Does ESG Performance Reduce Bankruptcy Risk?

Abstract

This study examines how environmental, social, and governance (ESG) performance affects firms’ bankruptcy risk and explores the mechanisms linking ESG engagement to financial stability. Using a panel dataset of Chinese-listed firms from 2009 to 2022, we employ multivariate regression analyses, instrumental variable estimation, and robustness tests to address potential endogeneity. The results indicate that higher ESG performance significantly reduces bankruptcy risk. Mechanism analyses reveal that ESG engagement lowers bankruptcy risk by improving information transparency, alleviating financing constraints, enhancing operating performance, and reducing leverage. The effect is more pronounced for non-state-owned enterprises, firms in economically developed regions, highly competitive industries, and those in the growth and maturity stages. Among the three ESG pillars, corporate governance exerts the strongest influence on mitigating bankruptcy risk. These findings provide new evidence from an emerging market and offer important implications for sustainable corporate finance and risk management.

1. Introduction

Over the past two decades, Environmental, Social, and Governance (ESG) practices have drawn substantial scholarly attention. Since the concept of ESG was first introduced in the UN Global Compact (2004) report, numerous studies have analyzed the financial and strategic outcomes of ESG engagement. Extensive studies have explored the relation between ESG activities and various aspects of corporate finance, including market valuation, risk, performance, and ownership characteristics (Gillan et al., 2021). Despite these advances, Gillan et al. (2021) note that many research gaps remain, particularly regarding the causal mechanisms linking ESG performance and firm-level financial outcomes.

While prior studies have investigated the association between ESG activities and risk, empirical evidence on the causal link between ESG performance and bankruptcy risk remains limited—especially in the context of emerging markets. Existing studies generally report that stronger ESG engagement is associated with lower distress risk (Sun & Cui, 2014; Badayi et al., 2021; H. Li et al., 2022; Do, 2022), yet most focus on developed economies or employ narrow samples. The absence of comprehensive evidence from emerging markets such as China represents a critical gap in the literature.

This study addresses that gap by examining whether and how ESG performance affects firms’ bankruptcy risk using a large sample of Chinese listed firms from 2009 to 2022. Understanding this relationship is particularly relevant for China, where bankruptcy risk has become an increasingly urgent policy issue. China’s economic growth has slowed noticeably in recent years, especially following the COVID-19 pandemic. In 2023 alone, more than 500,000 enterprises were deregistered or terminated, primarily due to financial distress, and this trend has persisted into 2024. How to revitalize financially distressed firms and prevent systemic risk has thus become an exigent concern for policymakers.

At the same time, promoting ESG practices has emerged as a national priority. The Chinese government has committed to accomplishing carbon peaking by 2030 and carbon neutrality by 2060, making ESG integration an essential component of the country’s sustainable development agenda (Ji et al., 2022). The surge of green finance—reaching $4.1 trillion in 2023—underscores the rapid expansion of ESG-related investment in China. Against this backdrop, our study investigates whether, and through which channels, ESG performance can mitigate firms’ bankruptcy risk.

Our research extends the extant literature in several important ways. First, we provide robust evidence on the negative relationship between ESG performance and bankruptcy risk in an emerging market setting. Prior studies, such as H. Li et al. (2022); Sun and Cui (2014), and Maquieira et al. (2024), often rely on shorter sample periods or smaller datasets, or use samples from economies with institutional structures distinct from China’s mixed economic system (Do, 2022; Badayi et al., 2021). By contrast, we use a comprehensive dataset of Chinese listed firms spanning 2009–2022, offering a more extensive and updated analysis. Moreover, while many prior studies focus on default or credit risk proxies, we employ the Altman Z-score as a direct and well-established measure of bankruptcy risk, which provides a more relevant lens for assessing firm solvency.

Second, our empirical analyses employ multiple econometric techniques—including instrumental variable (IV) estimation, propensity score matching (PSM), and robustness checks using alternative measures of both bankruptcy risk (Ohlson’s O-score) and ESG performance (Bloomberg ESG scores)—to strengthen causal inference. Our baseline results show that better ESG performance significantly reduces firms’ bankruptcy risk, a finding that remains consistent across all robustness tests.

Third, we identify and empirically validate four mechanisms through which ESG performance mitigates bankruptcy risk: (1) improving information transparency (Drempetic et al., 2020), (2) alleviating financing constraints (Zhang et al., 2020; Yu, 2023), (3) enhancing operating performance (Wang et al., 2022; Xue et al., 2022), and (4) lowering leverage ratios (Ellul & Pagano, 2019). These findings not only confirm the multifaceted financial benefits of ESG engagement but also highlight specific channels that firms and policymakers can target to enhance corporate resilience.

Fourth, we extend the literature by decomposing ESG performance into its three core components—environmental, social, and governance (ESG pillars)—and examining their respective effects on bankruptcy risk. In contrast to findings for European banks, where environmental and social factors dominate (Palmieri et al., 2024; Maquieira et al., 2024), our results show that the governance pillar plays the most prominent role in reducing bankruptcy risk among Chinese firms. This underscores the importance of strong governance structures in emerging markets, where institutional frameworks and regulatory enforcement remain developing.

Finally, our study uses a novel instrumental variable—the density of Confucian temples—to address potential endogeneity in ESG performance. This cultural proxy captures the influence of Confucian values on ethical conduct and social responsibility (Liu et al., 2023; F. Li et al., 2019), offering a theoretically grounded and empirically robust instrument that enriches future ESG research in cross-cultural contexts.

In sum, our study makes contributions to the ESG literature by providing comprehensive evidence from the world’s largest emerging market on the causal relation between ESG performance and bankruptcy risk, the mechanisms underlying this relationship, and the distinct roles of the three ESG pillars. The findings yield valuable policy implications for promoting corporate sustainability and financial stability in China and other emerging economies.

