Impact of carbon disclosure on debt financing costs

Abstract

Creditors, such as banks, often use disclosed environmental information to assess a company's environmental risk and ensure the safety of debt funds. Consequently, carbon disclosures have become an important consideration for creditors when making investments. This study explores the relationship between carbon disclosure and debt financing costs using data on listed companies from 2008 to 2019. The results show that carbon disclosure can reduce the debt financing costs of enterprises, and that this influence is more significant for private companies than for state-owned enterprises. Instrumental variables and Propensity Score Matching (PSM) were used to evaluate the robustness of negative relationships. Furthermore, carbon disclosure has a more significant impact on debt costs with less environmental supervision pressure, weak residents' environmental awareness, and weak product market competition. These findings provide guidance for companies ' carbon information disclosure and support the establishment of official carbon disclosure standards.

Headnote

ARTICLE INFO

Keywords:

Carbon disclosure

Debt financing cost

State-owned enterprise

Private enterprise

ABSTRACT

Creditors, such as banks, often use disclosed environmental information to assess a company's environmental risk and ensure the safety of debt funds. Consequently, carbon disclosures have become an important consideration for creditors when making investments. This study explores the relationship between carbon disclosure and debt financing costs using data on listed companies from 2008 to 2019. The results show that carbon disclosure can reduce the debt financing costs of enterprises, and that this influence is more significant for private companies than for state-owned enterprises. Instrumental variables and Propensity Score Matching (PSM) were used to evaluate the robustness of negative relationships. Furthermore, carbon disclosure has a more significant impact on debt costs with less environmental supervision pressure, weak residents' environmental awareness, and weak product market competition. These findings provide guidance for companies ' carbon information disclosure and support the establishment of official carbon disclosure standards.

(ProQuest: ... denotes formulae omitted.)

1. Introduction

The concept of "energy saving and emission reduction" was first outlined in China's "Eleventh Five-Year" plan, which targeted the reduction of energy consumption per unit of GDP, and total discharge of major pollutants. At the 15th United Nations General Assembly, Chinese government proposed that China should strive to reach its peak carbon dioxide emissions by 2030 and achieve carbon neutrality by 2060. Reducing carbon emissions and achieving a low-carbon transition are clear requirements for national policies, indicating that enterprises must follow these trends in the future. Therefore, the disclosure of carbon information by companies plays an important role in demonstrating their intention to actively assume social responsibility. Currently, the carbon information disclosure systems of listed companies in China are underdeveloped. In general, carbon information disclosure is characterized by an unorganized structure, large industry differences, and asymmetric disclosure quantity and quality (Chen et al., 2013), owing to the lack of a clear and unified evaluation system and basis for disclosure by companies (Li et al., 2015).

Environmental information disclosure has also become a key element affecting a company's debt financing, because it can reduce information asymmetry and form a signal transmission mechanism. As lenders of most corporate debt financing, commercial banks hope to reduce the risk of credit mismatches through adequate information disclosure (Li and Wang, 2011) to obtain benefits while ensuring the safety of investment funds. With the announcement of more policies supporting green credit, company performance in terms of environmental responsibility becomes an important reference for creditors in judging the company's environmental uncertainty, predicting its future performance, and evaluating credit risk (Clarkson et al., 2008). Companies can disclose environmental information through corporate social responsibility (CSR) or corporate sustainability reports to reduce credit risk and corporate debt financing costs (Ni and Kong, 2016).

At present, carbon information disclosure by Chinese listed companies is voluntary. In the absence of unified standards for carbon information disclosure, improving its quality requires additional cost input. However, many companies are still willing to actively disclose carbon information to the public and are committed to improving the quality of their information. There must be benefits that outweigh the costs that motivate them to disclose carbon information.

Previous research has focused on whether companies disclose relevant information rather than how well they disclose it. Even when researchers evaluate the quality of carbon disclosure, the criteria they use are not sufficiently authoritative. To address this shortcoming, this study aims to establish a carbon information disclosure evaluation standard that evaluates the quality of carbon disclosure based on the Carbon Disclosure Project (CDP) and the Global Reporting Initiative (GRI). Following the theory of information asymmetry and signal transmission, this study verifies the conclusions of previous studies on environmental disclosure and debt costs. Moreover, this study considers heterogeneity from three aspects-government supervision, social awareness and market competition-thereby expanding the conclusions on the economic consequences of carbon information disclosure. In practice, this research can serve as a reference for creditors to better evaluate the credit and environmental risks of enterprises and provide guidance for companies to reduce their debt costs. In the context of low-carbon transactions and green development, the results can provide a basis for green policymaking and help governments and markets allocate resources more efficiently.

The remainder of this paper is organized as follows. Section 2 summarizes the related studies on carbon disclosure and debt costs, explains asymmetric information problems and the theory of information transition, and presents the hypotheses. Section 3 provides the data used in the study. Section 4 outlines the research method, introduces the regression model, and explains the study variables. Sections 5 and 6 present the empirical results and further analysis, respectively. Finally, Section 7 concludes the paper and discusses potential future research.

2. Literature review and hypothesis

2.1. Theoretical analysis and literature review

2.1.1. Theory of information asymmetry

Jaffee and Russell (1984) believed that the main reason for credit mismatches is information asymmetry between corporate managers and creditors. As external interest-related parties, creditors have little knowledge of the internal information of companies. Generally, companies can be analyzed using information required by law, published in the market, and disclosed independently. To maintain their corporate image and ensure their credit level and investor confidence, companies selectively disclose positive information to avoid the outflow of negative information. This leads to an information asymmetry between creditors and internal managers.

