain enterprises’ value creation capabilities through two channels: increasing financing costs and intensifying resource crowding-out. Moreover, this process exhibits heterogeneity due to differences in the nature of enterprises’ ownership.

First, regarding the debt financing cost pathway, based on the asymmetric information and signal transmission theories, enterprises with higher carbon risk are often perceived by the market and creditors as having greater environmental compliance risks and potential transition costs [47]. Such enterprises face more severe financing constraints, with debt financing costs increasing significantly [3]. To maintain stable cash flow, companies may be forced to abandon investment projects with positive net present value, leading to underinvestment and inhibiting long-term value creation.

Second, regarding the R&D investment pathway, based on the resource constraint theory, enterprises under carbon emission reduction pressure have to allocate their limited financial and human resources to activities such as low-carbon technology R&D, equipment upgrading, and carbon emission quota purchases [58]. Although such investments contribute to long-term transition, they crowd out R&D resources allocated to core businesses and efficiency improvement in the short term. This is particularly true when emission reduction investments fail to generate economic benefits in a timely manner, which will have a negative impact on enterprises’ current-period value creation.

It is worth noting that although the mediation effect model is widely used for mechanism testing, Dell (2010) [59] argues that judging the existence of a mediation effect by observing changes in the coefficient of the independent variable after adding the mediating variable to the regression of the dependent variable on the independent variable may lead to erroneous conclusions. Thus, we formally test two transmission channels: (1) debt financing costs and (2) innovation expenditures, using the following mediation models:

Cost_i,t/RD_i,t = β₀ + β₁CRISK + β_n∑ Controls_i,t + Ind + Year + ε_i,t

Following Zhou et al. (2017) [60], we measure debt financing cost (Cost) as the ratio of annual financial expenses to year-end total liabilities. Consistent with Quan and Yin (2017) [61], innovation R&D cost (RD) is calculated as R&D expenditure scaled by beginning-period total assets. All control variables remain identical to the baseline specification.

Table 6. Mechanism test results.

	(1)	(2)
VARIABLES	Cost	RD
CRISK	0.018 ***	0.003 ***
	(7.35)	(3.52)
Size	0.001	−0.001 ***
	(1.03)	(−4.10)
Lev	0.085 ***	−0.011 ***
	(19.89)	(−8.76)
ROE	0.007 **	0.005 ***
	(2.03)	(4.83)
BM	0.019 ***	−0.010 ***
	(4.08)	(−7.09)
CashFlow	0.007	0.014 ***
	(0.85)	(5.49)
Indep	−0.015	−0.002
	(−1.45)	(−0.77)
ListAge	0.001	−0.003 ***
	(0.89)	(−13.21)
Big4	−0.007 ***	0.001
	(−3.13)	(1.16)
Constant	−0.035 ***	0.045 ***
	(−2.98)	(12.65)
Ind	Yes	Yes
Year	Yes	Yes
Observations	5685	5685
R-squared	0.208	0.343
r2_a	0.205	0.340
F	57.28	113.6

Notes: ***, ** denote statistical significance at the 1%, 5% levels, respectively; t-statistics are reported in parentheses.

reports the mediation analysis results for debt financing costs and innovation expenditures. As shown in Table 6, CRISK exhibits significantly positive coefficients (p < 0.01) with both Cost (β = 0.082) and RD (β = 0.015), confirming that carbon risk elevates financing costs while crowding out R&D resources—both pathways suppress value creation. These findings empirically validate our hypothesized transmission mechanisms.

5.4. Heterogeneity Analysis

To examine in depth the heterogeneous impacts of carbon risk on corporate value creation, this study conducts analysis from three theoretical dimensions—firm size, ownership type, and firm maturity—based on the Resource-Based View and Institutional Theory. These three dimensions respectively reflect differences in firms’ resource capabilities to cope with risks, the institutional pressures and support they face, and their organizational rigidity and learning capabilities. Together, they provide an important theoretical perspective for understanding the boundary conditions under which carbon risk exerts its effects.