The remainder of this paper is organized as follows. Section 2 reviews the relevant literature and formulates the hypotheses. Section 3 explains the data and variable construction. Section 4 outlines the empirical strategy and presents the baseline results. Section 5 provides extended analyses, including mechanism tests, heterogeneity analyses, and the examination of the three ESG pillars. Section 6 discusses the policy implications, and Section 7 concludes. To enhance readability, Table 13 summarizes the hypotheses, mechanisms, and key findings of the study.

2. Literature Review and Hypotheses Development

Prior studies have commonly drawn on stakeholder theory and resource dependence theory to explain the relationship between ESG performance and bankruptcy risk. According to stakeholder theory, a firm’s development is linked to its relationships with various stakeholders (Freeman & Evan, 1990). Strong ESG performance provides stakeholders with positive and credible signals about a firm’s values and practices—signals that often complement, rather than duplicate, financial performance information—thereby enhancing stakeholder trust and support.

Similarly, resource dependence theory suggests that firms rely on external resources for sustained growth and competitiveness (Pfeffer & Salancik, 1978). High ESG performance can serve as a positive signal that facilitates the acquisition of critical resources and supports stable operations (DiMaggio & Powell, 1983). In addition, prior literature suggests that robust governance structures mitigate agency problems and improve risk oversight, thereby reducing default risk (Gompers et al., 2003). Environmental and social performance may also act as proxies for proactive risk management and stakeholder engagement, contributing to firms’ long-term resilience (Eccles et al., 2014; Lins et al., 2017).

In the Chinese context, institutional theory offers further insight. Firms may pursue ESG initiatives not only for their intrinsic strategic benefits but also to comply with evolving regulatory expectations and to strengthen legitimacy among stakeholders (Marquis & Qian, 2014). Moreover, behavioral factors—such as managerial overconfidence and short-termism—can influence how ESG investments are perceived and implemented, ultimately shaping their financial implications (Ben-David et al., 2013).

Overall, the literature suggests that strong ESG performance reflects firms’ commitment to fulfilling social and environmental responsibilities, which can enhance stakeholder support, improve access to resources, and thereby reduce bankruptcy risk.

While the theoretical foundations suggest that ESG engagement can lower bankruptcy risk, the specific mechanisms through which this effect occurs remain less explored. To better understand how ESG performance contributes to financial stability, we examine four potential channels identified in the literature: (1) improving information transparency, (2) alleviating financing constraints, (3) enhancing operating performance, and (4) reducing leverage ratios. The following subsections review relevant studies and develop hypotheses related to these mechanisms.

2.1. ESG Performance, Information Transparency, and Bankruptcy Risk

Information transparency plays a crucial role in resource allocation and market efficiency, thereby influencing bankruptcy risk (Naqvi et al., 2021). Greater transparency reduces information asymmetry, allowing external stakeholders to better assess a firm’s financial condition. This transparency enhances stakeholder confidence, facilitates external financing, and lowers bankruptcy risk (Reber et al., 2022; H. Li et al., 2022). Conversely, information asymmetry can weaken monitoring effectiveness (Qian et al., 2019), enabling managerial opportunism and increasing agency costs and bankruptcy risk.

As a reflection of corporate social responsibility, ESG performance is inherently linked to transparency. Firms with stronger ESG performance typically disclose more ESG-related information, providing stakeholders with a broader view of their operations (Drempetic et al., 2020). Such disclosure attracts analysts’ attention, further improving information dissemination and monitoring (Qian et al., 2019). Analysts’ evaluations can reduce information asymmetry and improve governance quality, ultimately mitigating bankruptcy risk.

2.2. ESG Performance, Financing Constraints, and Bankruptcy Risk

Financing constraints are a major challenge for many Chinese firms (Poncet et al., 2010). Restricted access to financing can hinder firm operations and exacerbate financial vulnerability, particularly under adverse market conditions (Zhang et al., 2020). Higher financing costs may also pressure firms into pursuing riskier investments, increasing their bankruptcy risk.

ESG engagement can help alleviate financing constraints by improving firms’ reputations and signaling compliance with government priorities related to sustainability (Zeng & Crowther, 2019). Firms with superior ESG performance are more likely to obtain government subsidies, bank loans, and favorable credit terms. Moreover, high-quality ESG disclosure decreases information asymmetry, which lowers the perceived risk premium and financing costs (Qian et al., 2019; Naqvi et al., 2021). Consequently, better ESG performance may reduce financing constraints and, in turn, decrease bankruptcy risk.

2.3. ESG Performance, Operating Performance, and Bankruptcy Risk

Operating performance is a key determinant of a firm’s financial health and bankruptcy likelihood (Altman et al., 2017). Strong operating performance enhances profitability, efficiency, and debt-servicing capacity. It also supports innovation and investment, helping firms develop competitive advantages (Hsueh & Tu, 2004). Moreover, firms with robust performance typically enjoy stronger reputations, facilitating access to financing and reducing bankruptcy risk (H. Li et al., 2022).

Firms with good ESG performance often exhibit more effective governance structures, leading to higher operational efficiency and superior financial outcomes (H. Li et al., 2022). According to stakeholder theory, ESG engagement creates value for both internal and external stakeholders, generating stronger support networks that reduce operational and financial risks (Singhal & Zhu, 2013). Thus, good ESG performance promotes better operating performance, which enhances competitiveness and reduces bankruptcy risk.

