Abstract

Green credit is an important manifestation of commercial banks’ environmental responsibility, but few studies have examined the impact of green credit policies on the financial performance of commercial banks. Based on the panel data of 62 commercial banks in China from 2013 to 2020, this paper investigates the changes in financial performance due to the implementation of green credit policies in 2016 using the difference-in-differences (DID) method and explores the mediating mechanism of green credit affecting commercial banks’ profit. The sample was divided into a treatment and a control group. The treatment group disclosed total green credit data during the sample period, while the control group did not. The findings of this paper are as follows: (1) the implementation of the green credit policy increases the profits of commercial banks. (2) The green credit policy improves the profits of commercial banks by increasing their noninterest income and reducing their nonperforming loan ratios. (3) The green credit policy does not improve commercial banks’ profit by reducing their cost-to-income ratios. (4) The implementation of the green credit policy significantly improves the profits of banks with low (vs. high) nonperforming loan ratios. (5) Compared to large national banks, regional urban and agricultural commercial banks’ profits improve more significantly after executing the green credit policy. The research contribution of this paper provides a quantitative basis for Chinese commercial banks to improve their financial performance through the implementation of green credit in recent years and for the government to further improve green credit policies to motivate banks to implement green credit and achieve sustainable development. Then, we further discuss the implications for the prominent theoretical and managerial policies, study’s limitations, and future research directions.

1. Introduction

Climate change is becoming an increasingly obvious threat to the world. In response, 178 parties worldwide signed the Paris Agreement in 2016, making unified arrangements for global actions on climate change after 2020. By implementing the green credit policy, resources can be redistributed through financing channels to promote industrial transformation and upgrades. The green credit policy is a means of guiding the flow of funds and optimising the allocation of financial resources, which play crucial roles in transitioning to a low-carbon economy [13]. Banks are the main sponsors of green credit, which plays a core role in mobilising financial resources and allocating them to productive investments, making them important contributors to economic growth and development [4]. However, green credit and bank profit may conflict. The green credit provided by commercial banks will not necessarily improve their financial performance [5, 6], which makes them less enthusiastic to participate.

In China, commercial banks are the principal intermediaries for the primary financing of the Chinese real economy and have significantly contributed to its development. Nevertheless, the profit-maximising business behaviour of commercial banks can tilt credit resources towards the polluting sector, decreasing the quality of economic growth. To reduce environmental pollution and promote sustainable development, the Chinese government has issued green credit policies to guide the flow of credit resources and provide policy directions for sustainable development. Green credit refers to loans invested in green projects to support environmental improvement and is an essential part of green financing in China. In 2007, the Opinions on Implementing Environmental Protection Policies and Regulations to Prevent Credit Risks were issued, which called for the credit control of enterprises and projects that do not comply with industrial policies and environmental protection. The 2012 Green Credit Guidelines formulated more clearly the implementation of green credit, and now, the impact of the 2012 green credit policy is discussed more in academic circles. However, some scholars have suggested that the policy lacks incentives and rigid constraints, preventing it from achieving the expected effect [7]. In 2016, the Opinions on Building a Green Financial System were issued, which comprehensively refined green credit policies. The report included measures such as improving the statistical system of green credit and strengthening the monitoring and evaluation of the green credit implementation. This system is now mandatory and more comprehensive and has become one of the important policies for developing green credit in China. According to data from the China Banking Regulatory Commission, the green credit of commercial banks in China reached over CNY 15 trillion by the end of 2021, and the carbon dioxide equivalent emission reduction exceeded 700 million tonnes. Green credit policies have played a positive role in promoting commercial banks’ development of green credit businesses. Whether and how the implementation of the green credit policy promotes the profitability of commercial banks has become a critical issue in assessing the effect of the green policy.

Banks’ concern about environmental and social issues is not only a matter of social responsibility but is also related to the sustainable development of bank business activities. From the perspective of stakeholder and environmental risk management theories, the green credit provided by commercial banks is conducive to attracting investments and reducing operational risks. Some studies have also supported the positive impact of green credit on the performance of commercial banks. The Equator Principles (EPs) can help implement sustainable environmental and social development strategies [8, 9]. Some scholars have studied the impact of social responsibility on bank performance based on the environmental, social, and governance (ESG) evaluation system [5, 10], and the difference-in-differences (DID) method has been used to evaluate the effect of the policy implementation [2, 5]. Green credit has a positive impact on the financial performance of commercial banks, which improves commercial banks’ profitability, competitiveness, and operational security and reducing capital costs [2, 8]. However, green credit can have a negative impact on commercial bank operations [3, 5, 6]. For example, green credit may enhance commercial bank risk and offset profitability [5]. In addition, the impact of implementing green credit on the financial performance of commercial banks appears to vary significantly across regions and commercial banks [3, 4, 9]. Therefore, to better utilise their incentive effect, a systematic analysis is needed of how green credit policies affect the financial performance of commercial banks.

Commercial bank profit depends mainly on revenues and costs, with revenues including interest and noninterest income. Finger et al. [9] found that in developed (developing) countries, EPs’ adoption (a set of new environmental and social standards for global project finance) by banks is associated with an increase (decrease) in funding activities and the interest income share. In addition, Del Gaudio et al. [5] found that green credit might increase commercial bank risk, thus offsetting commercial bank earnings and decreasing commercial bank equity financing costs; however, it does not reduce the cost of bond financing for commercial banks. Asset quality directly affects costs and income, which can change nonperforming loan ratios, and the risk control level is an important influence. Therefore, in studying the impact of green credit on commercial financial performance, the mechanisms must be considered of the roles of nonperforming loans, income, and costs.