5.4.1. Firm Size Heterogeneity

This study divides the sample into two groups (large-scale and small-scale enterprises) based on the median of total assets. A comparison of the two regression results in Column (1) of Table 6 shows that the coefficients of carbon risk (CRISK) on corporate value creation (TobinQ) are −0.070 for large-scale enterprises and −0.134 for small-scale enterprises, both indicating a significant negative correlation. Additionally, the Chow test is significant at the 1% level (p = 0.000 < 0.01), indicating that the coefficient differences across groups are statistically significant. A comparison of the coefficients reveals that the impact of carbon risk on high-carbon-emission enterprises varies significantly across firms of different sizes—specifically, the inhibitory effect of carbon risk on corporate value creation is stronger for small-scale enterprises. From the perspective of the Resource-Based View, firm size serves as a key proxy variable for resource endowments and capabilities. Large enterprises typically demonstrate stronger risk resilience due to their superior asset collateral capacity, greater information transparency, and more abundant redundant resources. Within China’s institutional context, this size advantage is further amplified. In the bank-dominated financial system, large firms enjoy preferential access to green credit supported by their substantial assets and stable cash flows. Moreover, they exercise greater influence in carbon market development and quota allocation processes, often participating in policy-making as industry representatives to secure favorable transition conditions. In contrast, small and medium enterprises face not only inherent financing constraints but also institutional marginalization in policy implementation, significantly exacerbating their vulnerability to carbon risk.

5.4.2. Ownership Type Heterogeneity

We stratify the sample by ownership type—state-owned enterprises (SOEs) versus non-SOEs—to examine institutional heterogeneity in carbon risk effects.

Table 7. Results of heterogeneity analysis.

	(1)		(2)		(3)
	Large-Scale Corporation	Small Businesses	State-Owned Enterprises	Non-State-Owned Enterprises	Older Businesses	Young Enterprise
VARIABLES	TobinQ	TobinQ	TobinQ	TobinQ	TobinQ	TobinQ
CRISK	−0.070 **	−0.134 **	0.062	−0.243 ***	−0.045	−0.123 ***
	(−2.26)	(−2.00)	(1.11)	(−4.40)	(−0.73)	(−2.61)
Size	−0.013	−0.381 ***	−0.102 ***	−0.148 ***	−0.203 ***	0.026 *
	(−1.61)	(−14.15)	(−8.82)	(−9.90)	(−16.08)	(1.95)
Lev	−3.664 ***	−3.053 ***	−3.758 ***	−3.104 ***	−3.100 ***	−4.076 ***
	(−54.73)	(−30.87)	(−39.90)	(−32.73)	(−32.40)	(−45.08)
ROE	0.122 **	0.338 ***	0.396 ***	0.205 **	0.338 ***	0.072
	(2.47)	(3.81)	(5.49)	(2.35)	(4.87)	(0.72)
BM	−4.597 ***	−6.312 ***	−5.484 ***	−5.987 ***	−5.234 ***	−6.404 ***
	(−70.48)	(−51.95)	(−50.96)	(−60.07)	(−47.44)	(−70.18)
CashFlow	−0.084	−0.457 **	−0.737 ***	−0.281	−0.813 ***	−0.034
	(−0.73)	(−2.17)	(−3.93)	(−1.52)	(−4.12)	(−0.20)
Indep	0.203	0.496 *	0.845 ***	0.166	0.874 ***	0.240
	(1.58)	(1.89)	(3.77)	(0.74)	(3.71)	(1.21)
ListAge	0.011	0.153 ***	0.088 ***	0.131 ***	0.255 ***	0.039 **
	(1.11)	(9.25)	(4.60)	(8.83)	(4.98)	(2.47)
Big4	0.070 ***	−0.005	0.134 ***	0.107 *	0.203 ***	0.002
	(3.18)	(−0.04)	(3.29)	(1.71)	(4.39)	(0.04)
Constant	5.242 ***	13.253 ***	7.263 ***	8.400 ***	8.701 ***	5.158 ***
	(29.21)	(24.08)	(30.34)	(27.49)	(29.19)	(19.27)
Ind	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Observations	2846	2839	2512	3173	3007	2722
R-squared	0.745	0.703	0.713	0.700	0.672	0.771
r2_a	0.742	0.700	0.710	0.698	0.669	0.769
F	740.7	569.0	541.5	674.5	550.7	811.0

Notes: ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively; t-statistics are reported in parentheses.

, Column (2) demonstrates carbon risk’s significant negative impact on non-SOEs (β = −0.121, p < 0.01) but insignificant effect on SOEs (β = −0.032, p > 0.10). Institutional Theory emphasizes that organizational behavior is deeply embedded in its institutional environment. Within China’s institutional context, state-owned enterprises (SOEs) benefit from inherent political connections that create distinct advantages in managing carbon risk. SOEs receive prioritized access to low-carbon transition subsidies, specialized technology funds, and green project support. They typically benefit from extended transition periods and adaptive compliance arrangements, while their non-economic objectives provide additional buffers against carbon risk impacts. Furthermore, SOEs enjoy preferential treatment in green credit approval and bond issuance processes. Conversely, non-SOEs face substantial institutional disadvantages, including limited policy support, stricter regulatory enforcement, and financing discrimination. This institutional asymmetry significantly increases their vulnerability to carbon risk and constrains their value creation capacity.