2.4. ESG Performance, Leverage Ratio, and Bankruptcy Risk

Firms with good ESG performance tend to manage environmental and social risks more effectively and maintain higher investor confidence (Wu et al., 2024). Prior research shows that poor ESG performance is often connected with higher equity costs and risk premiums (Shen et al., 2023; Chava, 2014), while firms with strong ESG profiles experience lower financing costs and improved capital structures (Asimakopoulos et al., 2023).

Evidence from Chinese listed firms supports this pattern: better ESG performance is associated with lower costs of debt and equity (Lai & Zhang, 2022; Chen et al., 2023). Firms with strong ESG records attract greater investor support (Long & Ouyang, 2022) and achieve lower leverage ratios by substituting equity for debt financing (Xu et al., 2015). ESG certification can also reduce over-indebtedness among Chinese firms (Lai & Zhang, 2022). Therefore, firms with better ESG performance are likely to maintain healthier capital structures, thereby reducing bankruptcy risk.

Building on the discussion of how overall ESG performance influences firms’ capital structure and bankruptcy risk, it is vital to recognize that ESG is a multidimensional construct comprising environmental, social, and governance components. Each dimension may affect financial stability through different channels and with varying degrees of influence. To obtain a more insightful understanding of the ESG–bankruptcy nexus, the following section examines how the individual pillars—environmental, social, and governance—contribute separately to reducing bankruptcy risk within China’s institutional and regulatory context.

2.5. The Three Pillars of ESG Performance and Bankruptcy Risk

From a theoretical perspective, each ESG dimension influences bankruptcy risk through distinct yet interconnected mechanisms. The environmental (E) dimension reflects a firm’s commitment to sustainability, resource efficiency, and regulatory compliance. Firms with strong environmental practices are better positioned to manage environmental liabilities, avoid regulatory penalties, and adapt to China’s evolving green policy agenda—particularly the national goals of carbon peaking and carbon neutrality. Within China’s institutional framework, where environmental regulation and green finance are increasingly prioritized, environmentally responsible firms gain preferential access to financing and policy support, thereby reducing their exposure to financial distress.

The social (S) dimension captures a firm’s relationships with employees, customers, and other stakeholders. From a behavioral perspective, socially responsible firms foster stronger stakeholder trust, employee loyalty, and customer satisfaction, which can stabilize revenue streams and reduce operational volatility. In the Chinese context, where social harmony and collective well-being are deeply embedded in cultural and political values, strong social engagement enhances firms’ legitimacy and reputation, helping them secure both market and governmental support during periods of financial stress.

The governance (G) dimension encompasses board structure, management oversight, and shareholder rights, serving as a core mechanism linking ESG performance to financial stability. Effective governance mitigates agency problems, enhances managerial accountability, and promotes transparent information disclosure—factors that directly improve decision-making efficiency and lower bankruptcy risk. Within China’s regulatory environment, governance quality also signals institutional alignment and compliance with state directives, reinforcing investor confidence and access to capital.

Taken together, these mechanisms illustrate that while all three ESG pillars contribute to reducing bankruptcy risk, their relative importance may differ in China’s institutional context. The governance dimension likely exerts the strongest influence by directly shaping corporate control and accountability, while environmental and social practices contribute indirectly through legitimacy building and institutional alignment.

2.6. Hypotheses Development

Synthesizing the above discussion, we propose the following hypotheses:

Hypothesis 1:

ESG performance is negatively associated with bankruptcy risk; that is, higher ESG performance corresponds to lower bankruptcy risk.

Hypothesis 2:

ESG performance mitigates bankruptcy risk through four primary mechanisms—enhancing information transparency, alleviating financing constraints, improving operating performance, and lowering leverage ratios.

Hypothesis 3:

The three pillars of ESG performance—environmental, social, and governance—exert differential effects on bankruptcy risk. Specifically, stronger performance in each dimension is expected to reduce bankruptcy risk, with the governance pillar having the most pronounced impact.

3. Data and Research Design

3.1. Sample Selection and Data

The research sample comprises firms listed on the Shanghai and Shenzhen Stock Exchanges between 2009 and 2022. The analysis starts in 2009, the first year the Chinese government required companies to disclose ESG information, which also helps to reduce any lingering effects of the 2008 global financial crisis. Financial and real estate sectors are excluded, along with firms lacking complete data, resulting in 23,803 firm-year observations. All financial and accounting variables, apart from ESG scores, are collected from the China Stock Market and Accounting Research (CSMAR) database, one of the most comprehensive data sources for Chinese listed companies. ESG scores are obtained from the Sino-Securities Index ESG rating database, a commonly adopted dataset in previous ESG-related literature (Feng et al., 2022; Lian et al., 2023). To limit the influence of outliers, continuous variables are winsorized at the top and bottom 1 percent.

3.2. Variables and Model Specifications

Consistent with prior research (Altman, 1968; Agarwal & Taffler, 2007; Houston et al., 2010; Badayi et al., 2021; Ji et al., 2022), we employ the Altman Z-score to measure firms’ bankruptcy risk. Compared with using realized bankruptcy events, a key advantage of the Z-score is that it reflects both cross-sectional and time-varying variations in the likelihood of bankruptcy. We use the normalized Altman Z-score, which is inversely related to bankruptcy risk—higher Z-scores indicate lower probabilities of financial distress. Altman et al. (2017) demonstrate that the Z-score remains a robust predictor of bankruptcy across most industries in an international setting.

Data for our main independent variable, firms’ ESG ratings, are obtained from the Sino-Securities Index ESG rating database. These ratings are derived from corporate disclosures and publicly available information covering three dimensions: environmental, social, and corporate governance. The original nine rating categories are converted into numerical scores ranging from one to nine, with one representing the lowest and nine the highest ESG performance.