With panel data for 62 commercial banks in China from 2013 to 2020, a quasinatural experiment based on China’s green financial policies in 2016 is conducted. The experiment investigates the impact of green credit policies in differentiating the financial performance of commercial banks using the DID method and explores the mediating mechanism of green credit affecting commercial banks’ profits. The aim is to better utilise the incentive effect of green credit policies on banks. First, the impact is examined of the policies introduced in China in 2016 on the financial performance of Chinese commercial banks. Green credit policies are found to increase commercial banks’ profits. Then, the mediating mechanism is analysed of green credit affecting commercial banks’ profits. Banks improve their financial performance by enhancing their noninterest income and reducing their nonperforming loan ratios but do not effectively reduce their cost-to-income ratios. Furthermore, the differences are discussed in the effects of green credit policies on the performance of different bank types. Green financial policies contribute to the quality differences in commercial banks and help transform green performance into economic performance.

The remainder of this article is structured as follows: Section 2 provides the policy background, literature review, and hypothesis development; Section 3 presents the research design; Section 4 presents the empirical regressions; Section 5 provides the conclusions and discussion.

2. Policy Background, Literature Review, and Hypothesis Development

2.1. Literature Review

Stakeholder and environmental risk management theories have become critical foundations for explaining the impact of green credit on commercial banks’ performance. Abundant empirical studies have also provided the basis for studying the effect of green credit on the financial performance of commercial banks from several perspectives.

Stakeholder theory argues that organisations should balance the interests of various stakeholders in an integrated manner rather than focussing only on the accumulation of shareholder wealth. Companies should focus not only on their financial performance but also on their social impacts. Stakeholders generally include the enterprise, management, board of directors, shareholders, creditors, government, and so on. [11]. Positive environmental management and satisfactory environmental results can help improve stakeholders’ expectations, corporate image, and reputation [12]. According to stakeholder theory, the inclusion of various stakeholders in organisational decision-making is both an ethical imperative and a strategic resource, both of which contribute to the organisation’s competitive advantage.

Environmental risk management theory argues that default risk due to environmental issues has become an integral part of bank risk. Environmental risk may lead to debtor default, transmitted to commercial banks through loans, thus triggering environmental risk in commercial banks. Hence, commercial banks need to engage in environmental risk management. The literature focusses on the risk to debtors of environmental factors, noting that they lead to debt service pressures. Borrower characteristics such as reputation, leverage, earnings, and collateral can affect commercial bank risk [13]. Commercial banks examine borrowers’ environmental credit risk to reduce the default risk for environmental risk management [14]. The EPs make the vague environmental and social standards in project financing explicit and concrete, thereby helping the banking industry improve its environmental and social risk management system and practise sustainable development while helping implement sustainable development strategies for the environment and society. Some scholars have studied the differences between countries and banks. Eisenbach et al. [8] studied the difference in financial performance between the EP adopters and nonadopters based on data from 44 financial institutions worldwide. The results showed that the EP adopters had market share growth that was more significant, but the banks’ EP declarations did not enhance their stock market returns. Based on data from 78 commercial banks globally, Finger et al. [9] found that the EPs boosted the total loan share and net interest income of banks in developed countries but suppressed these variables in developing countries. The authors also examined the lagged impact of green credit policies on commercial banks. The authors found that green credit had a more significant positive impact on the long-term performance of commercial banks. Chen et al. [15] found that banks implementing the EPs were likely to hold more liquidity reserves due to the higher risks and costs associated with such an adoption and stricter regulations, which lead commercial banks to increase their liquidity reserves.

Environmental and social issues are not only the social responsibility of banks but are also related to their business activities and sustainable development. The ESG evaluation system examines a company’s ability to manage risk and long-term development from the perspectives of the environment, social performance, and corporate governance. Many scholars have studied the impact of corporate social responsibility (CSR) on bank performance based on the ESG evaluation system. Wu and Shen [16] used data from 162 banks in 22 countries to investigate the association between CSR and financial performance. CSR is positively associated with financial performance in terms of return on assets, return on equity, net interest income, and noninterest income. Cornett et al. [17] examined the relation between banks’ CSR and financial performance in the context of the recent financial crisis and found that financial performance was related positively and significantly to CSR scores. Brogi and Lagasio [18] investigated the association between ESG disclosures and returns on assets of American companies and found a significant and positive association between ESG and the environmental awareness of banks strongly related to profitability. Azmi et al. [19] used a dynamic panel generalised method of moments to examine the relationship between environmentally friendly activity, ESG activity, and bank value. The results showed that low ESG levels positively impacted bank value, environmentally friendly activities had the greatest effect on bank value, and ESG activity negatively affected the cost of equity but had no effect on the cost of debt. From a risk-suppression perspective, Gangi et al. [20] used a sample of 142 banks in 35 countries and found that banks more sensitive to environmental issues also showed less risk. From a reputation-enhancing perspective, Forcadell and Aracil [21] examined the performance of European banks listed in the Dow Jones Sustainability Index for 2003–2013, and the results suggested that banks’ efforts to build a reputation for CSR benefited their performance. However, some studies have posited that social responsibility has an inhibitory effect on the financial performance of commercial banks, which is mainly due to the differences between the short-term and long-term perspectives [21]. Zhou et al. [10] used the data of listed banks in China from 2008 to 2018 to investigate the CSR impact on bank financial performance. The results showed that CSR had a negative impact on short-term bank financial performance.

Green credit is one way through which banks embody social responsibility [17, 22]. Luo et al. [2] studied the impact of green credit policies on commercial bank competitiveness in China in 2012. The authors constructed a comprehensive competitiveness index system for commercial banks and found that green credit policies improved the core competitiveness of commercial banks. Furthermore, the results showed that China’s 2012 green credit policy more significantly improved the competitiveness of urban and agricultural commercial banks, banks with high credit risk, and banks with high reputation risk. Song et al. [3] used a dynamic panel model to examine the impact of green credit policies on bank performance based on 12 Chinese listed commercial banks and seven international commercial banks from the first quarter of 2008 through 2015. The results showed that green credit benefited the profitability in developed countries but inhibited the profitability of Chinese commercial banks. Del Gaudio et al. [5] used a sample of 217 green facilities financing syndication worldwide to examine the exposure of the lead bank and investigate the effect of the green propensity and the syndication structure on bank accounting performance. They found that a higher (vs. lower) propensity for green lending was associated with lower profitability, more moderate default risk, and lower credit risk. However, the risks might offset profitability, requiring public support to empower the banking sector to boost the ecological transition.