5.4.3. Corporate Maturity Heterogeneity

Using median listing duration as the threshold, we classify firms as mature (>median) or young (≤median). Table 7, Column (3) indicates stronger value destruction for young firms (β = −0.148, p < 0.01) relative to mature firms (β = −0.057, p < 0.05). From an organizational learning perspective, established firms possess accumulated environmental management experience and regulatory adaptation capabilities through long-term “learning by doing.” Their mature organizational routines and relational networks provide natural buffers against carbon risk. Within China’s institutional context, this age advantage is further reinforced. Older enterprises, as stable “insiders” within the system, have developed long-term relationships with regulatory bodies, enabling better anticipation and adaptation to policy changes. Their systemic importance in maintaining regional economic and social stability also ensures greater regulatory attention and support. In contrast, younger firms face dual challenges: not only do they lack experience and organizational capabilities, but their peripheral position in the industrial ecosystem limits their access to institutional support. This “new entrant disadvantage” significantly increases their vulnerability to carbon risk pressures.

6. Conclusions, Policy Implications and Prospects

6.1. Research Findings and Theoretical Discussion

In recent years, with the growing severity of global climate change and the deepening implementation of the “Dual Carbon” goals, carbon risk has become a critical factor that cannot be ignored in the process of corporate value creation. Using a sample of listed companies in high-carbon industries on China’s Shanghai and Shenzhen A-shares from 2012 to 2022, this study empirically examines the impact mechanism of carbon risk on corporate value creation by constructing a comprehensive carbon risk evaluation indicator system and employing a two-way fixed effects model. The findings reveal several key insights:

First, carbon risk exerts a significant inhibitory effect on value creation in high-carbon-emitting enterprises. This conclusion remains robust after a series of robustness tests, indicating that under the current institutional environment and technological conditions, carbon risk manifests more as an operational burden than an innovation opportunity for firms.

Second, the inhibitory effect of carbon risk demonstrates notable heterogeneity. Specifically, non-state-owned enterprises (NSOEs) experience more pronounced negative impacts on value creation due to stronger financing constraints and limited policy support. Small-sized firms, constrained by limited resource reserves and weaker risk resilience, are at a relative disadvantage when coping with carbon risk. Meanwhile, younger enterprises, lacking experience and transformation capabilities, are more vulnerable to the shocks induced by carbon risk.

Notably, the competitive hypothesis H1b derived from the Porter Hypothesis—that carbon risk promotes value creation—lacks empirical support in this study. This finding diverges from classical theoretical expectations and particularly reveals the boundaries of the Porter Hypothesis’s applicability within China’s institutional context, carrying profound theoretical and practical implications:

1. Institutional Environment and Market Mechanism Constraints: The effectiveness of the Porter Hypothesis relies on a well-developed institutional environment. In the early stages of China’s carbon market development, the price discovery function of carbon emission rights remains underdeveloped, with persistently low carbon prices failing to provide effective innovation incentives for firms. Concurrently, supporting mechanisms such as green technology trading markets and green financial systems are still under construction. Consequently, even if firms engage in low-carbon innovation, they struggle to attain adequate economic returns through market channels. Particularly for state-owned enterprises (SOEs), soft budget constraints and policy-imposed burdens further dilute the incentivizing effect of carbon price signals, hindering the manifestation of the “innovation compensation effect.”

2. Structural Differences in Firm Capabilities and Transformation Motivation: This study finds that while the inhibitory effect of carbon risk is relatively weaker in SOEs and large firms, the promoting effect remains insignificant. This insight is highly revealing: although SOEs possess stronger risk resilience due to resource advantages and policy access, their unique governance structures and evaluation systems often lead them to adopt “compliance-oriented” rather than “innovation-oriented” strategies. Relatively conservative innovation cultures and lengthy decision-making chains constrain their ability to transform environmental pressures into innovation opportunities.

3. Timing Mismatch Between Costs and Benefits: Compliance costs and equipment upgrade expenses triggered by carbon risk are immediate, whereas returns on innovation investments are characterized by significant time lags and uncertainty. Faced with short-term performance assessments and financing constraints, firms tend to perceive carbon risk as a cost item rather than an investment opportunity. This decision-making preference further weakens the manifestation of the Porter Hypothesis effect.

4. Regional Disparities in Policy Enforcement: Significant regional variations exist in the enforcement of environmental regulations in China, which influences the emergence of the Porter Hypothesis effect. In regions with weaker environmental enforcement, firms tend to adopt end-of-pipe treatments rather than fundamental innovation. Conversely, in regions with stricter enforcement, firms often face urgent compliance pressures, leaving limited time windows for strategic innovation.

