Operational Efficiency and Debt Cost: The Mediating Effect of Carbon Information Disclosure in Chinese Listed Companies

Abstract

Voluntary corporate carbon information disclosure not only meets the carbon information needs of investors, but also enhances the legitimacy of organizations. Building a green image may increase the public consumption of products and be conducive to reducing the cost of debt. As a part of creditors’ assessment of debtors’ solvency, operational efficiency is the basic factor of debt cost reduction. By constructing the correlation between carbon information disclosure and operational efficiency, this paper attempts to test the mediating effect of carbon information disclosure in the relationship between operational efficiency and debt cost, so as to increase the understanding of the mechanism of action between carbon information disclosure and debt cost. Stepwise regression method and Bootstrap statistical method were used to test. The results show that the higher the operational efficiency, the lower the debt cost; the carbon information disclosure of enterprises in low-carbon industries has a significant partial mediating effect on the relationship between operational efficiency and debt cost while that of carbon-intensive industries is not significant. It verifies that the operating efficiency of low-carbon industry can affect the cost of debt through carbon information disclosure, and finds a significant correlation between carbon information disclosure and operating efficiency. This study broadens ways for enterprises in low-carbon industries to reduce debt costs, highlights the role of operational efficiency in various industries, and reveals favorable evidence in the positive value of carbon disclosure in low-carbon industries which in essence can better reflect the enthusiasm of enterprises in their own financing or the restrictive channels of enterprise carbon performance evaluation. This has implications for the research on how to promote the link between carbon information disclosure and debt cost of enterprises in high-carbon industries, which will help enterprises in the future to make carbon information transparency or voluntarily disclose carbon information. It is of great significance for regional and industrial enterprises to choose the disclosure system of voluntary or mandatory carbon information disclosure.

1. Introduction

The increase of debt cost will lead to the debt default risk of enterprises, which will be reflected in the additional investment risk compensation demanded by creditors, thus resulting in a vicious circle of corporate financial risk [1]. In addition, the increase of debt cost may have a significant impact on investment strategies, investment decisions and sustainable development of enterprises [2]. Under the impact of the global COVID-19 pandemic, the risk of debt default in the real economy has obviously accumulated. However, under the condition of balanced operating capital payment, the debt default boundary is jointly determined by operating efficiency, borrowing rate and capital structure [3]. Operation efficiency management is the key strategy of enterprise development, mainly to create shareholder value as the ultimate goal. However, operational efficiency exerts its own utilization value in the operation of enterprises and affects the business performance of enterprises [4]. Improving the operation efficiency management can increase profitability and profit quality, thus having a significant impact on the cost of debt [5].

As environmental issues brought by economic development have increasingly become a hot topic discussed by scholars and a common challenge faced by countries around the world, enterprises and creditors have begun to pay attention to the climate change risk [6]. The carbon information disclosure enables stakeholders to obtain more information, so as to promote the smooth progress of carbon emission trading, obtain excess returns and enhance corporate value [7]. Climate-related information disclosure can reduce the information asymmetry between enterprises and creditors, thus reducing the capital cost for enterprises [8]. Therefore, in combination with scholars’ studies on the relationships among operational efficiency, carbon information disclosure and enterprise value, operational efficiency and debt cost, and carbon information disclosure on debt cost, this study assumes the relationship among operational efficiency, carbon information disclosure and debt cost: If operational efficiency can affect the debt cost, does carbon information disclosure play an intermediary role? In addition, compared with enterprises in low-carbon industries, the value of enterprises in high-carbon industries is more affected by carbon emission trading, and their market value will be further reduced [9]. Both the green credit policy and the development of green credit increase the debt financing cost of “high-pollution and high-emission” enterprises, but reduce the debt financing cost of green enterprises [10]. This study argues that the operating efficiency and enterprises carbon information disclosure in different carbon-intensive industries may have different impacts on corporate debt costs. The following research questions are put forward: 1. Does operational efficiency have a positive impact on carbon information disclosure? 2. Is there a significant positive correlation between operational efficiency and carbon information disclosure? 3. Does carbon information disclosure play an intermediary role in the relationship between operational efficiency and debt cost?

The results show that the higher the operating efficiency of high and low carbon enterprises, the lower the debt cost. The enterprises carbon information disclosure in low-carbon industries has a significant partial mediating effect on the relationship between operational efficiency and debt cost. The mediating effect of carbon information disclosure on the relationship between operational efficiency and debt cost in carbon-intensive industries is not significant. It verifies that the operating efficiency of low-carbon industry can affect the cost of debt through carbon information disclosure, and finds a significant correlation between carbon information disclosure and operating efficiency. This study broadens ways for enterprises in low-carbon industries to reduce debt costs, highlights the role of operational efficiency, and reveals the significant effects of carbon information disclosure. This has implications for the research on the relationship between operational efficiency, debt cost of different carbon-intensive enterprises and carbon information disclosure, especially the complex relationship between operational efficiency and debt cost of carbon-intensive industries.

2. Literature Review

2.1. Operational Efficiency and Cost of Debt

Operational efficiency, as part of the creditor’s assessment of the debtor’s ability to repay debts, is the basic factor for debt cost reduction, so it is necessary to grasp the balance between operational efficiency and debt cost. In terms of influencing factors, Rey et al. believe that the cost advantage is a factor that directly affects the rapid improvement of operational efficiency of Asian airlines [11]. Studies such as that by Barros et al. show that a company’s operational efficiency is affected by the size of the company, mergers and acquisitions, and timing [12]. In terms of research methods, Saranga and Nagpal determined the driving factors of operational efficiency and carried out factor analysis through the DEA model combined with GLS regression and the Tobit model [4]. Ramachandran and Janakiraman comprehensively calculated operational efficiency through the performance index, efficiency index and utilization index [13]. In addition, Seth et al. used the data envelopment analysis method to evaluate the relative effectiveness of multi-index input and multi-index output [14].