DID is a measurement method specifically used for policy effect evaluation that has gradually become widely used in many fields. Yao et al. [23] used a DID model to explore the impact of green credit policies on the performance of Chinese listed companies. The authors found that the impact was significantly negative, which was determined by increasing financing constraints and lower investment levels. Wu et al. [24] used the DID method to examine the impact of green credit policies on external financing for manufacturing enterprises from 2003 to 2016. The results showed that the green credit policy had a significant negative impact on the external financing of the manufacturing industry. Chen et al. [25] constructed a DID model to investigate and analyse the impact of green credit policies on corporate technological innovation by Chinese enterprises from 2004 to 2019. They found that green credit policies not only vigorously promoted low-carbon technological innovation but also played a more significant role in promoting low-carbon technological innovation in state-owned and ESG-certified enterprises.

These studies do not clearly answer whether the impact is positive or negative of firms’ green financial policies on the financial performance of commercial banks. Some scholars have advocated for increased green credit support from the government or relevant management authorities to promote green innovation and the green transformation of commercial banks. In contrast, other scholars have argued that green credit policies have no positive policy incentives for commercial banks. Although these scholars have studied the impact of green credit on commercial banks from various aspects, the literature still needs to be supplemented. In particular, literature is lacking that assesses the effects of the 2016 green credit policies in China. Further research is needed on how the external factor of the green credit policy acts on commercial banks to affect their performance. The literature does not fully reveal the relationship between the main mechanisms by which green credit affects the financial performance of commercial or the financial performance indicators.

Therefore, based on the literature, this paper uses the DID model to explore the impact of green credit on commercial bank performance. It then investigates the mediating mechanisms through which green credit affects commercial bank performance through revenues, costs, and risk. It aims to better utilise the incentive effect of green credit policies on banks and support sustainable development.

2.2. Hypothesis Development

The profit of commercial banks epitomises their financial performance, and they can grow their profits by improving their corporate reputation and increasing investments, both of which grow their market share and expand their business. Commercial banks who actively implement green credit thereby demonstrate that they are safeguarding stakeholders’ interests and can interact with stakeholders [26], which helps firms to attract investments, depositors to increase their deposits, and shareholders to subscribe to corporate equity [16, 27], all of which lays the foundation for business expansion and profit increase. Therefore, implementing green credit may increase commercial banks’ profits. However, CSR investments might waste scarce resources by causing companies to invest instead in socially responsible areas with low short-term returns and high externalities, which would influence companies to pursue high returns and might decrease bank profits [28]. As a result, implementing green credit might trigger a decline in commercial banks’ profits.

Green credit has a catalytic effect on commercial banks’ profits. Some scholars believe that green credit can improve the overall strength of commercial banks in terms of stock market prices and risk suppression, which helps them generate more profit. For example, Luo et al. [2] and Zhou et al. [10] found that China’s green credit policy in 2012 improved commercial banks’ overall competitiveness in terms of profits and risk. Eisenbach et al. [8] found that EPs’ implementation by commercial banks effectively grew their market share and that financial institutions that adopt the EPs have positive excess returns in terms of market share and other aspects. However, green credit has had adverse effects on commercial banks, affecting their profit growth. For instance, Del Gaudio et al. [5] found that green credit may enhance the risk of commercial banks, leading to a suppressive effect on commercial bank profit, and Song et al. [3] found that green credit hurts bank performance in some regions, mainly by suppressing commercial bank returns on total assets. Therefore, green credit policies may decrease commercial bank profits. Theoretical analyses and empirical studies have found uncertain impacts of green credit on bank profits. While some positive effects may result, some adverse effects might also result. We propose the following hypotheses:

Hypothesis 1. Green credit policies lead to higher (lower) profit for commercial banks.
Noninterest income is a direct source of bank profit, mainly derived from intermediary business, such as bills, commission sales, and consultation. The implementation of the green credit policy promotes the growth of green certification, bonds, fund sales, and other business. For example, commercial banks can increase their noninterest income by selling green bonds and insurance on a commission basis. Commercial banks actively implement green credit, which helps establish a new product and service system by taking the lead in green finance and promoting brand differentiation. Brand differentiation can be applied to business-related noninterest income in the banking sector, as it attracts customers and can be created because service quality differs across banks, allowing CSR banks to charge higher commissions and fees, which increases noninterest income [10, 16]. Therefore, the green credit policy may promote the noninterest income of commercial banks and thus increase profit.

Hypothesis 2. Green credit policies improve commercial banks’ profit by increasing noninterest income.
Cost control is a critical factor affecting commercial banks’ profits, and green credit may benefit cost control. Institutional stakeholder theory suggests that the social initiatives that affect stakeholder relations decrease transaction and agency costs. Commercial banks are shifting from a shareholder-centred approach to shareholder, government, public, and multiparty investor interests and can thereby earn a good social reputation and reduce transaction and agency costs [29]. Green credit reflects the social responsibility of commercial banks, can reduce information asymmetry with stakeholders, and win the respect of the community, reducing public relations costs. In addition, employees, when they have a choice, prefer to work for a socially responsible company. Thus, implementing green credit and creating a good image of social responsibility can reduce overhead costs [16]. The green credit policy may reduce commercial bank costs and thus enhance their profits.
Green credit may also increase costs, in the form of higher requirements for the environmental risk management level of commercial banks due to green credit. However, some banks may lack the skills to evaluate green technologies, which require strong specialisation and a high degree of technical skills. Environmental performance is costly to improve, and these costs may exceed the financial gain derived from environmental activities [30, 31]. Recent studies have found a U-shaped relationship between carbon emission and financial performance [32, 33]. Therefore, the green credit policy may increase commercial banks’ costs and thus affect banks’ profits. Thus, the following hypothesis is proposed.