These findings not only explain why the Porter Hypothesis fails to fully materialize in the Chinese context but, more importantly, deepen our understanding of the synergistic effects among institutional environments, firm characteristics, and innovation policies. Particularly in transitional economies, the innovation-promoting effect of environmental regulations needs to be coupled with deeper institutional reforms and corporate governance improvements.

6.2. Policy Implications

Based on in-depth analysis of the research findings—particularly the inhibitory effect of carbon risk on corporate value creation and the limited applicability of the Porter Hypothesis in the Chinese context—this study proposes the following targeted policy recommendations.

First, it is essential to improve carbon pricing mechanisms and establish a multi-layered incentive system. This can be achieved by expanding the coverage of the existing carbon market to include high-emitting sectors such as steel and cement, while gradually reducing the proportion of free allowances to enhance the effectiveness of carbon price signals. For non-state-owned and small-scale enterprises that are more significantly affected by carbon risk, a dedicated green transition fund should be established to provide low-interest loans and technical subsidies, alleviating their financing constraints and transition pressures. Such differentiated policy design ensures balanced transmission of emission reduction pressures while providing appropriate buffering space for different types of enterprises.

Second, institutional environment building should be strengthened to transform compliance costs into innovation drivers. Given that carbon risk currently manifests more as a “cost effect” rather than an “innovation compensation effect,” supporting institutional development is crucial. It is recommended to establish a corporate carbon risk rating system linked to financing costs and market access, creating competitive advantages for firms that proactively engage in low-carbon innovation. Simultaneously, improvements in the green technology trading market and intellectual property protection will help ensure that enterprises obtain due economic returns from their low-carbon innovations, thus shifting corporate strategies from passive compliance to active innovation.

Third, differentiated guidance policies should be implemented to precisely support corporate green transition. Considering the heterogeneous impact of carbon risk, policy formulation must fully account for variations among enterprise types. For state-owned enterprises, their demonstrative role in green transition should be strengthened by incorporating carbon performance into executive evaluation systems. For non-state-owned and small-scale enterprises, targeted technical assistance and fiscal incentives should be provided to enhance their carbon risk management capabilities. For younger firms, green entrepreneurship support programs could be established to encourage low-carbon development from their early stages.

Fourth, green market demand should be cultivated to create value realization channels for low-carbon products. Measures such as low-carbon product certification and green consumption subsidies are recommended to stimulate market demand. Meanwhile, governments should prioritize low-carbon products in public procurement, creating institutional conditions for firms to gain market premiums through green innovation. This combination of demand-side pull and supply-side push will help form a virtuous cycle of “innovation–profit–re-innovation,” establishing the market foundation for the Porter Hypothesis to take effect.

6.3. Research Deficiencies and Prospects

This study has several limitations that warrant further exploration.

First, in terms of variable measurement, it should be noted that estimating firm-specific carbon risk using industry-level data entails certain measurement inaccuracies. Due to variations across firms in the same industry—such as differences in energy efficiency, adoption of clean technologies, and progress in low-carbon transition—an industry-average-based approach may not fully capture firm-level heterogeneity in carbon risk exposure. This limitation primarily stems from the lack of detailed, firm-level carbon emission disclosure in China. Future studies could develop more accurate measures if finer-grained, firm-specific emissions or energy use data become available.

Second, regarding research design, although a two-way fixed effects model was employed to control for endogeneity, the potential bidirectional causality between carbon risk and firm value cannot be completely ruled out. For example, firms with higher value may be better able to invest in low-carbon technologies, thereby lowering carbon risk. Future studies could address this issue by exploiting exogenous policy shocks as instrumental variables or applying causal identification strategies such as regression discontinuity design to provide more robust evidence.

Third, in terms of mechanism testing, this study focused on financial (debt financing costs) and innovation (R&D investment) channels. However, carbon risk may also affect firm value through broader mechanisms such as corporate governance, supply chain dynamics, or consumer behavior. Future research could apply moderating effect models or multidimensional panel data methods to examine these channels in greater depth.

Looking ahead, future work could advance in several directions. First, strengthen causal identification by leveraging natural experiments, such as carbon market expansion policies. Second, explore the role of digital technologies in monitoring and managing carbon risk. Third, conduct cross-country comparative studies to investigate how institutional contexts shape the value relevance of carbon risk, thereby informing the development of a global sustainable financial system. Moreover, as China advances its national carbon market and international policies such as the European Union’s Carbon Border Adjustment Mechanism take effect, issues such as cross-border carbon risk transmission and innovation in carbon financial products also merit continued scholarly attention.