Low-cost debt financing is the core factor for enterprises to have competitive advantages. The existing research on the cost of debt mainly analyzes the factors affecting the cost of debt from three aspects: the quality of information disclosure, the characteristics of the external environment and the characteristics of internal governance. Since information asymmetry is an important factor affecting the cost of debt, low-quality environmental information disclosure will lead to an increase in corporate debt costs [15]. The industry environment characteristics [1], legal system [16] and media reports [17] will all have an impact on the cost of debt of enterprises. From the perspective of the degree of competition in the industry environment, a good competitive environment in the industry environment is conducive to reducing debt costs, and this effect is more obvious in private enterprises. In terms of internal governance characteristics, previous studies have mainly focused on the role of institutional arrangements [18], internal culture [19] and senior management characteristics [20] on debt costs. There are few studies on the impact of operating efficiency on debt costs, and the research content is mainly focused on internal control defects. The lower the enterprise’s operating efficiency, the greater the financial exposure, and the more sensitive the debt cost is to the company’s performance [21]. Therefore, reasonable debt cost is an indispensable core link to improve operational efficiency. Although there are many studies on the cost of debt at present, the research on the relationship between corporate operating efficiency and debt cost is still in the exploratory stage.

2.2. Operational Efficiency and Corporate Performance

Enterprise performance is the real result of enterprise operation, which is usually measured by the unit input cost or the profit brought by the unit output that reflects the production result [22]. As an important indicator to measure the financial efficiency and corporate performance of an enterprise, operational efficiency represents the relationship between liquidity and profitability [23,24,25]. However, blindly pursuing operational efficiency will also reduce actual business efficiency and further reduce enterprise performance [26]. Therefore, how to measure the impact of an enterprise’s operational efficiency on enterprise performance has become a hot topic in academic circles. The existing research mainly analyzes the profitability of listed companies through financial reports, and further measures the impact of corporate operating efficiency on company performance through financial indicators such as total asset turnover rate, current asset turnover rate and accounts receivable turnover rate [27].

However, scholars hold different views on the specific impact of operational efficiency on corporate performance. Some scholars have found that the total asset turnover rate [28], current asset turnover rate [29] and accounts receivable turnover rate [30] have a positive impact on corporate profitability and further have a positive effect on corporate performance. On the contrary, another group of scholars has found that there is a negative correlation between corporate operating efficiency and corporate performance. In terms of profitability, the average payback period, inventory turnover rate and cash cycle are negatively related to corporate performance [31,32,33]. Furthermore, some studies show that asset liquidity, cash turnover and accounts receivable turnover do not significantly affect the profitability and performance of enterprises [34]. Operational efficiency is the internal performance of the company’s operation, while corporate performance is the external performance of the company’s operation. The relationship between the two needs to be further confirmed and studied.

2.3. Corporate Performance and Carbon Information Disclosure

In recent years, with the enhancement of environmental regulations, carbon information disclosure, as an important way for enterprises to transmit their carbon performance and carbon emission management to the outside world, has been highly valued by external stakeholders and has become an important issue in the field of economic management. By sorting out the existing literature, the research on carbon information disclosure mainly focuses on the relationship between corporate environmental responsibility and corporate performance. A review of previous literature shows that factors that significantly affect corporate carbon information disclosure include corporate size, institutional shareholding, corporate performance and social reputation [35]. Most of the research results show that corporate performance will be positively affected by carbon information disclosure [36]. Some studies show that corporate carbon information disclosure will reduce investors’ uncertainty [37], alleviate financial constraints [38], obtain practical economic benefits and further improve corporate performance [39], and this promoting effect is more obvious in non-state-owned enterprises and enterprises with high corporate governance level [40].

However, some scholars have found that carbon information disclosure is negatively related to corporate performance. As enterprises fulfill their environmental responsibilities, they will inevitably pay certain corporate costs, which will reduce corporate performance. Therefore, the disclosure of carbon information is not conducive to the improvement of corporate performance [41]. In addition, the disclosure cost is relatively high. This measure cannot provide good economic benefits for related enterprises, even reduces the profitability of enterprises, and further hurts corporate performance [42]. Sheng et al. also found that carbon information disclosure will inhibit the realization of corporate value under the product market after a series of research and analyses [43]. In summary, the relationship between carbon information disclosure and corporate performance is still worth further exploration under different conditions.

2.4. Carbon Disclosure and Cost of Debt

The research on the relationship between information disclosure and debt cost initially extended from accounting information disclosure to social responsibility information disclosure and environmental information disclosure. With the development of China’s carbon trading market, the research on carbon information disclosure has gradually deepened. The existing literature mainly explores the relationship between carbon information disclosure and corporate value, financial performance and capital costs. Research on the relationship between carbon information disclosure and corporate debt costs is still lacking in depth.

The concept of carbon information disclosure was first proposed by Clarkson et al. [44]. Scholars divide companies into two types according to the degree of environmental information disclosure. Companies that strictly abide by environmental information disclosure increase the cost of competitor information disclosure, while companies that minimize environmental information disclosure expenditures do not enjoy the benefits of carbon information disclosure. With the improvement of environmental regulations, creditors will increase their concerns about corporate carbon risks and factor climate change risks into their credit decisions [45]. Subsequently, companies that choose to voluntarily disclose carbon emissions enjoy more favorable loan terms than companies that do not disclose carbon emissions [38].

Some companies have lower financing costs, but their carbon information disclosure shows a higher level [46]. Scholars have begun to analyze the internal mechanism of the above phenomenon: Lemma et al. found that the improvement of carbon information disclosure can effectively improve the transparency of enterprises, effectively alleviate the degree of information asymmetry between enterprises and creditors and enable creditors to have a deeper understanding of the carbon performance of enterprises, thus reducing financing costs [47]. In addition, Fonseka et al. found that environmental information disclosure helps to establish a good image for enterprises to comply with environmental laws and regulations, send signals to creditors that their production and business activities have obtained legal status, enhance creditors’ confidence in the legal operation of enterprises and thus reduce debt costs [48]. Palea and Drogo also found that high carbon emissions were associated with high debt financing, and the improvement of carbon information disclosure played a role in mitigating the cost of debt [8].

2.5. Operational Efficiency, Carbon Disclosure and Cost of Debt

Under the concept of global governance of advocating the awareness of a community with a shared future for mankind, the national low-carbon strategy and green credit policy are continuously being carried out in depth. To reduce their indirect risks, investors use the information disclosed by corporate carbon information as an important reference for financing decisions [38]. Capital is the source of sustainable development of enterprises, and debt financing is one of the main channels for enterprises to obtain external capital. How to improve enterprises’ financing ability and reduce debt costs while complying with national policies and undertaking social responsibility, to achieve win-win results in enterprise operation benefits, environmental benefits and economic benefits, is an important research topic.