In the context of the increasing intensity of carbon emission regulations, if a company is penalized for greenhouse gas pollution during its operation, its profitability and development capabilities are likely to be affected. Most creditors are rational investors, and their investment goals are to maximize returns and minimize risks; hence, they raise doubts about the safety of borrowed funds and compensate for risk through higher interest charges. In the case of asymmetric environmental information, creditors may find it difficult to assess environmental risk levels. If a company's environmental performance information is unavailable, prudent investors will increase borrowing costs if the worst situation occurs.

The active disclosure of carbon information by companies can considerably solve the problem of information asymmetry, help creditors better estimate corporate environmental risks, and more accurately assess the safety of funds. Because negative disclosures can affect the cost of debt financing, the absence of information will make creditors raise interest expenses as much as possible based on the conservative principle. However, if the company objectively discloses its own environmental risks, creditors can increase interest expenses appropriately based on the actual situation and reduce credit mismatches.

2.1.2. Theory of signal transmission

Clarkson et al. (2008) argued that companies with good environmental performance can distinguish themselves from other companies by disclosing more environmental information. Information dissemination is extremely rapid because of online media, which supports the application of the signal transmission theory. Companies can be differentiated from companies with poor carbon performance through the active disclosure of carbon information; thus, carbon disclosure is also a means for companies to promote and compete. By transmitting the beneficial signals of stable growth and sustainable development to creditors, companies can reduce credit risk and enhance creditors ' investment confidence. Therefore, rational creditors will adopt lower interest rates for companies with lower risks.

Corporations actively disclose carbon information and relay three types of signals to creditors. Firstly, companies show that their business and production processes are legally compliant by disclosing carbon emission data, management systems, laws, and regulations, thereby avoiding the risk of penalties, fines, or even suspension of operations. Consequently, companies can reduce their borrowing risk and attract stable long-term investments. Secondly, companies disclose their own green product information and attract the public to purchase their products and services by promoting emissions reduction and green production concepts. In the context of strong environmental awareness, the concept of emissions reduction is deeply rooted in the hearts of people, allowing companies to rely on lowcarbon products to realize income growth and increase creditors' investment confidence. Thirdly, by disclosing low-carbon and green development strategies, companies show their attitude of actively responding to national policies and development trends, proving their qualification for low-carbon and environmentally friendly subsidies. Thus, companies can send stable and sustainable development signals to their creditors.

2.1.3. Carbon disclosure

Research topics on carbon disclosure are mainly divided into two categories: factors influencing the carbon disclosure level and its financial consequences. The factors affecting carbon information disclosure focus on the enterprise itself or on external supervision, and enterprise factors can be further classified based on internal governance and financial factors.

In terms of internal governance, empirical evidence has shown that both the proportion of independent directors and activity of the board of directors positively impact the quality of carbon information disclosure (Cui et al., 2016). In addition to operational management, a high degree of equity concentration can positively affect the level of environmental information disclosure (Wang et al., 2013; Yang, 2017). Brammer and Pavelin (2006) found that companies with dispersed ownership are more willing to voluntarily disclose environmental information.

Financial factors can also determine carbon disclosure because only well-developed companies can assume more social responsibility (Preston and O'Bannon, 1997). Companies with greater size, profitability, and refinancing needs receive more social supervision, therefore they are required to disclose more about social responsibility (Cui et al., 2016; Shen, 2007). Li et al. (2019) found that companies with good carbon performance were more willing to disclose carbon information. Furthermore, according to the agency theory (Jensen and Meckling, 1976), the higher the financial leverage, the higher the agency cost of a company; therefore, more information must be disclosed. Clarkson et al. (2008) and Meek et al. (1995) found that financial leverage positively promotes the disclosure of environmental information.

External supervision generally involves supervision by the government and society. Wang (2008) explained the impact of external factors using two theories. In the context of the theory of accountability, enterprises must be responsible to society in the process of creating wealth and benefitting the public. Liu and Yang (2018) found that heavily polluting companies take more responsibility for carbon reduction; thus, their carbon disclosure levels are higher. In relation to the theory of external pressures, companies are forced to disclose environmental information because of pressure from the government and public. Wang (2008) found differences among supervisory systems among different industries influence environmental information disclosure of heavily polluting and non-heavily polluting companies. Li and Liu (2016) believed that corporate carbon information disclo- sure can help a country formulate carbon emission management policies, and that news reports play a supervisory role in companies ' environmental behaviors.

The economic consequences of carbon information disclosure can be analyzed from three perspectives: financial performance, corporate value and financing constraints. Most scholars believed that environmental information disclosure can improve financial performance and corporate value by alleviating information asymmetry, improving corporate reputation, reducing capital costs and improving corporate legitimacy (Wen and Zhou, 2017). Wegener et al. (2013) concluded that carbon information is positively correlated with corporate performance using CDP questionnaire responses as indicators. Zhang (2018) analyzed the positive effects of carbon information disclosure on profitability, solvency, and operational capabilities. Wen and Zhou (2017) found that this positive relationship is more significant among stateowned enterprises (SOE) and non-polluting enterprises, and media has played a significant inverted U-shaped moderating effect, having the dual functions of "information dissemination" and being a "social public tool". Li and Liu (2016) found that corporate value can be enhanced only when disclosure reaches a certain level. However, some scholars have arrived at the opposite conclusion, believing that actively fulfilling social responsibilities will cost companies additional expenditures. Using the residual income valuation model, Zhang et al. (2013) found that in high-emission companies, carbon information disclosure is negatively correlated with corporate value, and Hassel et al. (2005) concluded that environmental information disclosure is negatively related to company value.