Hypothesis 3. Implementation of the green credit policy facilitates a lower (higher) cost-to-income ratio, which enhances (reduces) bank profit.
Green credit policies may affect commercial banks’ profits by influencing their risk and thus their profit. Some scholars have found that green credit policies suppress commercial banks’ risks and thus affect their profit. Green credit policies may prompt commercial banks to improve customer quality, but profit maximisation goals may lead commercial banks to ignore environmental risks, increase loans to those with low credit ratings, and decrease bank loan quality, leading to an increase in nonperforming loans and ultimately to a decrease in bank profit. Strict loan approval and management help avoid loans to high-risk customers. From the portfolio diversification perspective, an increase in the share of green loans helps reduce the climate-related risks associated with brown loans, thus helping minimise credit risk and improve the stability of a financial system [34]. Cui et al. [35] suggested that China’s green credit policy reduces banks’ nonperforming loan ratios by increasing the proportion of green loans to total loans. Meanwhile, commercial banks can effectively attract high-quality customers by fulfilling their social responsibility and implementing green credit policies, thus reducing their risks. CSR actions are a means by which firms can provide a positive reputational signal [36]. Therefore, banks engaged in CSR can select and attract borrowers that are more creditworthy, which contributes to them having higher profits and better asset quality. Moreover, implementing green credit can discourage bank executives from excessive risk-taking behaviours in pursuit of high profits, which increases nonperforming loan rates. The shift from a shareholder-centric to a stakeholder-centric governance model balances the interests of the bank’s investment and noninvestment stakeholders, thereby curbing excessive risk-taking behaviours by management and protecting bank value [4]. From the customer perspective, green credit can achieve higher customer loyalty and curb bank risk [37].
Green credit policies may also increase the risk of commercial banks. For example, polluters as bank customers can significantly increase debt costs and even the risk of bankruptcy of heavy polluters due to green credit policy constraints [38]. A series of financing penalties and the investment disincentive effects of green credit would force heavy polluters to transform into green enterprises. Otherwise, they face the risk of being squeezed out of the market, which generally helps optimise the industrial structure. However, if bank interest income still relies on heavily polluting customers, increasing customer-financing constraints introduces business risks, which are then transmitted to banks. Some banks’ lack of environmental risk management may lead to green loan losses. For instance, Zhou et al. [4] found a nonsignificant association between bank green lending and risk performance across the banking sector as a whole. Compared to their large peers, city/regional-level commercial banks have more limited capacity and industry expertise and less developed risk management systems, all of which may contribute to the economic losses they experience in green lending [39]. Therefore, this paper proposes the following hypothesis:

Hypothesis 4. Implementation of the green credit policy contributes to lower (higher) nonperforming loan ratios, thus boosting (lowering) bank profit.

3. Research Design

3.1. Sample Selection and Data Source

This paper assesses the impact of the green credit policy on commercial banks’ profits and examines whether their subsequent profit growth is more pronounced compared to those banks not affected by the green credit policy. To this end, this paper identifies the treatment and control groups based on the green credit data disclosed in the social responsibility reports of commercial banks to distinguish whether commercial banks are the target of the green credit policy effect. If a bank consistently and completely disclosed total green credit data during the sample period, then the bank is determined to be better at implementing green credit and is influenced by the green credit policy; such banks comprise the treatment group. Otherwise, the bank is considered part of the control group, indicating that it has implemented green credit to a shallow extent and is not significantly affected by the green credit policy. After the policy was implemented, if the profits of the banks in the treatment group appear to increase more significantly, it indicates that the green credit policy improves commercial banks’ profits. For example, the Bank of Communications, which increased its green credit from CNY 165.836 billion to CNY 387.28 billion during the sample period, implemented the green credit policy to a large extent and was its main target, so the bank was regarded as a treatment group bank. According to the requirement of the DID method and drawing on the mainstream time interval selection method, the three years before the policy node and all years after the implementation were selected as the sample time range, i.e., 2013 to 2020. The details are shown in Table 1.

3.2. Variable Definitions

The following variables were selected in terms of asset size, cost, and risk:(1)The independent variable (PROFIT). The total profit was selected to study the effect of the green credit policy on the financial performance of commercial banks.(2)The dependent variable (TREAT∗POST). This is the double differential variable, which was the product of the bank classification and period variables.(3)The control variables. Considering the factors of commercial banks’ operational safety and asset quality and combining the mainstream selection of existing control variables, this article selects the following control variables:(1)Total assets (SIZE). Asset size is the basis of the commercial bank economy and directly impacts total profit.(2)Nonperforming loan ratio (NPLR). The nonperforming loan ratio can affect the interest income and cost of commercial banks. A higher nonperforming loan ratio indicates greater credit losses of commercial banks and higher costs paid to cover nonperforming loans, which are expected to reduce commercial banks’ profits. Through the monthly data analysis of international monetary fund countries, Albulescu [40] found that nonperforming loan ratios hurt commercial banks’ profitability.(3)Cost-to-income ratio (CIR). Profit is revenue minus costs.(4)Capital adequacy ratio (CAR). Each bank has an optimal capital-to-asset ratio, all else being equal. The capital adequacy ratios are beneficial for profit enhancement [41, 42].(5)Loan-to-deposit ratio (LDR). The loan-to-deposit ratio is the ratio of loans to deposits in commercial banks. Deposits require interest payments, while loans can be issued to obtain interest profit from asset business. More funds used for lending or even overlending can result in more profit with the same deposit size, but an extreme loan-to-deposit ratio increases the risk burden, so it was chosen as a control variable.(6)Net interest income ratio (NIIR). Most of Chinese bank income is derived from interest income, so the interest income ratio reflects the profit structure of commercial banks and is expected to impact their profits.