However, in terms of the research on the quality of carbon information disclosure, at the beginning of the period, limited carbon information disclosure usually hinders investors from making a comparison [49], and some enterprises are often not willing to take the initiative to disclose information due to light environmental pollution. In the long run, it may aggravate the information asymmetry inside and outside the enterprise, and the external stakeholders of the enterprise cannot accurately judge the future operating benefits and risks of the enterprise, thus leading to adverse selection and further increasing the debt cost of the enterprise [50].

Under the threat of global warming, more and more attention has been paid to the guarantee of carbon information disclosure [51], but there is still a lack of consistency in the indicators of quantitative carbon information disclosure. Some scholars conduct research through CDP’s carbon disclosure leadership index [52], other scholars develop carbon disclosure index according to the carbon reporting framework [53] and some scholars evaluate through carbon emission level [54].

At present, there have been preliminary studies on the relationship among corporate operating efficiency, carbon information disclosure, and cost of debt in academic circles at home and abroad, but there is a lack of research on the relationship between carbon information disclosure and the cost of debt through operating efficiency. Therefore, considering the relationship among corporate operating efficiency, carbon information disclosure and cost of debt can enrich the research on the relationship among the three.

In summary, the research progress of the relationship among operational efficiency, carbon information disclosure and cost of debt is summarized. However, there are differences between the conclusions found in the study, which may be caused by different research samples, research methods and research indicators. In addition, there are few studies on the relationship among operational efficiency, carbon disclosure and cost of debt. By examining the mediating effect of carbon disclosure in the relationship between operational efficiency and debt cost, this study attempts to overcome the above-mentioned defects, and use the important influencing factor of operational efficiency to explore the mechanism between carbon information disclosure and debt cost, to enrich the research on carbon information disclosure.

3. Research Hypothesis

In recent years, with the increase of carbon emissions year by year, climate extreme events such as global warming and glacier melting have revealed deep-seated problems for human beings and the environment. As a major participant in market operations, enterprises should bear the social responsibility of environmental protection, energy conservation and emission reduction. Especially under the promotion of the carbon peak and carbon neutrality policy, enterprises in both carbon-intensive and low-carbon industries are facing a series of regulations and social public opinion pressure. In addition to fulfilling social obligations, enterprises also need to deal with issues such as enterprise operation efficiency and debt financing. Operational efficiency is a part of the creditor’s evaluation of the debtor’s solvency, and the relationship between the two is worth exploring.

According to the stakeholder theory, the production and operation of an enterprise are closely related to its shareholders and creditors. In the process of the whole life cycle management of internal assets, if the enterprise does not pay enough attention to some assets, the operation efficiency is low and ineffective assets occupy more, which will further affect the creditor’s investment decision-making and easily affect the financing cost of the enterprise. Specifically, companies operate less efficiently, and investors will underestimate their value as a signal of high risk, thus increasing the difficulty of corporate financing and further increasing the cost of debt [55]. However, some studies believe that starting from the preferred order financing theory, higher operating efficiency of an enterprise means having high liquidity assets, which can be understood as hedging of debts in essence [56]. Considering practical value and potential need from investors, this study tends to value rather than undermine liquidity assets, regardless of industries. Therefore, based on the above analysis, this article proposes Hypotheses 1 and 2:

H1: 

In carbon-intensive industries, the higher the operational efficiency, the lower the cost of debt.

H2: 

In low-carbon industries, the higher the operational efficiency, the lower the cost of debt.

As climate change has received more and more research and policy attention, the public’s understanding of the carbon peaking and carbon neutrality goals has been deepened; the concept of green, low-carbon and circular economy has been deepened; and the level of carbon information disclosure has become a powerful indicator to measure the response of enterprises to climate change [57]. The study found that enterprises in carbon-intensive industries are more inclined to provide high-quality carbon information disclosure levels. The main reasons are as follows: First, if enterprises in carbon-intensive industries cannot disclose high-quality carbon information, creditors will think that the enterprise’s operating efficiency is poor and there is investment uncertainty, thus reducing the purchase preference for enterprises, and the enterprise’s debt cost will increase. Second, carbon information disclosure reflects the social responsibility of enterprises in carbon-intensive industries and their determination to respond to environmental protection and emission reduction policies. If enterprises in carbon-intensive industries do not actively respond to the national carbon reduction policy and disclose low-quality carbon information to the outside world, investors will think that enterprises have performed their social responsibility poorly, thus losing investment trust in them and restricting investment in enterprises. Further, the financing difficulty of enterprises increases, the cost of debt increases, the working capital is limited and the operating efficiency will also be adversely affected. Therefore, this article proposes Hypothesis 3:

H3: 

Carbon disclosure has a mediating effect on the relationship between operational efficiency and debt cost in carbon-intensive industries.

In addition, other studies have found that lenders mitigate the impact of greenhouse gas emissions of companies in carbon-intensive industries on their future cash flows, but also require companies with higher carbon emission intensity to pay significantly higher costs to finance their operations through debt [58]. Therefore, based on the above analysis, this article proposes Hypothesis 4:

H4: 

Carbon disclosure has no mediating effect on the relationship between operational efficiency and debt cost in carbon-intensive industries.

According to the signal transmission theory, information asymmetry in the capital markets will aggravate the degree of external financing constraints of enterprises. When investors cannot well receive the information disclosed by enterprises, they will be stricter on the operating efficiency and financing behavior of enterprises to reduce investment losses [59]. The study found that low-carbon industry enterprises are more inclined to provide high-quality carbon information disclosure levels. The main reasons are as follows: First, voluntary carbon information disclosure by low-carbon industry enterprises reduces the adverse impact of information asymmetry between investors and enterprises, and then reduces regulatory costs. Second, the disclosure of carbon information of enterprises in low-carbon industries also shows good operating efficiency and operating conditions to the public, improves investors’ purchase preference for enterprise products, reduces the difficulty of enterprise financing, is conducive to the reduction of enterprise debt costs, and further improves the operating efficiency of enterprise funds. Third, carbon information disclosure reflects the determination of enterprises in low-carbon industries to fulfill their social responsibility and respond to environmental protection and emission reduction policies, while avoiding negative market reactions caused by concealing carbon information [47]. Therefore, this article proposes Hypothesis 5:

H5: 

Carbon information disclosure has a mediating effect on the relationship between operational efficiency and debt cost in low-carbon industries.

However, some studies believe that low-carbon industry enterprises face fewer legitimacy problems, and investors are not sensitive to the degree of carbon information disclosure of such enterprises, so the relationship between operational efficiency and debt cost is weak for low-carbon industry enterprises [60]. Therefore, based on the above analysis, this article proposes Hypothesis 6:

H6: 

Carbon information disclosure in low-carbon industries has no mediating effect on the relationship between operational efficiency and debt cost in low-carbon industries.