Scholars ' analyses of the impact of carbon information disclosure on financing constraints have focused on reducing investment uncertainty by alleviating information asymmetry. Enhancing information transparency can reduce corporate financing constraints (Zhang and Lv, 2007). Li and Wang (2011) found that companies with highquality information disclosure can obtain more loans because they foster a more consistent understanding of the company between investors and corporate executives, thereby reducing financial constraints (Leuz and Verrecchia, 2000). From the perspective of the corporate life cycle, Ma and Gai (2019) found that the disclosure of carbon information can effectively alleviate financing constraints during the growth and maturity stages. External investors pay more attention to the core competitiveness of the company's products or technologies and future development than to the social responsibility of a start-up company. Ni and Kong (2016) found that the quality of environmental information disclosure has a more significant impact on the shortterm bank borrowings than long-term bank borrowings of listed companies in heavily polluting industries. Goss and Roberts (2011) suggested that banks consider GSR the second most important factor in determining loan spreads.

In summary, carbon information disclosure primarily affects enterprises by alleviating information asymmetry. Most studies believe that this effect is positive and beneficial to enterprises, whereas a small number of scholars consider that additional expenditures decrease corporate profits and indicate a negative correlation.

2.1.4. Debt financing costs

The factors influencing the cost of debt are the company itself and the external environment. According to the corporate finance theory, a company's credit rating can affect the cost of debt (Li and Liu, 2009). Stakeholders can calculate relevant indicators based on a company's financial statements and analyze its profitability, liquidity, and solvency. Besides specific financial data that can reflect corporate debt risk, corporate internal controls also impact debt financing costs. High-quality internal controls provide more creditor protection signals, improve creditors' perceptions of the company's risks, and help creditors relax debt contract conditions (Lin et al., 2017). If the internal controls have defects, creditors will be less dependent on the company ' s disclosed information and will charge higher debt costs to compensate for the unreliability of information (Kim et al., 2011). Chen and Zhou (2014) stated that good internal controls can reduce a company's illegal acts, agency conflicts, and internal costs.

In addition to the profitability and management characteristics of the enterprise itself, the external environment, such as government actions and media reports, also affects financing costs. Compared with Western countries, China's SOEs occupy a dominant position in the market economy, and government intervention is common. The government specifically supports the economic development of a certain area and assists local enterprises in obtaining support from financial institutions, which will reduce their business risk and debt financing costs and increase corporate protection (Ni and Xu, 2016). Changes in monetary policy adversely affect enterprises; however, this impact can be mitigated by higher-quality information disclosure (Li and Wang, 2011). Additionally, private companies face credit discrimination from banks. Listed companies have increased debt financing costs after privatization, and privately listed companies bear higher debt financing costs than non-privately listed companies (Xia et al., 2018). In terms of the social environment, Xia et al. (2018) found that the number of media reports negatively correlates with the cost of debt, and that positive reports can reduce the cost of corporate debt.

2.1.5. Carbon disclosure and debt financing costs

Few studies have examined the relationship between carbon information disclosure and debt financing costs, and their conclusions are controversial. Scholars believed that environmental information disclosure is positively related to the cost of debt owing to increased expenses in disclosing information. Currently, carbon information disclosure in China is voluntary and lacks uniform standards, with many companies spending considerable money exploring how to prepare and disclose true and reliable carbon information (Tan, 2017). If the funds invested in environmental protection activities do not increase the value of a company in the short term, they will occupy resources for the company to create more value, decreasing its profitability and creditors' lending willingness (Ni and Kong, 2016). Compared with security markets in developed countries, the quality of accounting information disclosure in China's securities markets is relatively low, and investors' rights and interest protection systems are behind (Wang and Jiang, 2004). Low information quality, such as when companies conceal negative news and only report good information when disclosing environmental information, may also increase the cost of debt as this does not help creditors make well-informed credit decisions (Zhang et al., 2011). Yang et al. (2020) concluded that carbon information disclosure has an inverted "U-shaped" effect on debt financing costs. That is, only when disclosure reaches a certain level does it become useful, otherwise it only becomes an extra cost.

Many scholars concluded that environmental information negatively correlates with debt financing costs. Enhancing information transparency can effectively alleviate information asymmetry. Botosán (1997) and Sengupta (1998) found a negative correlation between information transparency and corporate debt financing costs. Some scholars believe that creditors require a higher rate of return from companies with insufficient environmental information disclosure to compensate for environmental risks (Goss and Roberts, 2011). Ni and Kong (2016) found that companies that actively disclosed environmental information had lower debt financing costs. Furthermore, He et al. (2014) and Lemma et al. (2019) concluded that voluntary carbon information disclosure is negatively correlated to the cost of capital.

2.2. Research hypotheses

Investors' opinions of a particular company are crucial for their financing activities. To ensure fund safety, investors must reduce their investment uncertainty. When credit resources are in shortage, investors are more likely to invest in low-uncertainty companies or charge higher interest rates to effectively allocate their credit asset portfolios (Li and Wang, 2011). Information asymmetry is an essential reason for this uncertainty. If the degree of information asymmetry between banks and enterprises is high, banks will find it difficult to understand the credibility of corporate credit fund repayments. Consistent information helps banks judge enterprises' credit motives and form accurate repayment expectations. Therefore, information asymmetry is a fundamental element affecting banks ' credit decisions (Li and Wang, 2011). Furthermore, positive signals can strengthen investor awareness of a company's brand and sustainable development capabilities, thereby reducing capital costs and increasing profitability (Wen and Zhou, 2017).

With China's targets to reach its carbon peak by 2030 and achieve carbon neutrality by 2060, the emphasis on carbon emissions has made carbon action an important factor in investment decisions. Factors such as whether the entity has a problem of high pollution and may be shut down by the government in the future, whether GHG emission treatment is a huge expense that decreases profitability, or whether eco-friendly actions can bring additional income to the corporation, are being considered. Increased carbon information can help investors evaluate environmental risks. However, if little carbon information is available, creditors may set higher interest rates to compensate for possible environmental risks (Goss and Roberts, 2011).