In addition, considering that economic growth and bank market position are important external factors affecting commercial banks’ profitability, the annual growth rate of the gross domestic product (GRGDP) and the market position (MP) of the bank are selected as the external control variables. A higher economic growth rate indicates a better external environment for banks, which is expected to be favourable for profit growth. A higher ratio of the bank asset size to GDP indicates a stronger market position for banks, which may be expected to enhance commercial bank profit.

In summary, the final model variables are shown in Table 2.

3.3. Research Model

Based on Hypothesis 1, we constructed the following model for the empirical analysis:where λ and YEAR are the individual and time effects, the subscript t denotes the year, PROFIT denotes the commercial bank profit, TREAT denotes whether the bank is affected by the green credit policy, POST denotes whether the sample is before or after the Guidance on Building a Green Financial System, and X represents the control variable. TREAT∗POST can decompose the net effect of the policy on the profit of the treatment group. We first conducted a full-sample regression to test whether the green credit policy affects the financial performance of commercial banks. If β1 is significantly greater than 0, the green credit policy substantially enhanced commercial bank financial performance; if β1 is less than 0, the policy worsened commercial bank financial performance. Furthermore, subsample regressions were conducted to observe the differences in the results of β1 coefficients under different sample regressions to identify the impact of green credit on commercial banks’ business performance.

Based on Hypotheses 2.4, we followed the mediation effect test procedure formulated by Baron and Kenny [43] to construct the test-by-test mediation effect regression model, as follows:

The meanings of the variables in equations (2)–(4) are the same as in (1). ME represents the mediating variable. In the analyses based on Hypotheses 24, the mediating variables are noninterest income (NI), nonperforming loan ratio (NPLR), and cost-income ratio (CIR), respectively.

This work uses the test-by-test method to test the mediation effect. If the coefficients of TREAT∗POST in equations (2)–(4) are significant, it indicates a significant mediation effect. If the coefficients of TREAT∗POST in (2) and (3) are significant but they are not significant in (4), there is a mediating effect. If one of the coefficients of TREAT∗POST in (2) or (3) is not significant, the bootstrap method is used to test whether there is a mediating effect. Based on this method, this paper analyses the specific mechanism of how the green credit policy affects the financial performance of commercial banks by verifying that the green credit policy affects the noninterest income, nonperforming loan ratio, and cost-income ratio.

4. Empirical Results and Analysis

4.1. Empirical Regressions

This study adopts the DID model and mainly focusses on the net profit effect of green banks relative to other banks before and after the green credit policy was implemented, namely, the coefficients of the interaction term (TREAT∗POST) in the DID model.

Table 3 reports the impact of the green credit policy on commercial bank profits for the full sample. In Column (1), we included only the core explanatory variables. In Column (2), we controlled for the time effects. In Column (3), we included only the control variables. In Column (4), we included the control variables and controlled for the time effects.

The conclusion shows that the coefficient of the cross-term TREAT∗POST was positive and greater than 0; thus, it passed the significance test. In Column (4), for example, the coefficient of the cross-term TREAT∗POST of 0.150 passed the 1% significance test. This result indicates that after the green credit policy was implemented, commercial banks in the treatment group had a more obvious profit growth, so the green credit policy increased bank profits.

Our research conclusion verifies Hypothesis 1, which is consistent with the conclusions of other studies. Luo et al. [2] and Eisenbach et al. [8] concluded that green credit helps to improve the financial performance of commercial banks. Although both theoretical and empirical studies have shown that commercial banks’ green credit policies may decrease profits, this study still supports the conclusion that commercial banks’ green credit policies are conducive to improving commercial banks’ profits.

The results of the control variables are essentially in line with expectations. Asset size and the ratio of loans to deposits were positively correlated with the profit level of commercial banks. This result indicates that the efficiency increase of credit resources used in commercial banks helps to achieve higher profit levels. The nonperforming loan and cost-to-income ratios were negatively correlated with commercial banks’ profits. Furthermore, the increasing nonperforming loan ratios and costs caused commercial banks to lose profits. The capital adequacy ratio did not effectively contribute to commercial banks’ profit growth, which might indicate an optimal capital adequacy ratio, as higher is not better. The interest income ratio had an insignificant effect on commercial banks’ profits, indicating that a higher interest income ratio is not better. In addition, the regression coefficient of the economic growth rate was negative, but the significance level was low. This result indicates that economic fluctuations have insignificant effects on commercial banks’ profit. The external market position of banks inhibits commercial bank profits. This result may imply that the excessive market influence of commercial banks, such as Chinese state-owned banks, assumes more nonprofit functions and has more diversified business objectives, thus partially inhibiting profit expansion.

4.2. Parallel Trend Test and Dynamic Effect Analysis

The validity of the DID method is based on the premise that the treatment and control groups satisfy the assumption of parallel trends before the policy is introduced. This assumption is tested by plotting the parallel profit trend and constructing a regression model, which analyses the dynamic effect of the green credit policy on the profits of commercial banks. The regression model was constructed as follows:

The difference between equations (1) and (5) is that explanatory variable TREAT∗POST is replaced with seven dummy variables. Dkit is whether the treatment group is in YEAR k; yes is 1, and no is 0. For instance, D2016 represents that the treatment group is in 2016, the year the policy was implemented. In the parallel trend test, this article focusses on whether the regression coefficients of D2013, D2014, and D2015 are significant. If not, the evolutionary trend of commercial banks in the treatment and control groups before the policy was introduced is not significantly different, and the parallel trend hypothesis is satisfied. Let us suppose that at least one of the regression coefficients of D2016, D2017, D2018, and D2019 is significant. This result would indicate a significant divergence between the profit of commercial banks in the treatment and control groups after the policy was introduced. Such a finding would both support the main regression results and be used to observe the dynamic effect of the policy on commercial banks’ profits.