4. Research Methodology

4.1. Research Sample

This study selects listed companies that have disclosed key data of carbon performance from 2010 to 2021, and the data are from RESSET database and annual reports including social responsibility reports, sustainable development reports or ESG reports. The number of observations was 475, of which 203 were in high-carbon industries and 272 were in low-carbon industries. Liu’s method is adopted to distinguish the high-carbon industry from the low-carbon industry [61]. High-carbon industries mainly include electricity and heat production and supply industry, smelting and pressing of ferrous metals, non-metallic mineral products industry, coal mining and washing industry, etc. Low-carbon industries mainly include oil and gas mining, black metal mining, computer, communication and other electronic equipment manufacturing, water production and supply industry, special equipment manufacturing, textile and garment industry, etc.

4.2. Variables

4.2.1. Explained Variable

The financing cost (explained variable) reflects the interest cost of the enterprise obtaining funds from the bank or other institutions and is measured by the cost of debt financing with the ratio of interest expenses and interest-bearing debt since the cost of equity financing is more complex and may cause more factors involved.

4.2.2. Explanatory Variable

Operational efficiency is an important indicator to measure the financial efficiency and corporate performance of an enterprise and represents the relationship between liquidity and profitability [23,24,25]. In this study, total asset turnover is adopted as a measurement of operational efficiency.

Carbon information disclosure refers to public disclosure of monetary, quantitative or qualitative description related to carbon emission, and its proxy indicator is number of carbon disclosure items including carbon emissions, carbon emission reductions, identification of climate-related risks, emission reduction targets and so on.

With limited capital liquidity and high risk of capital chain breakage, small and medium-sized enterprises are extremely sensitive to the demand for capital [62]. In the case of poor profitability, the link between credit spreads and asset liquidity is more obvious [63]. Therefore, this paper considers factors that have an impact on corporate financing costs as control variables such as company size, assets, liability as well as sales, ROA and fixed asset ratio. In terms of ownership and stocks, equity concentration and stock liquidity are included. Variables are seen in

Table 1. Description of Variables.

Variables	Name	Proxy Variables	Code
Explained Variable	Debt financing cost	Ratio of interest expenses and interest-bearing debt	DeCost
Explanatory Variables	Capital utilization	Total asset turnover ratio	TuRati
	Carbon information disclosure	Number of carbon disclosure items	CaDisc
	Size	Asset	Asset
		Liabilities	Liabi
	Profitability	Operating income	Sale
		Return on total assets	ROA
	Capital utilization	Fixed asset ratio	FiRati
	Equity concentration (EqCon)	Shareholding ratio of the first largest shareholder	EqCon1
		Shareholding ratio of top 5 shareholders	EqCon5
	State of Enterprises/Stock liquidity (ShPerc)	State Shares Percentage	StSha
		Corporate shares	LpSha
		Proportion of tradable A-shares	StLiqu

.

4.3. Empirical Models

Model 1:

$$DeCos{t}_{i,t}={\propto }_{1}TuRat{i}_{i,t}+{\propto }_{2}Sal{e}_{i,t}+{\propto }_{3}RO{A}_{i,t}+{\propto }_{4}FiRat{i}_{i,t}+{\propto }_{5}Asse{t}_{i,t}+{\propto }_{6}Liab{i}_{i,t}+{\propto }_{7}EqCo{n}_{i,t}+{\propto }_{8}ShPer{c}_{i,t}+\epsilon$$

Model 2:

$$CaDis{c}_{i,t}={\beta }_{1}TuRat{i}_{i,t}+{\beta }_{2}Sal{e}_{i,t}+{\beta }_{3}RO{A}_{i,t}+{\beta }_{4}FiRat{i}_{i,t}+{\beta }_{5}Asse{t}_{i,t}+{\beta }_{6}Liab{i}_{i,t}+{\beta }_{7}EqCo{n}_{i,t}+{\beta }_{8}ShPer{c}_{i,t}+\epsilon$$

Model 3:

$$DeCos{t}_{i,t}={\gamma }_{1}CaDis{c}_{i,t}+{\gamma }_{2}TuRat{i}_{i,t}+{\gamma }_{3}Sal{e}_{i,t}+{\gamma }_{4}RO{A}_{i,t}+{\gamma }_{5}FiRat{i}_{i,t}+{\gamma }_{6}Asse{t}_{i,t}+{\gamma }_{7}Liab{i}_{i,t}+{\gamma }_{8}EqCo{n}_{i,t}+{\gamma }_{9}ShPer{c}_{i,t}+{\epsilon }_{i,t}$$

where $DebtCos{t}_{i,t}$ is debt financing cost, $DebtCos{t}_{i,t-1}$ is the first lag term of Debt financing cost, $DebtCos{t}_{i,t-2}$ is lag term of debt financing cost in the second phase and $DebtCos{t}_{i,t-3}$ is the lag term of debt financing cost in the third phase. ε is the disturbance term.

Considering that there may be some continuity between indicators and their previous value, the study set regression models with lag period of explained variables of 3. In order to solve the endogeneity problem, Arellano-Bover/Blundell-Bond (A-B/B-B) estimation of dynamic panel data were used for statistics, and lags of endogenous variables and the predetermined variables were used as instrumental variables. In order to pass the autocorrelation test and overidentifying restrictions test, this paper identifies the share ratio of the largest shareholder, the share ratio of the top five shareholders, the share ratio of state-owned shares, the share ratio of legal shares and the share ratio of A shares in circulation as the former fixed variables, and the operating efficiency and operating revenue and so on as the endogenous variables. The first-stage lag term of the predetermined variable, the first-stage lag term and the second-stage lag term of the endogenous variable are put into the model. The instrumental variables are at most two lag terms of the predetermined variable and the endogenous variable.