Companies can actively disclose carbon information to compensate for the information asymmetry between the company and creditors, and signal to creditors that they actively take social responsibility and have low carbon risks. Both the information asymmetry theory and signaling theory explain how carbon information disclosure can decrease debt financing costs. In contrast, the signals received by creditors are not necessarily beneficial to the company. Yang et al. (2020) and Jaffe et al. (1995) believed that creating benefits for society will increase private costs and reduce market competitiveness. Companies must spend a considerable amount of funds to collect and report carbon information, and the information disclosed may be of low quality, with no reference value for investors. Creditors will regard information disclosure as an additional cost, negatively influencing investors' expectations of repayment and increasing interest rates to compensate for possible losses.

Therefore, we propose the following competitive hypothesizes.

H1a: The level of corporate carbon information disclosure is negatively correlated with the cost of debt financing.

H1b: The level of corporate carbon information disclosure is positively correlated with the cost of debt financing.

Sun (2018) believed that due to China's special institutional background, differences in the nature of corporate property rights affect the influence of CSR on debt financing costs. Many studies have shown that the government has the power to encourage enterprises, especially SOEs, to fulfill their social responsibilities. Pan et al. (2019) argued that the allocation of property rights often has a fundamental impact on the cost of debt financing.

Financing difficulties and expensive financing have always been important bottlenecks in the development of private enterprises (Cull and Xu, 2005) because the problem of information asymmetry in debt financing is more serious in private enterprises. This is primarily because private enterprises have more governance defects and endowment disadvantages than SOEs (Bai and Lian, 2012); thus, they lack a complete enterprise management system, and their accounting information remains low quality (Li et al., 2015). Lu et al. (2009) found that private companies face "credit discrimination" when monetary policy is tightened due to lack of government security compared with SOEs, making investors become more prudent. Moreover, private enterprises receive less attention from society, and many have difficulty collecting comprehensive and reliable information, resulting in more serious information asymmetry. Therefore, information quality becomes a competitive factor among private enterprises. Consequently, the voluntary disclosure of environmental information becomes an added resource for creditors to evaluate an entity ' s risk.

SOEs are more conscious of complying with government rules and regulations, and have developed sound internal control systems that determine the quality of the information disclosed (Goh and Li, 2011). They generally represent the developing trend of enterprises in a country and, therefore, receive more attention from society. Information collections of SOEs are always more comprehensive and detailed than those of private enterprises, and analysts organize information to make it more valuable.

Owing to the ownership differences mentioned above, we propose the following hypothesis.

H2: The impact of carbon information disclosure on the cost of debt financing is different between SOEs and private enterprises.

3. Data

3.1. Interest expense and liability

The financial data for each company were collected from the China Stock Market and Accounting Research Database (https://data. csmar. com/) and Wind database (https://www. wind. com. cn/). The data include comprehensive accounting data from the financial reports of companies listed on the Shanghai and Shenzhen stock exchanges each year. We use interest expenses, which exclude the capitalized interest and liability amounts of consolidated financial statements from 2008 to 2019. Financial institutions were excluded because they were beyond the scope of this study.

3.2. Carbon information disclosure

CSR reports were from the Juchao Information Network (http:// www.cninfo.com.cn/new/index) for 2008-2019. Each report was marked according to the criteria developed to obtain the carbon information disclosure score. The evaluation criteria were established based on the CDP, an authoritative environmental reporting system that focuses on carbon disclosure worldwide. The CDP framework includes four aspects: ф risks, opportunities, emission reduction targets, and strategies caused by climate change, (2) greenhouse gas emission reduction accounting, (3) greenhouse gas emission reduction management, and @ climate change governance.

In this paper, the marking criteria was formed based on the document "Linking GRI and CDP" (https://www. globalreporting. org/ search/? query = Linking + GRI + and + CDP) and used the method of content analysis. The GRI and CDP are mature systems that direct the global standardization of corporate natural capital disclosure, and both have a comprehensive information scope referring to economic, environmental, and social aspects. All 17 scoring criteria are listed in Appendix A. We ensured that each criterion was assessed by at least two people, and the score was checked at least once by group members. We totaled the scores of all items to obtain a yearly carbon information disclosure index (CID) for 881 companies, including 476 stated-owned enterprises and 405 private enterprises.

4. Methods

To obtain the CID, the scores of all criteria for each company were summarized to attain the total score, ... and divided by the maximum score, ... .

... (1)

Some studies measure the cost of debt using a firm's credit rating assigned by the S&P or the yield spread for publicly traded and se- nior unsecured bonds (Bradley and Chen, 2011). Because China does not have an authoritative debt rating agency and the average yield to maturity of corporate debt is difficult to obtain, we used to the calculation method of Li and Liu (2009) to measure the cost of debt by interest expense ratio (GOST). Average liability is the average figure of liabilities at the beginning (end of the previous year) and end of the year.

... (2)

Considering that carbon information disclosure influences company performance after a lag period, we use a one-period lag of debt financing costs to run the regression on carbon information disclosure.

Larger companies generally have more mortgage assets, and their cash flows are less likely to be affected by negative shocks; thus, they have a lower risk of default (Zhou et al., 2017). The larger the company, the stronger its ability to withstand various risks and the lower its default risk and debt cost. Therefore, enterprise size (SIZE) was selected as the control variable, measured as the logarithm of total assets at the end of the period.

Minnis (2011) found that lenders place more weight on financial information when determining interest rates because it reflects the operating capabilities of a company; thus, lenders can assess their ability to repay. Therefore, creditors set lower interest rates for companies with good financial performance. This study selected the asset - liability ratio (LEV), return on assets (ROA), earnings per share (EPS), and growth rate of revenue (GROWTH) as measures of operating capabilities and financial performance.

In addition to corporate financial indicators, Bradley and Chen (2011) showed that corporate governance affects the cost of debt. Using Ni and Xu (2016) as a reference, this study selected the concentration of equity, the sum of the proportions of the top ten shareholders (SHARE), as an indicator to measure the level of internal controls of the enterprise.