Before conducting the regression analysis, we first found the parallel trends of commercial bank profit in the treatment and control groups in Figure 1. Figure 1 shows that before the policy was introduced in 2016, the profit of the treatment and control groups essentially satisfied the parallel trends. The regression results in Table 4 with insignificant values for the coefficients of D2014, D2015, and D2016 further supported the parallel trend assumption. After the policy was introduced, Figure 1 shows that the treatment and control groups’ profits diverged in 2017. Moreover, the regression coefficient of D2017 in the regression results of Table 4 was significantly positive. This result indicates that commercial banks’ profits in the treatment group increased significantly in the second year after the policy was introduced. However, the profit growth in the other years was relatively insignificant. This result indicates that although the green credit policy increased the profit of commercial banks in the treatment group, the impact was limited and lacked persistence overall.

4.3. Heterogeneity Test

We wanted to verify the variability of green policies on the profit growth of commercial banks and further enhance the robustness of the regression results. Therefore, we established a grouped regression model to study the impact of the green credit policy on the profits of different commercial banks. The regression results are shown in Table 5.

4.3.1. Bank Type

Under the current banking system in China, large, state-controlled banks and national joint-stock banks are larger in size and are more capable and willing to take on social responsibility than are urban commercial banks and are significantly different from urban and agricultural commercial banks. For example, Zhou et al. [4] found a nonsignificant association between bank green lending and risk performance across the banking sector as a whole; compared to their large peers, city/regional-level commercial banks have more limited capacity and industry expertise and less developed risk management systems, all of which may contribute to the economic losses they experience in green lending [39]. Hence, green finance policies may have different results on the profits of the two bank types.

Shown in Columns (1) and (2), we classified the banks in the treatment group according to the commercial banks—consisting of large, national banks, including the five largest state-owned banks and national joint-stock banks—and other banks, including regional urban commercial banks and agricultural and commercial banks. According to the regression results in Table 5, the regression coefficients of the TREAT∗POST variable in both Columns (1) and (2) were significantly positive. This result indicates that the green credit policy enhances the financial performance of different bank types in China. However, the coefficient of TREAT∗POST in Column (2) was slightly larger, suggesting that the implementation of green credit policies by urban and agricultural commercial banks substantially affects their profit enhancement. This finding suggests that large, national banks may tend to place more emphasis on the social benefits of lending and thus on the long-term and social benefits of green credit, such as the implementation of more favourable green credit interest rates. This emphasis results in a less-pronounced effect of green credit on their short-term profit enhancement than it does for other banks, such as urban merchant and agricultural commercial banks.

4.3.2. Nonperforming Loan Ratio

The nonperforming loan ratio reflects banks’ risk control levels, and different levels of risk control may lead to differences in the effectiveness of implementing green credit policies. The treatment group was classified according to the nonperforming loan ratio in Table 5, and the regression results are shown in Columns (3) and (4).

These results demonstrate that the regression coefficient of TREAT∗POST was significant in both Columns (3) and (4), but the coefficient in Column (4) was high. This result indicates that banks with a lower nonperforming loan ratio significantly affected their profit growth by implementing green credit. However, the theoretical analysis suggests that the implementation of green credit by banks with high nonperforming loan ratios can improve overall asset quality and reduce that ratio by investing more loan resources in green projects, which can increase commercial banks’ profits. However, this mechanism is currently not evident, probably because although banks with high nonperforming loan ratios have implemented green credit, the degree of implementation remains insufficient to reduce the overall nonperforming loan ratio.

4.4. Robustness Tests

The heterogeneity analysis in the previous section confirms the robustness of the impact of green credit policies on commercial banks’ profit from a subsample perspective. In addition, this article further conducts robustness tests based on the following ideas to obtain Table 6, which includes five robustness tests.(1)Replacing the measurement method. The propensity score matching DID model can effectively correct the selectivity bias. Therefore, this article matches the treatment and control groups based on the 1 : 2 most adjacent matching method and eliminates the samples that do not satisfy the “common support hypothesis.” The results are shown in Column (1).(2)Substituting the explained variables. In Column (2), this article uses the net profit (N-PROFIT) indicator instead of the original total profit indicator to conduct the regression. The coefficient of TREAT∗POST remained significant.(3)Substituting the explanatory variables. Considering the lag in the green credit policy implementation, we shifted the policy implementation node back to 2017 to obtain the double difference variable TREAT∗POST2. Then, we conducted the regression again to obtain Column (3), which shows that the regression coefficient of TREAT∗POST2 was significant at the 5% level.(4)Shortening the sample time horizon and excluding the samples from 2013 to 2020. The regression coefficient of the TREAT∗POST variable remained significantly positive after this treatment. The results are shown in Column (4).(5)Considering the cohort effect. The cohort effect is currently receiving much attention in the commercial bank research, and the number of studies is increasing on the cohort effect of commercial bank operations. In reality, the business decision of commercial banks to develop green credit is easily influenced by peer competition, so the cohort effect must be included for empirical testing. This article introduces a spatioeconometric model that uses a spatial weight matrix to measure the mutual influence of commercial banks. Commercial banks more often regard banks with strength comparable to themselves as competitors. The elements of the weight matrix were set based on the total asset size of the banks: aij = 1/|bibj|. The bi and bj are the averages of the ith and jth 2013–2020 asset sizes. The aij was positively correlated with the difference in asset size, i.e., commercial banks more similar in market position had a more obvious cohort effect with each other. Based on the weight matrix setting, the spatial error model (Column (5)), spatial lag model (Column (6)), and spatial Durbin model (Column (7)) were regressed, and the regression coefficient remained significantly positive, so after considering the cohort effect, green credit still had a significant positive effect on commercial banks’ profits

4.5. Placebo Test

In the original regression, the profit level of the treatment group was higher than that of the control group after the green credit policy was implemented. Nevertheless, this difference might have arisen from other unobservable random factors. A practical solution to this problem is to adopt a placebo test. If the regression model here has systematic biases or other omitted factors, the results obtained using random sampling to constitute a dummy treatment group will be similar to the results of the baseline regression. Otherwise, the results of the baseline regression of this paper are valid. Accordingly, a placebo test is conducted by random sampling in this paper. One thousand random samples were used to construct the virtual treatment group and carry out the regression again, and the coefficient values were obtained of the regression equation, t values, and values based on the virtual treatment group. The scatter plots of the regression coefficients and values are shown in Figure 2.