Model 4:

$$\begin{array}{ll}DeCos{t}_{i,t}& =DeCos{t}_{i,t-1}+DeCos{t}_{i,t-2}+DeCos{t}_{i,t-3}+TuRat{i}_{i,t}+TuRat{i}_{i,t-1}\\ & +TuRat{i}_{i,t-2}+Sal{e}_{i,t}+Sal{e}_{i,t-1}+Sal{e}_{i,t-2}+RO{A}_{i,t}\\ & +RO{A}_{i,t-1}+RO{A}_{i,t-2}+FiRat{i}_{i,t}+FiRat{i}_{i,t-1}+FiRat{i}_{i,t-2}\\ & +Asse{t}_{i,t}+Asse{t}_{i,t-1}+Asse{t}_{i,t-2}+Liab{i}_{i,t}+Liab{i}_{i,t-1}\\ & +Liab{i}_{i,t-2}+EqCo{n}_{i,t}+EqCo{n}_{i,t-1}+ShPer{c}_{i,t}\\ & +ShPer{c}_{i,t-1}+\epsilon \end{array}$$

Model 5:

$$\begin{array}{ll}CaDis{c}_{i,t}& =CaDis{c}_{i,t-1}+CaDis{c}_{i,t-2}+CaDis{c}_{i,t-3}+TuRat{i}_{i,t}+TuRat{i}_{i,t-1}\\ & +TuRat{i}_{i,t-2}+Sal{e}_{i,t}+Sal{e}_{i,t}+Sal{e}_{i,t-1}+Sal{e}_{i,t-2}\\ & +RO{A}_{i,t}+RO{A}_{i,t-1}+RO{A}_{i,t-2}+FiRat{i}_{i,t}+FiRat{i}_{i,t-1}\\ & +FiRat{i}_{i,t-2}+Asse{t}_{i,t}+Asse{t}_{i,t-1}+Asse{t}_{i,t-2}+Liab{i}_{i,t}\\ & +Liab{i}_{i,t-1}+Liab{i}_{i,t-2}+EqCo{n}_{i,t}+EqCo{n}_{i,t-1}+ShPer{c}_{i,t}\\ & +ShPer{c}_{i,t-1}+\epsilon \end{array}$$

Model 6:

$$\begin{array}{ll}DeCos{t}_{i,t}& =DeCos{t}_{i,t-1}+DeCos{t}_{i,t-2}++DeCos{t}_{i,t-3}\\ & +CaDis{c}_{i,t}+CaDis{c}_{i,t-1}+CaDis{c}_{i,t-2}+TuRat{i}_{i,t}\\ & +TuRat{i}_{i,t-1}+TuRat{i}_{i,t-2}+Sal{e}_{i,t}+Sal{e}_{i,t-1}+Sal{e}_{i,t-2}\\ & +RO{A}_{i,t}+RO{A}_{i,t-1}+RO{A}_{i,t-2}+FiRat{i}_{i,t}+FiRat{i}_{i,t-1}\\ & +FiRat{i}_{i,t-2}+Asse{t}_{i,t}+Asse{t}_{i,t-1}+Asse{t}_{i,t-2}+Liab{i}_{i,t}\\ & +Liab{i}_{i,t-1}+Liab{i}_{i,t-2}+EqCo{n}_{i,t}+EqCo{n}_{i,t-1}+ShPer{c}_{i,t}\\ & +ShPer{c}_{i,t-1}+\epsilon \end{array}$$

When it comes to collinear problems, statistical software will automatically omit collinear variables when analyzing them.

5. Empirical Results

5.1. Descriptive Analysis

The average cost of debt of enterprises in the high-carbon industry is 0.047, and the operating efficiency is 0.502, as shown in the

Table 2. Descriptive estimations in high-carbon industry.

Variable	Obs	Mean	Std. Dev.	Min	Max
DeCost	203	0.047	0.015	0.01	0.087
CaDisc	203	6.32	4.947	0	30
TuRati	203	0.502	0.3	0.09	1.739
Sale	203	470.96	703.631	3.67	3352.16
Asset	203	1111.016	1499.288	18.187	6070.52
Liabi	203	598.685	797.926	8.005	3016.552
EqCon1	203	0.473	0.156	0.152	0.97
EqCon5	203	0.698	0.143	0.356	1
ROA	203	0.047	0.05	−0.102	0.281
FiRati	203	0.477	0.16	0.027	0.876
StSha	203	0.083	0.179	0	0.771
LpSha	203	0.028	0.087	0	0.607
StLiqu	203	0.92	0.146	0	1

. The number of carbon information disclosure items was 6.32. The average operating income was 470.96 million RMB. The share ratio of the largest shareholder was 0.473, while that of the top five shareholders was 0.698. The proportion of A-shares in circulation reached 92%, and percentage for state-owned shares is 8.3%.

Average cost of debt of enterprises in low-carbon industry is 0.063, and the operational efficiency (total asset turnover) is 0.646, as shown in

Table 3. Descriptive estimations in low-carbon industry.

Variable	Obs	Mean	Std. Dev.	Min	Max
DeCost	272	0.063	0.078	0.007	0.74
CaDisc	272	9.853	5.51	1	31
TuRati	272	0.646	0.516	0.023	2.561
Sale	272	1618.212	4288.257	12.035	29,661.93
Asset	272	11,273.443	36,301.971	19.251	302,539.81
Liabi	272	9897.212	33,251.017	5.013	276,398.59
EqCon1	272	0.388	0.181	0.078	0.99
EqCon5	272	0.63	0.189	0.232	1.005
ROA	272	0.039	0.047	−0.2	0.216
FiRati	272	0.161	0.142	0.001	0.837
StSha	272	0.048	0.111	0	0.562
LpSha	272	0.026	0.108	0	0.902
StLiqu	272	0.818	0.194	0	1

. The average operating revenue was RMB 16.18212 million. The largest shareholder has a share ratio of 0.388, while the top five shareholders have a share ratio of 0.63. The proportion of A-shares in circulation reached 81.8 per cent, compared with the percentage for state-owned shares, which is 4.8%.

To sum up, the average debt cost and operational efficiency of enterprises in high-carbon industries are lower than those in low-carbon industries, and more carbon information is disclosed. However, the proportion of A-shares in circulation, state-owned shares, the proportion of the largest shareholder and the proportion of the top five shareholders in the low-carbon industry are all lower than that of the high-carbon industry.