In addition, some industries may pay more attention to carbon information disclosure owing to the nature of high pollution.

We estimate the following regression equation: this study also considered the heterogeneity of different industries using fixed effects.

... (3)

Where, represents industry fixed effects, represents year fixed effects, and eu is the residual error. Equation (3) was used to test the proposed hypotheses. We used the lagged value of the cost of debt (L. COST) to control for possible reverse causality. All control variables are explained above. If ćz^O and is statistically significant, the level of carbon information disclosure negatively impacts the corporate debt financing cost. If ^>0 and is statistically significant, the opposite relationship is demonstrated. Table 1 presents the descriptive results.

Table 2 shows the Pearson correlation analysis of the main research and control variables, with the cost of corporate debt financing being negatively correlated with the level of carbon information disclosure. The correlation coefficients among all explanatory variables are <0.7. The VIE of each variable is <5 and the average VIE is 1.89, as shown in Table 3. Therefore, the multicollinearity of the model is weak.

5. Results

5.1. Relationship between cost of debt and carbon disclosure

We estimated the cost of debt using ordinary least squares (OLS) and the regression results are presented in Table 4. The CID cient of was significant at the 1% level. This indicates that each percentage-point increase in CID will result in a 0.01 reduction in debt financing costs. For example, a company with liabilities at a sample mean level of 24.67 billion can save 9.43 million in debt financing costs by disclosing one more point of carbon information. The results demonstrate that companies can alleviate the problem of information asymmetry by actively disclosing more carbon information, positively conveying to creditors that they are of low environmental risk and are able to pay back their debt in a timely manner. Therefore, Hl a is validated. This outcome is consistent with the conclusions of Goss and Roberts (2011), in which creditors set higher rates for companies with insufficient environmental information disclosure to compensate for possible risks, and the theory of Li and Wang (2011), in which enhancing information transparency can help reduce information asymmetry between enterprises and creditors.

As shown in Columns (2) and (3) in Table 4, the CID coefficients for SOEs and private enterprises at the 5% and 1% significance levels indicate that the level of carbon information disclosure negatively impacts the cost of debt financing in both types of enterprises. The results pass Fisher's permutation test with an empirical P-value of 0.009, showing that the two coefficients are comparable, even with a sample quantity difference between the two groups. Evidently, CID has a greater impact on the cost of debt of private enterprises. This may be because SOEs receive more international and social attention, and therefore have more channels to transmit information. In addition, the government has clarified the requirements for information disclosure of most SOEs, which can also alleviate information asymmetry problems. Consequently, creditors are not sensitive to the SOEs voluntary disclosure of carbon information.

The results imply that investors value voluntary disclosure by private enterprises more; thus, this can compensate for drawbacks, such as fund security, government support, and credit discrimination, when competing for financing resources with SOEs. Although the marginal effect is limited for SOEs, carbon disclosure can benefit them by demonstrating social responsibility and building an environmentally friendly image. Because SOEs are safe investment objects, their interest rates may already be low after considering size, revenue, and profitability. Carbon disclosure can further reduce the cost of debt for SOEs, saving government funds such that money can be invested in other areas that benefit society.

5.2. Robustness validation

Interest cost may not be the only element of debt financing cost, and companies may have to pay additional fees and other expenses (Li and Liu, 2009). Hence, we added fees and other financial expenses to the original interest to obtain the total financial expenses. The ratio of total financial expenses to average total liabilities was used to substitute for debt financing costs. Table 5 presents the results. The empirical P-values indicate that the two coefficients are comparable, and all results are consistent with previous results.

Counterfactual causal inference can better solve the endogeneity problem of OLS owing to missing variables. Because a companys carbon information disclosure level may not be randomly assigned or determined exogenously, we referred to Ni and Kongs (2016) method using PSM kernel matching to resolve the problem. The main idea of PSM is to identify samples that are similar to individuals in the control group. The two matched sets of samples were similar or even the same in other aspects, except for the substantial differences in the carbon information disclosed. Consequently, this study analyzed the net impact of carbon disclosure levels on the cost of debt financing.

We used the annual median carbon information disclosure scores to divide all samples into a group with a high disclosure level (treatment group) and a group with a low disclosure level (control group). Equation (4) affects the level of corporate carbon disclosure.

... (4)

Where, ... is the set of covariates and ... is the propensity score indicating the possibility of a sample entering the treatment group.

Equation (5) calculates the average impact (ATT) of the level of carbon information disclosure.

... (5)

Where, Nx is the sample size of the treatment group, ... and ... are sums of the samples from the treatment and control groups, respectively, and ... is the weight of matching which refers to Ni and Kong (2016).

Table 6 shows the results of PSM kernel matching after controlling for industry and year. The difference of ATT between treatment and control groups is -0.001 773 222 and the T-stat is -2.98 (|-2.98| >2.58), making it is significant at the 1% level.

Table 7 shows whether the treatment effectively balanced the data. The biases of all variables is <10% and the reduction of each bias is significant. The results of the t-test did not reject the hypothesis that there is no system difference between the treatment group and control group. Therefore, the results of PSM are reliable. Figure 1 shows that most samples are supported and only a few samples are lost.

Table 8 presents the regression results obtained using the matched samples. The coefficients of CID are all significant at the 1% level. The difference between the SOEs and private enterprises is also significant consistent with the OLS results.

6. Further analysis

This section analyzes heterogeneity from three aspects-environmental supervision pressure from the government or public institutions, residents environmental awareness and market competition- to evaluate how investors react to carbon disclosure under different conditions.