The analysis showed that compared with the coefficient of TREAT∗POST (0.150) in Table 3, Column (4), the TREAT∗POST values of 1,000 random samples were distributed around a centre of 0. Most of the regression coefficients fell into the region with a value greater than 10%, and the mean value was −0.0014872, which was much smaller than this paper’s true DID coefficient value. In summary, the placebo test shows that the benchmark regression in this paper is true and valid. That is, the effect of the green credit policy on the profits of commercial banks was not significant in the vast majority of the dummy treatment group after random sampling. This result indicates that the results of the benchmark regression were not disturbed by certain random factors. The original regression results have a certain inevitability.

4.6. Further Analyses

To further examine the specific effects of green credit policies on commercial banks’ profits, this section constructs a mediating effect model. The model investigates the impact of the green credit policy on commercial banks’ revenues, nonperforming loan ratios, and cost-to-income ratios from the perspective of revenues, risks, and costs.

4.6.1. Green Credit Policy Impact on Commercial Banks’ Income

Noninterest income is an important source of commercial bank profit. This section focusses on the impact of the green credit policy on interest and noninterest income and on the shares of both. A mediating effect model is further employed to analyse how the green credit policy affects commercial banks’ profit by influencing their income. Figure 3 and Table 7 report the empirical results.

Shown in Table 7, Column (1), the total interest income of commercial banks was used to study the impact of the green credit policy on the interest income of commercial banks. The regression results show that the interest income level of commercial banks was significantly enhanced after the green credit policy was implemented, which is more consistent with the previous analysis. The green credit policy motivates commercial banks to actively offer green credit. Nevertheless, green credit raises the financing constraints of traditional customers. At the same time, environmental protection projects have a long revenue cycle, and a short-term green credit policy does not significantly enhance commercial banks’ interest income.

Shown in Table 7, Column (3), net fee and commission income are the primary sources of noninterest income. We used commercial banks’ net fees and commission income to study how the green credit policy affected commercial banks’ noninterest income. The results in Column (3) show that the coefficient of TREAT∗POST was 0.173, which passed the significance test at the 5% level, indicating that the green financial policy enhanced the noninterest income of commercial banks. Columns (2) and (4) combined show that the regression coefficient of TREAT∗POST in Column (4) was 0.205, which passed the significance test at a 1% level, indicating that noninterest income significantly enhanced the profit levels of commercial banks. Thus, the green credit policy mainly enhances commercial banks’ profit by boosting their noninterest income and therefore their profit. This result indicates that commercial banks actively implement green credit policies to create new profit growth points [9, 16].

Figure 3 shows the parallel trend plots to obtain the development trends of noninterest income. The noninterest income levels of the treatment and control groups maintained the same trend before 2016. Meanwhile, after 2016, the noninterest income of the banks in the treatment group increased. These results indicate that the green credit policy had a significant impact on the noninterest income of commercial banks.

4.6.2. Green Credit Policy Impact on Commercial Banks’ Costs

To test Hypothesis 3, this paper used the cost-to-income ratio as a mediating variable to investigate whether green credit policies suppress commercial banks’ costs and thus enhance their profit. Figure 4 and Table 8 report the relationships among the green credit policy, the cost-to-income ratio, and total profit.

Figure 4 shows that before 2016, the cost-to-income ratio of banks in the treatment and control groups showed a certain parallel trend. After the policy was introduced—in 2017, 2018, and 2019—the cost-to-income ratio of the two types of banks still maintained a parallel trend. This result indicates that the green credit policy may have no significant impact on the short-term cost-to-income ratio for commercial banks.

A further analysis in Column (2) shows that the regression coefficient of the TREAT∗POST variable was not significant, indicating that the green credit policy failed to effectively curb the costs of commercial banks. Furthermore, that the mediating effect was not significant of green credit through costs affecting the profits of commercial banks. This result suggests that the interaction may stem from the immature development and relatively small scale of green credit and the large, upfront, fixed investment. These factors result in the underutilisation of the cost-curbing effect of green credit on commercial banks. To further verify the stability of the findings, we used the bootstrap method to test the mediation effect, and the conclusion shows that the mediation effect was still insignificant.

4.6.3. Green Credit Policy Impact on Commercial Banks’ Risk

To test Hypothesis 4, this paper focusses on whether green credit policies can affect commercial banks’ nonperforming loan ratios and thus their risks and profit. Figure 5 and Table 9 report the regression results of the green credit policies affecting commercial banks’ nonperforming loan ratios.

We plotted the parallel trends of the nonperforming loan ratios of the treatment and control groups and showed that the nonperforming loan ratios of the treatment and control groups maintained parallel upward trends until 2016. At that time, the nonperforming loan ratios of the banks in the treatment group were significantly suppressed, while those of the control group showed a significant increase.

A further regression analysis was conducted to obtain Table 9, Columns 1–3. In Column (2), the regression coefficient of the TREAT∗POST variable was −0.213, which passed the 10% significance test. In Columns 1 and 3, the TREAT∗POST coefficients were significantly positive, indicating that the green credit policy significantly suppressed the nonperforming loan ratios of commercial banks and enhanced their profits. The current development of green credit may create financing constraints for traditional high-pollution customers and trigger loan risks for commercial banks. However, this finding suggests that the development of green credit by commercial banks can improve the quality of loans and thus reduce the loss of income caused by nonperforming loans, which increases commercial bank profits.