To test our main hypotheses, we estimate the baseline regression model below:

$${Risk}_{i,t+1}={\alpha }_0+{\alpha }_1{ESG}_{i,t}+\sum {Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(1)

where i and t indicate firms and years, respectively. To lessen the possible endogeneity caused by reverse causality, we use one-year lagged independent variables. ${Risk}_{i,t+1}$ proxies for bankruptcy risk and is measured by Altman’s Z-score, computed as follows (Altman, 1968):

$$Zscore=1.2X_1+1.4X_2+3.3X_3+0.6X_4+0.999X_5$$

(2)

where $X_1$ is the Working Capital/Total Assets ratio, $X_2$ is the Retained Earnings/Total Assets ratio, $X_3$ is the EBIT/Total Assets ratio, $X_4$ is the Market Value of Equity/Total Liabilities ratio, and $X_5$ is the Total Sales/Total Assets ratio. A higher Z-score indicates lower bankruptcy risk.

${ESG}_{i,t}$ is measured by the ESG rating provided by the Sino-Securities Index ESG rating database. This measure has been widely used in ESG studies as a proxy for comprehensive ESG performance for Chinese firms (Y. Lin et al., 2021; H. Li et al., 2022). $\sum {Control}_{i,t}$ is a vector of control variables that have been shown to affect the bankruptcy risk in prior literature. The variables include firm size (Size), ROE, firm age (Age), sales growth rate (Growth), ownership structure (Ownership), duality, independent director ratio (Independence), independent auditing opinion (Audit), and GDP growth rate (GDP growth). ${\phi }_i$ and ${\mu }_t$ are used to control the firm and year fixed effects, respectively; and ${\epsilon }_{i,t}$ is the regression error term. The variable definitions are reported in Appendix A.

Table 1. Descriptive statistics.

Variable	N	Mean	Std. Dev.	P25	Median	P75	Min	Max
Risk_Zscore	23,803	4.533	5.255	1.753	2.891	5.058	0.004	34.338
ESG	23,803	4.07	1.082	3	4	5	1	8
Size	23,803	22.366	1.276	21.461	22.199	23.100	19.952	26.367
ROE	23,803	6.075	13.437	2.560	6.921	12.052	−63.713	36.254
Age	23,803	11.196	6.935	5	10	16	2	32
Growth	23,803	0.212	0.504	−0.022	0.112	0.291	−0.550	3.322
Ownershare	23,803	0.339	0.164	0.211	0.322	0.451	0.038	0.750
Duality	23,803	0.245	0.43	0	0	0	0	1
Independence	23,803	37.420	5.291	33.330	33.330	42.860	33.330	57.140
Audit	23,803	0.972	0.166	1	1	1	0	1
GDP_Growth	23,803	7.661	2.970	6.600	7.800	8.900	0.200	14.800

This table reports the summary statistics for the key variables in our sample. N denotes the number of observations. The sample period is from 2009 to 2022. All variables are defined in Appendix A.

reports the descriptive statistics. The mean Z-score is 4.533, with a standard deviation of 5.255, indicating a relatively high level of bankruptcy risk and substantial variation across firms. The mean and median ESG ratings are 4.07 and 4, respectively, suggesting that the average ESG performance of Chinese firms remains relatively low. The average firm age is 11.2 years. The mean return on equity (ROE) is 6.07%, ranging from −63.71% to 36.25%, reflecting considerable volatility in firms’ profitability. The values of other firm-level variables are broadly consistent with those documented in prior ESG studies on Chinese firms.

4. Empirical Results

In this section, we present the empirical analysis exploring how ESG performance relates to the likelihood of corporate bankruptcy. Following the discussion of our baseline regression outcomes, several robustness checks are conducted to verify the reliability of the main findings.

4.1. Baseline Regression Results

To investigate the connection between ESG performance and bankruptcy risk, we estimate the baseline model described in Equation (1). The regression results are summarized in

Table 2. Baseline Regressions.

	(1)	(2)
	Risk_Zscore	Risk_Zscore
ESG	0.086 **	0.134 ***
	(2.527)	(4.020)
Size		−1.599 ***
		(−13.985)
ROE		0.036 ***
		(11.539)
Age		−0.014
		(−0.669)
Growth		−0.037
		(−0.591)
Ownershare		−0.068
		(−0.121)
Duality		0.011
		−0.090
Independence		0.016 *
		(1.800)
Audit		0.245
		(0.989)
GDP_Growth		−0.080 ***
		(−2.698)
Constant	5.294 ***	39.368 ***
	(27.150)	(15.857)
Firm fixed effects	Yes	Yes
Year fixed effects	Yes	Yes
N	18,423	18,423
Adj R2	0.097	0.143

This table presents the estimation outcomes from the baseline regression model. The dependent variable used in the analysis is Risk_Zscore. Column (1) shows the regression estimates excluding control variables, while Column (2) includes all control variables. Definitions of all variables can be found in the Appendix A. Standard errors are adjusted for heteroskedasticity using the Huber–White robust method and are clustered at the firm level. Corresponding t-statistics are shown in parentheses. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively.

. Column (1) displays the estimates without any control variables, whereas Column (2) incorporates the complete set of controls. The coefficients on ESG performance are 0.086 and 0.134, statistically significant at the 5% and 1% levels, respectively, revealing a negative association between ESG performance and bankruptcy risk. The economic effect is also meaningful: a one–standard-deviation improvement in ESG performance translates into a 9.3% decrease in bankruptcy risk in Column (1) and a 14.5% decrease in Column (2). Overall, these results lend support to Hypothesis 1, indicating that firms exhibiting higher ESG standards are less prone to financial distress.