6.1. Environmental supervision pressure

The introduction of Fair Disclosure in 2000 prompted research on government regulations and information disclosure in academia. Many scholars discovered that environmental supervision mechanisms have a significant impact on environmental information disclosure. Shen and Feng (2012) believed that because of factors such as legal regulations and economic development levels, the regulatory pressures in different regions significantly differ. Hence, we can speculate that in regions with different regulatory pressures, the impact of carbon information disclosure on debt financing cost also differs.

This paper adopts the Pollution Information Transparency Index (PITI), published by Institute of Public and Environmental Affairs, which is generally used to measure environmental supervision pressure (Ye et al., 2015). The PITI assesses the information disclosure of pollution sources in 120 cities across the country, reflecting the level of information disclosure by local environmental protection agencies. We extracted the PITI for each city from 2008 to 2019 to obtain an environmental regulatory pressure indicator based on the city in which the company was registered. For cities not listed among the 120 cities, the PITI of the provincial capital cities in which they were located was used instead. Taking the median PITI of the 120 cities as the dividing line, all companies were divided into groups with high and low environmental supervision.

The results in Table 9 show that the impact of carbon information disclosure on the cost of debt financing in areas with high PITI is not as significant as that in areas with low PITI. A high PITI indicates that government departments disclose more comprehensive carbon information as required. Creditors can judge a company's environmental risks based on the carbon emission information disclosed by the environmental protection department where the company is located, weakening the problem of information asymmetry. When the level of supervision is low, creditors cannot obtain sufficient carbon information from regional environmental authorities, making them more sensitive to the carbon information that companies actively disclose.

6.2. Residents' environmental awareness

The external stakeholders of an enterprise include government departments and the public. Residents are also supervisors of corporate behavior, and their environmental awareness and behavioral tendencies affect the development of the company and other stakeholders ' evaluations of the company. Therefore, in regions where residents have varied levels of environmental awareness, carbon information disclosure has different effects on debt financing costs.

Ouyang et al. (2015) calculated the environmental awareness scores of residents in 31 regions across the country with data from the China Comprehensive Social Survey using the New Environmental Paradigm. Based on their measurement results, we used the median scores of the 31 regions as the dividing line. The sample data were divided based on the city where the sample company was registered into groups with strong and weak environmental awareness of residents.

The CID coefficient of residents with strong environmental awareness was not significant, indicating that the relationship between carbon information disclosure and debt financing costs is weak in areas with strong environmental awareness. Residents with strong environmental awareness are more concerned about carbon emission reduction and carbon pollution and are more active in practicing environmental behaviors, such as proactively buying low-carbon products and disseminating low-carbon concepts on social media. Therefore, creditors can evaluate a company's carbon emission-related risks through residents ' behavior or the direction of public opinion, alleviating the problem of information asymmetry to a certain extent and forming a signal transmission mechanism.

6.3. Product market competition

Many companies use low-carbon concepts to compete in the market. Companies can pass on green ideas through various information channels, such as product packaging, media, and news. Environmental protection organizations also evaluate companies and deliver information to interested parties in the form of analytical reports. Owing to the fierce competition among enterprises, information on the market becomes more open and transparent, making it easier for stakeholders to understand the risk of an enterprise. This study analyzed this aspect based on the degree of competition in different product markets.

The degree of product market competition is measured using the industry-adjusted Lerner index, that is, the company's operating profit rate minus the industry's average operating profit rate. This index reflects product market competition between different companies in the same industry. The smaller the Lerner value, the greater the business similarity between the company and other competitors in the industry and the more intense the competition it faces. In contrast, the larger the Lerner value, the smaller the business similarity and the less competition it faces. In this study, the median market competition was used to divide the sample into two groups.

The CID coefficient was not significant in the group with a high degree of market competition, indicating that carbon disclosures have no obvious effects on debt financing costs when product market competition is high. According to the comparison hypothesis of sufficient information on competition, market competition can help stakeholders understand the business situation of an enterprise from the outside. Creditors can obtain carbon information from third parties through various signal transmission channels, and conduct more objective risk assessments. However, the information disclosed by an enterprise may have the problem of selective disclosure, deliberately concealing negative carbon information to obtain a better reputation, thereby making the reference value low.

7. Conclusion

This study used data of listed companies from 2008 to 2019 to establish a carbon information disclosure evaluation standard based on the CDP and GRI, and manually scored social responsibility reports issued by all companies from 2008 to 2019. We explored the impact of carbon information disclosure on debt costs using empirical analysis and found that the higher the quality of carbon information disclosure, the lower the debt financing cost. This relationship differs for SOEs and private enterprises. By actively disclosing carbon information, companies can alleviate the information asymmetry between their management and external stakeholders, send positive signals to creditors to improve their investment confidence, and solve the problem of credit mismatches to obtain lower interest rates. In terms of the nature of enterprises, SOEs are more regulated by the government, and thus require mandatory information disclosure, making creditors less sensitive to the carbon information they disclose.

Through further analysis, the impact of heterogeneity on the above-mentioned relationship from the perspectives of government supervision pressure, residents' awareness, and market competition were elucidated. Stronger supervision pressure and environmental awareness can alleviate information asymmetry. Furthermore, corporate information becomes more transparent in environments with fierce competition and numerous channels for information transmission. Thus, creditors can obtain more objective information from third parties, replacing the signal transmission channel of companies disclosing carbon information through GSR reports.

This study confirms the results of previous studies on the impact of carbon information disclosure on debt financing costs and enriches conclusions in this field. Based on empirical evidence, we demonstrated that companies' disclosure of carbon information can affect creditors' investment decisions. The role of stakeholder information should be considered when establishing carbon information disclosure standards in China. Policymakers must consider whether the information can help stakeholders better monitor a company ' s environmental behavior such that the disclosed carbon information can be fully utilized. The development of a set of proper standards by the government can help guide companies' low-carbon behavior and growth. Furthermore, because government, social, and market factors can weaken companies' self-disclosure of information, companies should limit the amount of information captured by the government, society, and market because the disclosure of repeated information will only increase unnecessary costs. Disclosing information that only the company ' s internal management knows but external stakeholders are unable to obtain increases the value of carbon information to stakeholders.