5. Empirical Results and Analysis

5.1. Theoretical Implications

As global environmental problems intensify, an increasing number of countries are attaching importance to guiding commercial banks to actively carry out green credit. Such countries are using green credit policies to promote the achievement of the “double carbon goal.” This goal has become an essential factor affecting the operations and sustainable development of commercial banks. In 2016, the Chinese government issued “Guidance on Building a Green Financial System.” This study is the first to propose the construction of a green financial system, the role of green credit as a significant component of the green financial system, and a comprehensive refinement of the green credit policy. The last has become one of the critical policies for the development of green credit in China, as it provides policy support and guarantees for commercial banks to provide green credit. Financial performance is an important factor for commercial banks, and systematically assessing the impact of green credit policies on the financial performance of commercial banks is significant to better understand the incentive role that green credit policies play for banks. This paper investigates the impact of green credit policies in differentiating the financial performance of commercial banks using the DID model for 62 commercial banks in China from 2013 to 2020 panel data. The theoretical implications are summarised as follows.

Based on the China context, this paper finds that green credit policies benefit commercial banks’ profits. This result is consistent with Luo et al. [2] and Eisenbach et al. [8], who concluded that green credit has a positive effect on commercial banks’ performance. The green credit policy and commercial bank development in China share a win-win relationship: commercial banks who actively practise green credit reap higher profit growth and partially alleviate the current concerns about social responsibility-induced corporate profit loss.

This paper finds that noninterest income and the nonperforming loan ratio constitute the mediating mechanism by which the green credit policy improves commercial banks’ profit. Furthermore, the green credit policy improves commercial banks’ profit by increasing their noninterest income and suppressing their nonperforming loan ratios.

This paper finds that the green credit policy does not improve commercial banks’ profit by reducing their cost-to-income ratio. In addition, its “cost containment effect” is not fully exploited, which limits the positive impact of the green credit policy on commercial banks to some extent. These findings are consistent with the conclusion that green credit has a positive impact on commercial banks in developing countries in the long term, while the green credit effect credit is affected in the short term by cost escalation [9].

The profit of banks with low (vs. high) nonperforming loan ratios increased more significantly after the green credit policy was implemented. This result indicates that a good risk management level helps commercial banks better employ green credit and thus increase their profit. Commercial banks should focus on controlling risks when implementing the green credit policy.

Compared to large, national banks, urban and agricultural commercial banks improved their profit more significantly after they implemented the green credit policy. This result may indicate that large national banks are more aware of their social responsibilities and pay more attention to long-term benefits, resulting in a relatively minor increase in profit. Green credit has different performance impacts for different banks. This paper finds that the green credit policy has a more significant positive impact on Chinese urban and agricultural commercial banks and banks with low nonperforming loan ratios, which is consistent with Luo et al. [2]. This result confirms the differences in the impacts of the green credit policy on different bank types.

5.2. Policy Implications

We propose the following policy recommendations based on the above analyses and conclusions.

The level of green credit development in China can still be improved, as the current scale of green credit accounts for only approximately 10% of total credit. Given that green credit policies can effectively grow commercial banks’ profits, the Chinese government should further improve its green credit policy to promote commercial banks’ performance and unify social responsibility and economic value. Currently, China has more than 4,000 commercial banks, but this paper includes only 21 commercial banks in the high environmental responsibility group, and the breadth of green credit implementation needs further improvement. Given the role of green credit policies in promoting commercial bank performance [2, 8], the Chinese government should encourage more banks to implement green credit and enhance the level of green credit development.

The green credit policy has had a complex impact on the financial performance of commercial banks, which is reflected in the impacts on different banks. China has numerous bank types, and different banks have different business philosophies and cost and income structures. Therefore, in the policy implementation process, different banks should be treated differently. In terms of performance assessment, for large, state-owned banks that are more fully committed to social responsibility, the focus should be on assessing the implementation of green credit, weakening the short-term assessment of economic performance such as profit, and extending the profit assessment cycle to motivate large, state-owned banks to pay more attention to long-term interests. For urban commercial and agricultural banks, emphasis should be placed on preferential measures such as financial subsidies and tax concessions. Doing so will reduce the burden of implementing green credit for these banks and alleviate their profit pressure. The implementation of green credit also needs a perfect risk management mechanism. Green credit also has environmental risks, so commercial banks should focus on improving environmental risk management to better play their role in green credit.

The green credit policy effectively enhances commercial banks’ noninterest income and suppresses their nonperforming loan ratios, which is the primary channel for promoting commercial profit. In this regard, commercial banks should actively develop their green credit business, strengthen their diversified services, and improve their risk control ability, business level, and other specific measures to fulfil their social responsibility while improving their financial performance.

Green credit policies have not effectively reduced the cost-to-income ratio of commercial banks, which might be too high for initial fixed investments, but the cost-suppression effect has not been brought into full play due to the insufficient overall scale of green credit. In this regard, we should focus on reducing commercial bank costs and improving the external environment for the commercial development of green credit. In this regard, the Chinese government can improve the information disclosure mechanism to reduce the cost of identifying green credit risks. The cost of green credit financing can be reduced through central bank refinancing, interest rate subsidies, and guarantees.

5.3. Suggestions for Future Research

This paper discusses and studies the effect of green credit policies on bank performance, but some limitations still need to be mentioned and improved upon. (1) The research object of this paper is China, and the findings are explanatory for the Chinese banking industry. A cross-country sample could be attempted in the future to study the net effect of China’s green credit policy on the performance improvement of Chinese commercial banks by examining the differences between Chinese and foreign commercial banks. (2) This paper analyses the overall impact of China’s green credit policy on the financial performance of Chinese commercial banks, mainly using the DID approach. To analyse the impact on the financial performance of individual banks, the synthetic control method can be used to discuss in depth the impact of green credit policies on the financial performance of commercial banks with different asset sizes, risk levels, and business objectives and to improve the relevance of the policies. (3) Commercial banks can not only fulfil their social responsibility by implementing green credit but can also assume their social responsibility through securities’ underwriting, etc. Considering these factors, future studies should examine the impact of commercial banks’ business on their performance by carrying out green bond and green fund reselling.

Data Availability

Data supporting the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

This research was financially supported by the Tianjin Finance Society Key Fund (TJX202107).