The coefficients on the control variables are largely consistent with prior studies. Firm size exhibits a negative association with the Z-score, implying that larger firms are more exposed to bankruptcy risk—consistent with the current economic context in China. For example, Evergrande Group, the country’s largest property developer, filed for bankruptcy in 2023, marking one of the most significant financial events in China that year. Because much of the debt held by large Chinese firms consists of loans from state-owned banks, their elevated bankruptcy risk has become a systemic concern in China’s economy. Overall, the baseline regression results align with our first hypothesis, reinforcing the conclusion that better ESG performance is associated with lower bankruptcy risk.

4.2. Robustness Tests

In this section, we implement several robustness tests to corroborate the causal relation from ESG performance to bankruptcy risk found in baseline regression results. Given the endogeneity concern which may lead to the spurious association between ESG performance and bankruptcy risk, we use the Instrumental variables (IV) approach with three instrumental variables, respectively, following prior CSR studies. The regression results by using the three instrumental variables all support the causal relation from ESG performance to bankruptcy risk. To take one step further in the following subsection, we use the propensity score matching (PSM) approach to tackle the endogeneity concern and the results remain consistent with our finding that ESG performance is causally related to bankruptcy risk.

4.2.1. Addressing Potential Endogeneity Through Instrumental Variables

The first instrumental variable (IV), termed Confucian, measures the count of Confucian temples located within a 100-km radius of each firm’s registered address. This proxy has appeared in previous ESG literature, as Confucian ideology emphasizes ethical responsibility and moral discipline—traits that can foster socially responsible business practices (Liu et al., 2023). Earlier findings also indicate that Confucian values and their long-term orientation influence corporate decision-making in ethical and social domains (F. Li et al., 2019). Consequently, the number of temples reflects the regional presence of Confucian culture, which contributes to shaping managerial ethics, governance attitudes, and ESG participation.

The relevance of this instrument is reinforced by existing empirical studies that link Confucian traditions with firms’ sustainable and socially oriented behavior (Liu et al., 2023; F. Li et al., 2019). In addition, temple density is unlikely to affect modern firm outcomes through any channel other than local cultural norms, thus fulfilling the exclusion restriction. Information on Confucian temples is retrieved from the Confucian Culture Database compiled by CNRDS. Results of the two-stage least squares (2SLS) estimation are shown in Columns (1) and (2) of

Table 3. Robustness test using instrumental variables.

	(1)		(2)		(3)
	First Stage	Second Stage	First Stage	Second Stage	First Stage	Second Stage
	ESG	Risk_Zscore	ESG	Risk_Zscore	ESG	Risk_Zscore
Confucian	0.009 ***
	(5.394)
FQ			0.115 ***
			(7.148)
L.ESG					0.254 ***
					(22.421)
ESG		2.441 ***		2.030 ***		0.475 ***
		(2.643)		(3.254)		(3.346)
Size	0.194 ***	−2.119 ***	0.159 ***	−1.727 ***	0.180 ***	−1.534 ***
	(29.270)	(−11.652)	(5.193)	(−10.092)	(6.801)	(−11.876)
ROE	0.012 ***	0.063 ***	0.002	0.031 ***	0.002 ***	0.034 ***
	(17.635)	(5.405)	(1.487)	(7.785)	(2.599)	(9.315)
Age	−0.018 ***	0.0317 *	0.097 **	−0.132	0.003	−0.028
	(−14.020)	(1.827)	(2.084)	(−0.642)	(0.068)	(−0.160)
Growth	−0.105 ***	−0.118	−0.068 ***	0.110	−0.048 ***	−0.075
	(−6.874)	(−0.963)	(−3.837)	−1.509	(−2.823)	(−1.154)
Ownershare	0.017	1.032 ***	0.091	−0.604	0.168	−0.271
	(0.352)	(4.275)	(0.654)	(−0.952)	(1.392)	(−0.458)
Duality	−0.072 ***	0.213 *	−0.067 *	0.071	−0.049 *	−0.022
	(−3.869)	(1.889)	(−1.849)	−0.501	(−1.680)	(−0.155)
Independence	0.016 ***	−0.010	0.015 ***	−0.015	0.016 ***	0.009
	(11.114)	(−0.623)	(5.118)	(−1.106)	(6.736)	(0.958)
Audit	0.671 ***	−1.528 **	0.211 ***	0.209	0.195 ***	0.211
	(12.367)	(−2.257)	(2.783)	(0.629)	(3.035)	(0.925)
GDP_Growth	0.005	−0.177 ***	0.015 *	−0.082 **	0.009	−0.048
	(0.857)	(−6.558)	(1.808)	(−2.134)	(1.284)	(−1.574)
Firm fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
N	18,423	18,423	14,025	14,025	14,025	14,025
Kleibergen–Paap rk Wald F-statistic		206.82		51.09		502.72

This table presents the outcomes of robustness analyses based on two-stage instrumental variable (IV) regressions. Three IVs—Confucian, FQ, and L.ESG—are employed individually in separate models. In the first-stage estimations, ESG performance serves as the dependent variable, with each instrument incorporated into the baseline specification. Columns (2), (4), and (6) display the corresponding results from the second-stage regressions. Definitions of all variables are provided in the Appendix A. Standard errors are adjusted for heteroskedasticity using the Huber–White robust method and are clustered at the firm level. t-statistics are shown in parentheses. The symbols ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

. A strong first-stage F-statistic verifies the instrument’s strength, while the second-stage coefficient of 2.442 remains significant, suggesting consistent results with the baseline estimation.