Despite the study findings, some research limitations remain. Firstly, we only evaluated GSR reports, but some companies disclose carbon information in separate environmental information reports that we did not consider; therefore, some scores might not cover all the information disclosed. If possible, we will attempt to expand the scope of our evaluation in future studies to obtain more reliable results. Secondly, the impact of important events should be ignored. The publication of official environmental documents might cause creditors to pay attention to carbon information only for a short period of time, making the empirical results only applicable during the sample year. Thirdly, some PITI reports were disclosed on a two-year basis, and the data specific to each year may differ from the actual data. In the future, we intend to use or develop other index representing supervision pressure with higher time frequency.

Disclosure statement

No potential conflict of interest was reported by the authors.

Acknowledgements

This study did not receive funding from any specific institution or foundation. We thank the seminars at the Beijing Institute of Technology and the China Center for International Economic Exchanges for their comments. The views expressed in this paper are solely the authors ' own and do not necessarily reflect the views of the authors ' affiliations. The authors alone are responsible for any remaining deficiencies.

Sidebar

Received 11 September 2023; Accepted 24 January 2024

Footnote

* Corresponding author.

E-mail address: [email protected] (Y. Liang)

1 P-value is used to test whether SOEs and private enterprises have significant differences, and it is obtained through Fisher's Permutation test (Bootstrap 1 000 times).

1 Empirical P-value is used to test whether SOEs and private enterprises have significant differences, and it is obtained through Fishers Permutation test (Bootstrap 1 000 times).

References

References

Bai, J., Lian, L.S., 2012. Credit capital allocation difference: Ownership discrimination or endowment difference? Management World 6, 30-42 (in Chinese).

Botosan, C.A., 1997. Disclosure level and the cost of equity capital. Accounting Review 72, 323-349.

Bradley, M., Chen, D., 2011. Corporate governance and the cost of debt: evidence from director limited liability and indemnification provisions. Journal of Corporate Finance 17, 83-107.

Brammer, S.J., Pavelin, S., 2006. Voluntary environmental disclosures by large UK companies. Journal of Business Finance & Accounting 33, 1168-1188.

Chen, H.W., Zhou, Z.S., 2014. Internal control quality and corporate bank credit loan cost. Nankai Business Review 17(3), 103-111 (in Chinese).

Chen, H., Wang, H.Y., Jing, X., 2013. Chinese corporate carbon information disclosure: Content definition, measurement methods and status quo research. Accounting Research 12, 18-24 (in Chinese).

Clarkson, P.M., Li, Y., Richardson, G.D., Vasvari, F.P., 2008. Revisiting the relation between environmental performance and environmental disclosure: an empirical analysis. Accounting, Organizations and Society 33, 303-327.

Cui, Y.G., Li, B., Sun, Y.Q., 2016. Can corporate governance and financial situation affect the carbon disclosure quality? -based on the data of listed companies of electric power industry in China. Research on Economics and Management 37(8), 125-133 (in Chinese).

Cull, R., Xu, L., 2005. Institutions, ownership, and finance: the determinants of profit reinvestment among Chinese firms. Journal of Financial Economics 77, 117-146.

Goh, B.W., Li, D., 2011. Internal controls and conditional conservatism. Accounting Review 86, 975-1005.

Goss, A., Roberts, G.S., 2011. The impact of corporate social responsibility on the cost of bank loans. Journal of Banking & Finance 35, 1794-1810.

Hassel, L., Nilsson, H., Nyquist, S., 2005. The value relevance of environmental performance. European Accounting Review 14, 41-61.

He, Y., Tang, Q. L., Wang, K. T., 2014. Carbon information disclosure, carbon performance and cost of capital. Accounting Research 1, 79-86 (in Chinese).

Jaffe, A.B., Peterson, S.R., Portney, P.R., Stavins, R.N., 1995. Environmental regulation and the competitiveness of U. S. manufacturing: What does the evidence tell us? Journal of Economic Literature 33, 132-163.

Jaffee, D.M., Russell, T., 1984. Imperfect information, uncertainty, and credit rationing: a reply. Quarterly Journal of Economics 99, 869-872.

Jensen, M.C., Meckling, W.H., 1976. Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305-360.

Kim, J.B., Song, B.Y., Zhang, L.D., 2011. Internal control weakness and bank loan contracting: evidence from SOX section 404 disclosures. Accounting Review 86, 1157-1188.

Lemma, T.T., Feedman, M., Mlilo, M., Park, J.D., 2019. Corporate carbon risk, voluntary disclosure, and cost of capital: South African evidence. Business Strategy and the Environment 28, 111-126.

Leuz, C., Verrecchia, R. E., 2000. The economic consequences of increased disclosure. Journal of Accounting Research 38, 91-124.

Li, G.Y., Liu, L., 2009. Debt financing cost and private credit discrimination on private listed firms. Journal of Financial Economics 12, 137-150 (in Chinese).

Li, H. Y., Liu, D., 2016. Marketization process, voluntary information disclosure and cost of equity capital. Accounting Research 1, 71-78 (in Chinese).

Li, Z. J., Wang, S. P., 2011. Monetary policy, information disclosure quality and corporate debt financing. Accounting Research (10), 56-62, 97 (in Chinese).

Li, H. Y., Fu, S. Y., Wang, R. F., 2015. Construction of carbon information disclosure evaluation system. Statistics and Decision 13, 40-42 (in Chinese).

Li, L., Liu, Q. Q., Tang, D.L., 2019. Carbon performance, carbon information disclosure quality and cost of equity financing. Management Review 31(1), 221-235 (in Chinese).