5.2. Correlation Analysis

The debt cost of enterprises in high-carbon industries is not correlated with operational efficiency, but significantly correlated with carbon information disclosure, ratio of fixed assets, ratio of legal shares and ratio of A-shares in circulation (see

Table 4. Correlations in high-carbon industries.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)
(1) DeCostt	1.000
(2) TuRati	0.046	1.000
	(0.518)
(3) CaDiscs	0.139 **	0.165 **	1.000
	(0.048)	(0.019)
(4) Sale	0.087	0.318 ***	0.623 ***	1.000
	(0.217)	(0.000)	(0.000)
(5) Asset	0.104	−0.031	0.592 ***	0.929 ***	1.000
	(0.138)	(0.662)	(0.000)	(0.000)
(6) Liabi	0.108	−0.124 *	0.561 ***	0.873 ***	0.976 ***	1.000
	(0.125)	(0.079)	(0.000)	(0.000)	(0.000)
(7) ROA	−0.038	0.333 ***	0.094	0.258 ***	0.155 **	0.034	1.000
	(0.594)	(0.000)	(0.181)	(0.000)	(0.027)	(0.627)
(8) FiRati	−0.173 **	−0.406 ***	−0.035	0.088	0.255 ***	0.280 ***	−0.078	1.000
	(0.013)	(0.000)	(0.622)	(0.211)	(0.000)	(0.000)	(0.266)
(9) EqCon1	−0.009	−0.191 ***	0.090	0.265 ***	0.375 ***	0.348 ***	0.026	0.149 **	1.000
	(0.897)	(0.006)	(0.203)	(0.000)	(0.000)	(0.000)	(0.713)	(0.034)
(10) EqCon5	0.094	0.027	0.329 ***	0.510 ***	0.513 ***	0.500 ***	0.064	0.083	0.599 ***	1.000
	(0.182)	(0.704)	(0.000)	(0.000)	(0.000)	(0.000)	(0.365)	(0.238)	(0.000)
(11) StSha	−0.054	−0.250 ***	−0.146 **	−0.035	0.103	0.139 **	−0.063	0.214 ***	0.444 ***	0.225 ***	1.000
	(0.445)	(0.000)	(0.038)	(0.617)	(0.144)	(0.048)	(0.373)	(0.002)	(0.000)	(0.001)
(12) LpSha	−0.173 **	−0.040	−0.162 **	−0.132 *	−0.134 *	−0.109	−0.053	0.059	−0.013	0.091	0.184 ***	1.000
	(0.013)	(0.573)	(0.021)	(0.061)	(0.057)	(0.122)	(0.453)	(0.402)	(0.856)	(0.199)	(0.009)
(13) StLiqu	−0.117 *	−0.070	−0.301 ***	−0.409 ***	−0.396 ***	−0.398 ***	−0.029	0.111	0.070	−0.343 ***	0.072	0.130 *	1.000
	(0.096)	(0.319)	(0.000)	(0.000)	(0.000)	(0.000)	(0.676)	(0.115)	(0.322)	(0.000)	(0.306)	(0.064)

*** p < 0.01, ** p < 0.05, * p < 0.1.

). The debt cost is positively correlated with carbon information disclosure, with a coefficient of 0.139. There was a significant positive correlation between operating efficiency and operating revenue, ROA and carbon information disclosure, and the coefficients were 0.318, 0.333 and 0.165, respectively. Carbon information disclosure is positively correlated with operating income, assets and liabilities, but negatively correlated with the proportion of state-owned shares and the proportion of legal shares.

The debt cost of enterprises in low-carbon industry is significantly positively correlated with operating efficiency and ROA, and negatively correlated with the ratio of state-owned shares, with a coefficient of −0.132, as shown in

Table 5. Correlations in low-carbon industries.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)
(1) DeCostt	1.000
(2) TuRati	0.125 **	1.000
	(0.039)
(3) CaDiscs	0.048	0.165 ***	1.000
	(0.426)	(0.007)
(4) Sale	0.002	0.178 ***	0.578 ***	1.000
	(0.968)	(0.003)	(0.000)
(5) Asset	−0.007	−0.370 ***	0.472 ***	0.799 ***	1.000
	(0.903)	(0.000)	(0.000)	(0.000)
(6) Liabi	0.002	−0.395 ***	0.446 ***	0.777 ***	0.993 ***	1.000
	(0.969)	(0.000)	(0.000)	(0.000)	(0.000)
(7) ROA	0.128 **	0.428 ***	−0.016	0.011	−0.201 ***	−0.251 ***	1.000
	(0.034)	(0.000)	(0.798)	(0.861)	(0.001)	(0.000)
(8) FiRati	0.024	0.341 ***	−0.177 ***	−0.273 ***	−0.505 ***	−0.530 ***	0.177 ***	1.000
	(0.697)	(0.000)	(0.003)	(0.000)	(0.000)	(0.000)	(0.003)
(9) EqCon1	0.105 *	0.226 ***	0.312 ***	0.200 ***	0.100	0.090	−0.033	−0.040	1.000
	(0.084)	(0.000)	(0.000)	(0.001)	(0.100)	(0.139)	(0.593)	(0.512)
(10) EqCon5	0.076	0.096	0.364 ***	0.259 ***	0.261 ***	0.246 ***	−0.087	−0.182 ***	0.786 ***	1.000
	(0.214)	(0.116)	(0.000)	(0.000)	(0.000)	(0.000)	(0.152)	(0.003)	(0.000)
(11) StSha	−0.132 **	−0.167 ***	−0.166 ***	−0.151 **	−0.071	−0.058	−0.169 ***	0.073	0.009	0.058	1.000
	(0.029)	(0.006)	(0.006)	(0.013)	(0.246)	(0.339)	(0.005)	(0.232)	(0.879)	(0.344)
(12) LpSha	0.071	0.241 ***	0.022	−0.131 **	−0.201 ***	−0.187 ***	0.074	0.031	0.314 ***	0.151 **	−0.053	1.000
	(0.240)	(0.000)	(0.716)	(0.031)	(0.001)	(0.002)	(0.224)	(0.614)	(0.000)	(0.013)	(0.386)
(13) StLiqu	−0.008	0.095	−0.408 ***	−0.314 ***	−0.393 ***	−0.398 ***	0.124 **	0.293 ***	−0.315 ***	−0.521 ***	0.075	−0.014	1.000
	(0.894)	(0.118)	(0.000)	(0.000)	(0.000)	(0.000)	(0.042)	(0.000)	(0.000)	(0.000)	(0.216)	(0.825)

*** p < 0.01, ** p < 0.05, * p < 0.1.

. Operating efficiency is significantly positively correlated with operating revenue, ROA, fixed asset ratio, carbon information disclosure, the share ratio of the largest shareholder and the share ratio of legal equity. There is a significant negative correlation between operating efficiency and the ratio of assets, liabilities and state-owned shares. Carbon information disclosure is significantly positively correlated with operating income, assets and liabilities, the share ratio of the largest shareholder and the share ratio of the top five shareholders, but significantly negatively correlated with the share ratio of state-owned shares and the share ratio of A-shares in circulation.