For the second instrument, the study uses the number of ESG investment funds holding a company’s shares, denoted as FQ, in line with Fang and Hu (2023) and Mao and Wang (2023). This variable is chosen for two main reasons. First, ESG funds, as institutional investors, may shape firm policies through engagement and capital allocation—commonly referred to as “foot voting”—thereby motivating improvements in ESG performance (Dimson et al., 2015; Dyck et al., 2019). Second, because these funds independently determine their portfolios and do not intervene in daily operations, the number of ESG funds investing in a firm is unlikely to directly influence bankruptcy likelihood, satisfying the exclusion condition. The 2SLS results for this specification are presented in Columns (3) and (4) of Table 3. The first stage shows a positive and significant link between FQ and ESG performance, with an F-statistic of 51.09, implying that the instrument is valid. The second stage yields a coefficient of 2.03 for ESG performance, indicating a positive relationship between ESG engagement and the Z-score, consistent with prior findings that stronger ESG performance lowers bankruptcy risk.

The third instrumental variable utilizes the one-period lag of ESG performance, following the methodology of earlier research (Angrist et al., 1996; Buch et al., 2013). The 2SLS estimations for this model are displayed in Columns (5) and (6) of Table 3. The first-stage F-statistic is significant at the 1% threshold, confirming that lagged ESG performance is a powerful instrument. The second-stage coefficient of 0.475, also significant at the 1% level, further substantiates a causal relationship between ESG engagement and reduced bankruptcy probability.

4.2.2. Propensity Score Matching (PSM) Approach

To address possible endogeneity or omitted-variable bias that could influence the estimated impact of ESG performance, this study employs the propensity score matching (PSM) approach, following the procedure outlined by Shipman et al. (2017). The sample is divided into two groups according to firms’ ESG ratings: companies with scores ranging from 1 to 4 form the control group, whereas those with scores between 5 and 9 make up the treatment group.

Three different matching strategies are applied to create balanced comparison samples: (1) one-to-one nearest-neighbor matching, (2) one-to-five nearest-neighbor matching, and (3) caliper matching using a bandwidth of 0.01. The regression outcomes for these matched samples are shown in Columns (1)–(3) of

Table 4. Robustness tests based on PSM samples.

	(1)	(2)	(3)
	Risk_Zscore	Risk_Zscore	Risk_Zscore
ESG	0.280 **	0.236 ***	0.232 ***
	(2.470)	(3.127)	(3.045)
Size	−1.434 ***	−1.434 ***	−1.390 ***
	(−8.032)	(−11.290)	(−11.409)
ROE	0.047 ***	0.038 ***	0.035 ***
	(7.581)	(9.632)	(9.784)
Age	0.033	0.021	−0.001
	(0.747)	(0.647)	(−0.029)
Growth	0.019	−0.074	−0.024
	(0.235)	(−1.326)	(−0.449)
Ownershare	−1.400 *	−0.235	−0.430
	(−1.684)	(−0.417)	(−0.746)
Duality	−0.139	−0.103	−0.084
	(−0.662)	(−0.720)	(−0.648)
Independence	0.007	0.009	0.014
	(0.567)	(0.955)	(1.446)
Audit	−0.352	0.509	0.681 **
	(−1.048)	(1.550)	(2.349)
GDP_Growth	−0.019	−0.036	−0.051
	(−0.444)	(−1.122)	(−1.566)
Constant	36.841 ***	35.632 ***	34.690 ***
	(9.468)	(13.126)	(13.216)
Firm fixed effects	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes
N	6604	12,728	14,289
Adj R2	0.126	0.135	0.134

This table displays the regression outcomes derived from the three propensity score matching (PSM) samples. Specifically, the analyses employ three matching techniques: 1:1 nearest-neighbor, 1:5 nearest-neighbor, and caliper matching with a bandwidth of 0.01, each producing a distinct matched dataset. The t-statistics are shown in parentheses beneath the coefficient estimates. Definitions of all variables are provided in the Appendix A. The symbols ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

. The coefficients of ESG performance are estimated at 0.280, 0.236, and 0.232, all statistically significant at either the 5% or 1% level. These results offer strong empirical support for a causal link between higher ESG performance and a lower likelihood of bankruptcy.

4.2.3. Alternative Measures for Bankruptcy Risk and ESG Performance

To further test the robustness of our results, we re-estimate the baseline regression models by employing alternative proxies for both bankruptcy risk and ESG performance.

Consistent with He et al. (2022), we use Ohlson’s Oscore (Ohlson, 1980) as an alternative indicator of bankruptcy probability. The Oscore is computed using the following specification:

$$\begin{array}{cc}\hfill Oscore=& -1.32-0.407Size+6.03Lev-1.43Wc+0.0757Cl-2.37ROA\hfill \\ & \hspace{1em}-1.83Sale+0.285Pat-1.72Lia-0.521∆ni/∆\left|ni\right|\hfill \end{array}$$

(3)

In this specification, Size denotes firm size; Lev represents the leverage ratio; Wc refers to working capital scaled by total assets; and ROA is calculated as net income divided by total assets. Sale captures firm sales. Pat is a dummy variable that equals 1 if the company recorded net losses in the previous two years and 0 otherwise. Lia is an indicator variable that takes the value of 1 when total liabilities exceed total assets, and 0 otherwise. Ni represents net income.

A higher O-score corresponds to a greater likelihood of financial distress (Ohlson, 1980). Replacing the Z-score with the O-score, we re-estimate the baseline regression. As shown in Column (1) of

Table 5. Robustness test using alternative bankruptcy risk and ESG variables.