Lin, Z.G., Ding, M.H., Wang, L., 2017. Government regulation, internal control defects and restoration, and the awareness of investor risk: An empirical study based on the perspective of institutional change of internal control. China Accounting Review 3, 35-66 (in Chinese).

Liu, Y.P., Yang, P.R., 2018. An empirical study on the influencing factors of carbon accounting information disclosure. Commercial Accounting 2, 11-14 (in Chinese).

Lu, Z.F., Zhu, J.G., Fan, Z., 2009. Monetary contractions, bank discrimination and the wealth loss of non-state listed firms' investors. Journal of Financial Research 8, 124-136 (in Chinese).

Ma, W., Gai, Y.X., 2019. Enterprise life cycle, carbon information disclosure and financing constraints-Empirical evidence based on heavily polluting industries. Ind. Technology Economics 38, 109-116 (in Chinese).

Meek, G.K., Roberts, C.B., Gray, S.J., 1995. Factors influencing voluntary annual report disclosures by U. S., U. K. and continental European multinational corporations. Journal of International Business Studies 26, 555-572.

Minnis, M., 2011. The value of financial statement verification in debt financing: evidence from private U.S. firms. Journal of Accounting Research 49, 457-506.

Ni, J., Kong, L.W., 2016. Environmental Information disclosure, bank credit decisions and debt financing cost: Evidence from the listed company in heavy polluting industries of a-shares in Shanghai stock market and Shenzhen stock market. Economic Review 1, 147-156 (in Chinese).

Ni, X. N., Xu, C. K., 2016. Research on government behavior, internal control quality and debt financing cost. Friends of Accounting 15, 98-101 (in Chinese).

Ouyang, B., Yuan, Z., Chen, S.J., 2015. Environmental awareness and environmental behavior of city residents in China. Economics Geography 35, 179-183 (in Chinese).

Pan, A.L., Liu, X., Wu, Q., 2019. Cross-ownership M&As, the institutional environment and the cost of debt financing of nonstate enterprises. Accounting Research 5, 3-10 (in Chinese).

Preston, L. E., O'Bannon, D.P., 1997. The corporate social-financial performance relationship. Business & Society 36, 419-429.

Sengupta, P., 1998. Corporate the disclosure quality and cost of debt. Accounting Review 73, 459-474 (in Chinese).

Shen, H.T., Feng, J., 2012. Public opinion supervision, government supervision and corporate environmental information disclosure. Accounting Research 2, 72-78 (in Chinese).

Shen, H.T., 2007. Corporate characteristics and social disclosure-Evidence from listed companies in China. Accounting Research 3, 9-16 (in Chinese).

Sun, Z.H., 2018. Financial ecological environment, social responsibility of private enterprises and debt financing costs. Friends of Accounting 16, 21-27 (in Chinese).

Tan, X., 2017. Industry competition, ownership and disclose of corporate society responsibility information: based on signaling theory. Industrial Economics Research 3, 15-28 (in Chinese).

Wang, J.M., 2008. Research on the correlation among environmental information disclosure, industry differences and supervisory system. Accounting Research 6, 54-62 (in Chinese).

Wang, W., Jiang, G.F., 2004. Information disclosure, transparency and the cost of capital. Economics Research Journal 39, 107-114 (in Chinese).

Wang, X., Xu, X. D., Wang, C., 2013. Public pressure, social reputation, inside governance and firm environmental information disclosure: the evidence from Chinese listed manufacturing firms. Nankai Business Review 16, 82-91 (in Chinese).

Wegener, M., Elayan, F.A., Felton, S., Li, J., 2013. Factors influencing corporate environment disclosures. Accounting Perspectives 12, 53-73.

Wen, S.B., Zhou, L.L., 2017. The influencing mechanism of carbon disclosure on financial performance-"Inverted U-shaped" moderating role of media governance. Management Review 29, 183-195 (in Chinese).

Xia, Q., Yang, Y.F., Zheng, J.M., 2018. Media coverage, media credibility and cost of debt. Management Review 30, 180-193 (in Chinese).

Yang, J., Yang, M., Liu, Y.C., 2020. How does carbon information disclosure affect debt financing cost-A study based on mediating effect of debt default risk. Journal of Beijing Institute of Technology (Social Sciences Edition), 22, 28-38 (in Chinese).

Yang, L., Wu, Y., Tang, Y.J., Guan, Y.M., 2017. The study on the characteristics of the corporate governance and carbon disclosure. Communication of Finance and Accounting 3, 20-25 (in Chinese).

Ye, C.G., Wang, Z., Wu, J.F., et al. 2015. External governance, environmental information disclosure and the cost of equity financing. Nankai Business Review 5, 85-96 (in Chinese).

Zhang, J., 2018. Relationship between quality of carbon information disclosure and financial performance of listed companies from the perspective of low carbon economy. Journal of Lanzhou University (Social Sciences), 46, 154-165 (in Chinese).

Zhang, C., Lv, W., 2007. Institutional investors, ultimate property rights and financing constraints. Management World 11, 119-126 (in Chinese).

Zhang, Q.L., Song, W.B., Tan, J., 2013. On carbon emission, the quality of carbon information disclosure and the firm value. Journal of Nanjing Audit University 10, 56-63 (in Chinese).

Zhang, S.H., Shi, X. X., Wen, L., 2011. Can environmental information disclosure enhance corporate value? Empirical evidence from China's Shanghai stock market. Comparative Economic & Social Systems 6, 166-173.

Zhou, Z.F., Wen, K., Zeng, H.X., 2017. Carbon risk, media attention and the cost of debt financing-Empirical evidence from Chinese listed firms of high-carbon industries. Modern Finance And Economics-Fournal of Tianjin Universitj of Finance And Economics 37, 16-32 (in Chinese).

(ProQuest: Appendix omitted.)