5.3. Regression Results

The more efficient enterprises in high-carbon industries are, the lower the cost of debt, as shown in the

Table 6. Estimates in regressions.

	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
	high−carbon industries			low−carbon industries
	DeCost	CaDisc	DeCost	DeCost	CaDisc	DeCost
TuRati	0.053 *** (0.014)	0.079 (0.089)	0.053 *** (0.014)	0.070 ** (0.028)	0.078 ** (0.036)	0.068 ** (0.028)
CaDiscs	−	−	0.001 (0.011)	−	−	0.036 (0.047)
Sale	−0.030 *** (0.007)	−0.024 (0.043)	−0.030 *** (0.007)	−0.033 *** (0.013)	0.020 (0.017)	−0.034 *** (0.013)
Asset	0.033 *** (0.009)	0.113 ** (0.056)	0.033 *** (0.009)	−0.073 ** (0.034)	0.171 *** (0.045)	−0.079 ** (0.035)
Liabi	−0.002 (0.005)	−0.028 (0.030)	−0.002 (0.005)	0.095 *** (0.030)	−0.134 *** (0.039)	0.100 *** (0.030)
ROA	−0.038 (0.033)	−0.291 (0.213)	−0.038 (0.034)	0.403 ** (0.167)	−0.486 ** (0.219)	0.421 ** (0.169)
FiRati	−0.032 *** (0.010)	−0.140 ** (0.064)	−0.032 *** (0.010)	0.075 (0.055)	−0.021 (0.072)	0.076 (0.055)
EqCon1	−0.033 ** (0.014)	−0.196 ** (0.091)	−0.032 ** (0.014)	0.051 (0.062)	0.106 (0.081)	0.047 (0.062)
EqCon5	0.045 *** (0.015)	0.209 ** (0.098)	0.045 *** (0.016)	0.034 (0.065)	−0.022 (0.085)	0.035 (0.065)
StSha	−0.000 (0.000)	−0.001 * (0.001)	−0.000 (0.000)	−0.001 ** (0.001)	−0.001 (0.001)	−0.001 * (0.001)
LpSha	−0.000 ** (0.000)	−0.001 (0.001)	−0.000 ** (0.000)	−0.000 (0.001)	0.001 (0.001)	−0.000 (0.001)
StLiqu	0.000 (0.000)	0.001 (0.001)	0.000 (0.000)	0.000 (0.000)	−0.002 *** (0.001)	0.000 (0.000)
_cons	0.877 *** (0.021)	−0.258 * (0.136)	0.877 *** (0.022)	0.880 *** (0.064)	−0.013 (0.083)	0.880 *** (0.064)
N	203.000	203.000	203.000	272.000	272.000	272.000
r2	0.178	0.447	0.178	0.108	0.454	0.110

Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.

. However, the effect of operational efficiency on carbon information disclosure is not significant. That is, it fails to verify the mediating effect of carbon information disclosure on the relationship between operational efficiency and debt cost in high-carbon industries. The higher the operating efficiency of enterprises in low-carbon industries, the lower the debt cost; Operational efficiency plays a positive role in carbon information disclosure. Carbon information disclosure has an intermediary effect on the relationship between operational efficiency and debt cost.

Regression residual is shown in Figure 1. The odds ratio of statistical power was 2.0561, and the estimated power of the Cochran–Mantel–Haenszel test was 0.9535 when the sample was increased to 800. The tests based on the bootstrap method show that the bias-corrected confidence interval of the direct effect is [0.0115317, 0.1548821], excluding 0. That is, the regression coefficient was statistically significant. There is a partial mediation effect.

5.4. Robustness Test

Using the substitution variable method, carbon information disclosure is changed into carbon information disclosure scores (CaDisc1), and the conclusion is consistent with the previous conclusion, as shown in

Table 7. Results of robustness test.

	Model 1	Model 2	Model 3
	DeCost	CaDisc1	DeCost
TuRati	0.070 ** (0.028)	−0.103 *** (0.035)	0.066 ** (0.028)
CaDisc1	−	−	−0.038(0.048)
Sale	−0.033 *** (0.013)	−0.006 (0.016)	−0.034 *** (0.013)
Asset	−0.073 ** (0.034)	−0.181 *** (0.044)	−0.080 ** (0.035)
Liabi	0.095 *** (0.030)	0.131 *** (0.038)	0.100 *** (0.030)
ROA	0.403 ** (0.167)	0.624 *** (0.215)	0.427 ** (0.170)
FiRati	0.075 (0.055)	0.046 (0.071)	0.077 (0.055)
EqCon1	0.051 (0.062)	−0.043 (0.079)	0.049 (0.062)
EqCon5	0.034 (0.065)	−0.039 (0.083)	0.033 (0.065)
StSha	−0.001 ** (0.001)	0.001 (0.001)	−0.001 * (0.001)
LpSha	−0.000 (0.001)	−0.001 (0.001)	−0.000 (0.001)
StLiqu	0.000 (0.000)	0.002 *** (0.001)	0.000 (0.000)
_cons	0.880 *** (0.064)	1.054 *** (0.082)	0.920 *** (0.082)
N	272.000	272.000	272.000
r2	0.108	0.478	0.110

Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.

. Carbon information disclosure scores refer to weighted carbon information disclosure and higher scores were assigned to the carbon information disclosure items that contain monetary or quantitative information. More specifically, 3 points, 2 points and 1 point were for each monetary, quantitative and qualitative carbon information disclosed, respectively. Carbon information disclosure in low-carbon industries has a significant mediating effect on the relationship between operational efficiency and debt cost. The direct effect coefficient of bootstrap was 0.066, the bias-corrected confidence interval was [0.0030378, 0.1485055] and the regression coefficient was statistically significant.

5.5. Endogeneity

For the endogeneity test of low-carbon samples, the GMM method was applied, and the lag terms of some variables were taken as endogenous variables for statistical analysis. In Model 4–6, the first-stage lag term, second-stage lag term and third-stage lag term of the dependent variable (cost of debt) are put into the model. In order to pass the autocorrelation test and overidentifying restrictions test, this paper identifies the share ratio of the largest shareholder, the share ratio of the top five shareholders, the share ratio of state-owned shares, the share ratio of legal shares and the share ratio of A shares in circulation as the predetermined variables, and the operational efficiency and operating revenue and other variables are set as the endogenous variables. The first-stage lag term of the predetermined variable, the first-stage lag term of the endogenous variable and the second-stage lag term are put into the model. The instrumental variables are at most two lag terms of the predetermined variable and the endogenous variable. The results show that the higher the operating efficiency is, the lower the debt cost is, and the coefficient is 0.128, as shown in

Table 8. Results of endogeneity.