	(1)	(2)
	Risk_Oscore	Risk_Zscore
ESG	−0.093 ***	0.030 ***
	(−4.687)	(3.458)
Size	0.374 ***	−1.546 ***
	(5.657)	(−9.998)
ROE	−0.026 ***	0.038 ***
	(−6.843)	(8.167)
Age	−0.030 **	0.004
	(−2.560)	(0.145)
Growth	−0.077	−0.035
	(−1.176)	(−0.365)
Ownershare	−0.641	0.258
	(−1.413)	(0.442)
Duality	−0.089	−0.143
	(−1.125)	(−0.746)
Independence	−0.007	0.010
	(−1.334)	(1.057)
Audit	−0.961 ***	0.649 **
	(−4.262)	(2.32)
GDP_Growth	0.070 ***	−0.043
	(4.206)	(−1.353)
Constant	−14.993 ***	37.450 ***
	(−10.064)	(10.996)
Firm fixed effects	Yes	Yes
Year fixed effects	Yes	Yes
N	22,169	8059
Adj R2	0.04	0.132

This table presents the regression outcomes obtained by substituting alternative measures for the dependent variable (bankruptcy risk) and the independent variable (ESG performance). Detailed definitions of all variables are provided in the Appendix A. Standard errors are adjusted for heteroskedasticity using the Huber–White robust estimation method and are clustered at the firm level. The t-statistics appear in parentheses beneath the coefficients. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively.

, the coefficient on ESG performance is −0.093 and is statistically significant at the 1% level, reinforcing the negative association between ESG performance and bankruptcy risk.

To account for potential measurement inconsistencies across ESG metrics, we also utilize Bloomberg’s ESG scores as an alternative measure of ESG performance, where higher values denote stronger ESG engagement. The regression results, reported in Column (2) of Table 5, reveal a coefficient of 0.030, significant at the 1% level, aligning closely with the results based on the Sino-Securities Index ESG ratings used in the baseline analysis.

Taken together, these robustness checks lend further empirical support to Hypothesis 1, indicating that firms with superior ESG performance are less prone to bankruptcy. The consistency across multiple measures also suggests that endogeneity or measurement bias is unlikely to explain the observed relationship.

5. Extended Analysis

5.1. Underlying Mechanisms

Having confirmed a negative association between ESG performance and bankruptcy risk, we next explore the potential mechanisms through which ESG performance may mitigate firms’ bankruptcy risk. Our analysis follows the four potential channels discussed in the hypothesis development section.

5.1.1. Improving Information Transparency

As discussed earlier, ESG performance can reduce bankruptcy risk by enhancing a firm’s information transparency. Strong ESG practices help mitigate information asymmetry by providing more credible and comprehensive information to investors (Drempetic et al., 2020) and by attracting greater attention from financial analysts. Improved transparency, in turn, strengthens stakeholders’ trust and engagement with the firm’s social responsibility initiatives. Prior studies also find that superior ESG performance enhances the quality of corporate disclosures and reduces agency costs (Lian et al., 2023; H. Li et al., 2022). Therefore, information transparency represents an important channel through which ESG performance alleviates bankruptcy risk, as higher transparency is associated with lower financial distress (Halov & Heider, 2011; L. Li & Faff, 2019).

To evaluate this underlying mechanism, we adopt two indicators of information transparency commonly used in prior studies. The first measure is the information quality score (ranging from 1 to 4) released by the Shanghai and Shenzhen Stock Exchanges, with higher values representing greater disclosure quality and transparency (Ji et al., 2022). The second measure employs analyst attention as an alternative proxy, following Naqvi et al. (2021). A higher level of analyst coverage reflects stronger external oversight and thus greater transparency in information dissemination (Naqvi et al., 2021; Qian et al., 2019).

We test this hypothesis using a mediation framework, and the corresponding model equations are specified as follows:

$${Risk}_{i,t+1}={\alpha }_0+{\alpha }_1{ESG}_{i,t}+{\alpha }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(4)

$${Information}_{i,t+1}={\beta }_0+{\beta }_1{ESG}_{i,t}+{\beta }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(5)

$${Risk}_{i,t+1}={\gamma }_0+{\gamma }_1{ESG}_{i,t}+{\gamma }_2{Information}_{i,t+1}+{\gamma }_3{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(6)

The mediation analysis is carried out in three stages.

In the first stage, Equation (4) is estimated to capture the total influence of ESG performance on bankruptcy risk. The coefficient ${\alpha }_1$ represents this overall effect.

In the second stage, Equation (5) is estimated with the information transparency variable serving as the mediator. Here, the coefficient ${\beta }_1$ reflects how ESG performance affects the mediating factor. A statistically significant ${\beta }_1$ indicates that ESG performance helps explain variations in information transparency.

In the third stage, Equation (6) is estimated to assess the mediating effect. The coefficient ${\gamma }_2$ measures the relationship between the mediator and bankruptcy risk after controlling for ESG performance. If both ${\gamma }_1$ and ${\gamma }_2$ are significant and their signs align with theoretical expectations, and ${\gamma }_1$ is smaller in magnitude than ${\alpha }_1$, this suggests a partial mediation effect. Conversely, if ${\gamma }_1$ becomes insignificant while ${\gamma }_2$ remains significant, it implies a full mediation effect.

Table 6. Analysis on ESG performance’s impact on bankruptcy risk through information transparency.