	Model 4	Model 5	Model 6
	Coef.
L	0.152 (1.43)	0.130 (0.69)	0.067 (0.56)
L2	0.412 *** (2.63)	0.325 ** (2.57)	0.399 ** (2.51)
L3	0.223 *** (2.62)	0.337 *** (3.17)	0.278 ***(3.31)
TuRati	0.128 ** (2.55)	0.002 (0.02)	0.109 ** (2.01)
TuRati L1	0.070 (1.26)	−0.251 * (−1.74)	0.053 (0.91)
TuRati L2	−0.256 *** (−3.42)	0.343 *** (3.18)	−0.208 *** (−2.92)
CaDisc	-	-	0.084 (1.14)
CaDisc L1			−0.114 *** (−3.53)
CaDisc L2			−0.084 (−1.43)
Sale	−0.078 * (−1.90)	−0.066 (−0.90)	−0.055 (−1.28)
Sale L1	0.014 (0.36)	0.026 (0.36)	0.022 (0.53)
Sale L2	0.090 ** (2.43)	0.022 (0.31)	0.061 * (1.74)
ROA	−0.204 (−1.22)	0.397 (1.23)	−0.216 (−1.19)
ROA L1	−0.332 (−1.57)	0.366 * (1.77)	−0.356 (−1.50)
ROA L2	0.356 * (1.80)	−0.508 ** (−2.40)	0.267 (1.38)
Asset	0.465 *** (4.63)	0.123 (0.60)	0.413 *** (4.11)
Asset L1	−0.183 (−1.34)	0.16 (0.71)	−0.173 (−1.32)
Asset L2	−0.078 (−0.61)	−0.03 (−0.14)	−0.011 (−0.09)
Liabi	−0.122 * (−1.67)	−0.159 (−0.93)	−0.078 (−1.03)
Liabi L1	−0.070 (−0.79)	−0.02 (−0.12)	−0.075 (−0.85)
Liabi L2	−0.014 (−0.16)	−0.016 (−0.13)	−0.071 (−0.87)
FiRati	0.062 (0.62)	−0.203 (−1.22)	0.079 (0.73)
FiRati L1	−0.571 *** (−4.39)	−0.311 * (−1.72)	−0.557 *** ( −4.33)
FiRati L2	0.443 *** (3.27)	0.431 * (1.83)	0.375 *** (3.12)
EqCon1	0.083 (0.52)	0.214 (0.54)	0.127 (0.96)
EqCon1 L1	−0.046 (−0.32)	−0.348 (−0.83)	−0.022 (−0.18)
EqCon5	−0.123 (−0.84)	−0.663 (−1.43)	−0.115 (−0.84)
EqCon5 L1	0.069 (0.49)	0.640 (1.40)	0.020 (0.14)
StSha	−0.001 (−0.73)	0.003 ** (2.05)	−0.001 (−1.41)
StSha L1	−0.001 (−1.42)	0 (−0.24)	−0.001 * (−1.80)
LpSha	0 (−1.17)	0 (0.65)	0 (−1.30)
LpSha L1	0.001 *** (3.70)	0.001 *** (2.93)	0 *** (2.77)
StLiqu	0.006 (1.02)	−0.033 *** (−2.77)	0.005 (0.9)
StLiqu L1	−0.006 (−1.03)	0.031 *** (2.58)	−0.005 (−0.91)
Constant	0.039 (0.27)	−0.122 (−0.57)	0.041 (0.31)
Mean dependent var	0.928	0.379	0.928
SD dependent var	0.112	0.199	0.112
Number of obs	84	84	84
Chi−square	23,103,375.463	2,070,644.289	74,477,998,632.943
Number of instruments	104	103	106
Autocorrelation (Order 2)	0.3346	0.4951	0.644
overidentifying restrictions	0.9991	0.9998	0.9999

* p < 0.1, ** p < 0.05, *** p < 0.01.

. The lag term of operational efficiency plays a significant role in carbon information disclosure. In Model 6, the addition of carbon information disclosure leads to an insignificant coefficient of operational efficiency. It is verified that carbon information disclosure is an intermediary variable between operational efficiency and debt cost.

6. Discussion and Conclusions

The results show that there is no significant mediating effect of carbon information disclosure on the relationship between operational efficiency and debt cost in high-carbon industries. However, this intermediary effect exists in low-carbon industry, and it is a partial intermediary effect. This may indicate that the carbon disclosure of low-carbon industries is more closely related to the cost of debt. The deep reason may be that compared with the carbon-intensive industry, the carbon information disclosure of enterprises in low-carbon industry can reflect the enthusiasm of enterprises in their own financing or the restrictive channels of enterprise carbon performance evaluation. However, for enterprises in high-carbon industries, their financing is more complicated in terms of enterprise scale, profitability, future development ability and relational capital. This is why carbon information disclosure of enterprises in high-carbon industries has little impact on the relationship between operational efficiency and debt cost.

The operational efficiency (total asset turnover, etc.) of enterprises in different carbon-intensive industries has a similar positive impact on the cost of debt. The positive value of carbon information disclosure of enterprises in low-carbon industries provides favorable evidence for enterprises to make carbon information disclosure. However, promoting the link between carbon information disclosure and debt cost of enterprises in high-carbon industries will help enterprises in the future to create carbon information transparency or voluntary carbon information disclosure. It is of great significance for regional and industrial enterprises to choose the disclosure system of voluntary or mandatory carbon information disclosure.

Given that this study has explored the internal relationship between operational efficiency, carbon information disclosure and debt cost, the ratio of fixed assets is only a control variable, and the alternative variables of operational efficiency are still few. The fixed asset ratio can be an important distinction between high-carbon and low-carbon industries. Considering the complexity of operational efficiency, quick ratio and other operational indicators should be given some attention. With deeper consideration of the insignificant impact of carbon information disclosure on the relationship between operational efficiency and debt cost of enterprises in high-carbon industries, future studies should investigate the level of debt of enterprises in high-carbon industries to make conclusions on relationship between carbon information disclosure and certain debt cost characteristics of enterprises. In addition, future studies need to distinguish between debt cost and equity cost, so as to have a deeper understanding of the influencing mechanisms of operational efficiency, environmental information disclosure and debt cost.