	(1)	(2)	(3)	(4)	(5)	(6)
	Risk_Zscore	Quality	Risk_Zscore	Risk_Zscore	Analyst	Risk_Zscore
ESG	0.128 ***	0.026 ***	0.123 ***	0.148 ***	0.398 ***	0.126 ***
	(2.692)	(3.157)	(2.592)	(3.549)	(2.874)	(3.038)
Quality			0.183 **
			(2.189)
Analyst						0.054 ***
						(8.152)
Size	−1.425 ***	−0.066 ***	−1.413 ***	−1.524 ***	1.999 ***	−1.632 ***
	(−8.307)	(−2.867)	(−8.250)	(−11.089)	(4.852)	(−11.980)
ROE	0.0319 ***	0.007 ***	0.031 ***	0.048 ***	0.226 ***	0.036 ***
	(7.3)	(8.59)	(7.003)	(9.756)	(12.217)	(8.186)
Age	−0.029	0.021 ***	−0.033	−0.010	−0.557 ***	0.020
	(−0.557)	(2.871)	(−0.630)	(−0.286)	(−4.743)	(0.568)
Growth	−0.074	0.020 *	−0.078	−0.055	0.424 **	−0.078
	(−1.193)	(1.67)	(−1.251)	(−0.914)	(2.11)	(−1.289)
Ownershare	−0.571	0.118	−0.592	−0.323	−1.494	−0.242
	(−0.724)	(1.07)	(−0.750)	(−0.549)	(−0.794)	(−0.413)
Duality	−0.164	0.003	−0.165	−0.164	0.586	−0.196
	(−1.023)	(0.127)	(−1.028)	(−0.941)	(1.391)	(−1.136)
Independence	0.015	0.004 *	0.014	0.011	0.050	0.009
	(1.198)	(1.731)	(1.146)	(1.062)	(1.529)	(0.830)
Audit	0.748 *	0.457 ***	0.664 *	0.228	−0.308	0.245
	(1.915)	(6.813)	(1.727)	(0.653)	(−0.342)	(0.700)
GDP_Growth	−0.040	−0.004	−0.039	−0.127 ***	0.119	−0.134 ***
	(−0.789)	(−0.474)	(−0.776)	(−3.068)	(1.085)	(−3.298)
Constant	35.324 ***	3.543 ***	34.676 ***	38.717 ***	−33.162 ***	40.515 ***
	(9.894)	(6.993)	(9.689)	(12.741)	(−3.718)	(13.459)
Firm fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
N	9728	9728	9728	10,024	10,024	10,024
Adj R2	0.159	0.046	0.159	0.136	0.126	0.154

This table presents the results of the analysis examining how ESG performance influences bankruptcy risk through the channel of information transparency. Detailed definitions of all variables are provided in the Appendix A. Standard errors are adjusted for heteroskedasticity using the Huber–White robust method and are clustered at the firm level. The t-statistics are shown in parentheses below the coefficients. The symbols ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

reports the estimation results. Columns (1), (2) and (3) present the results of Equations (4), (5) and (6), respectively, using information quality score as the proxy for information transparency. Columns (4) to (6) replicate the analysis with analyst coverage as the proxy for information transparency. The estimation results in Columns (2) and (5) show that coefficients of ESG performance are 0.026 and 0.398, respectively, both significant at 1% level, consistent with our conjecture and previous literature (L. Li & Faff, 2019). Overall, the results suggest that information transparency has a partial mediating effect on the relation between ESG performance and bankruptcy risk. Thus, the results support our conjecture that ESG performance helps reduce firms’ bankruptcy risk through improving information transparency.

5.1.2. Alleviating Financing Constraints

We further posit that ESG performance may reduce a firm’s bankruptcy risk by alleviating financing constraints. Strong ESG performance improves a firm’s reputation as a socially responsible enterprise, aligning it with governmental objectives. In China, such firms are often more likely to receive government subsidies or preferential access to bank credit, particularly from state-owned banks, thereby easing financial constraints. The relaxation of financing constraints can in turn promote firm growth and reduce bankruptcy risk (Zhang et al., 2020; Yu, 2023).

To test this mechanism, we construct two commonly used proxies for financing constraints: the KZ index and the SA index (J. J. Li & Han, 2019; Mao & Wang, 2023). The KZ index, developed by Kaplan and Zingales (1997), measures the extent of financial constraints among firms with low dividend payouts. It assigns positive weights to Tobin’s Q and leverage, and negative weights to operating cash flow, cash balances, and dividends—higher KZ values indicate tighter financial constraints (Kaplan & Zingales, 1997; Huang et al., 2022). The SA index, proposed by Hadlock and Pierce (2010), is based on firm size and age, with higher values indicating greater financial constraints (Ju et al., 2013). All variables used to construct these indices are obtained from the CSMAR database.

We use a mediation model to examine this hypothesis. The equations used in the model are presented below:

$${Risk}_{i,t+1}={\alpha }_0+{\alpha }_1{ESG}_{i,t}+{\alpha }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

$${Financing}_{i,t+1}={\beta }_0+{\beta }_1{ESG}_{i,t}+{\beta }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(7)

$${Risk}_{i,t+1}={\gamma }_0+{\gamma }_1{ESG}_{i,t}+{\gamma }_2{Financing}_{i,t+1}+{\gamma }_3{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(8)

The analytical procedure follows the same approach described in Section 5.1.1, and the results are displayed in

Table 7. Analysis on ESG performance’s impact on bankruptcy risk through alleviating financing constraints.

	(1)	(2)	(3)	(4)	(5)	(6)
	Risk_Zscore	SA	Risk_Zscore	Risk_Zscore	KZ	Risk_Zscore
ESG	0.134 ***	0.007 ***	0.070 **	0.134 ***	0.077 ***	0.1183 ***
	(4.020)	(7.800)	(2.087)	(4.020)	(−4.287)	(3.589)
SA			9.857 ***
			(9.250)
KZ						0.054 ***
						(−6.973)
Size	−1.599 ***	0.008 *	−1.675 ***	−1.599 ***	0.482 ***	−1.498 ***
	(−13.985)	(1.957)	(−14.940)	(−13.985)	−8.523	(−13.005)
ROE	0.036 ***	−0.001 *	0.037 ***	0.036 ***	−0.029 ***	0.030 ***
	(11.539)	(−1.652)	(11.697)	(11.539)	(−16.057)	(10.214)
Age	−0.014	−0.038 ***	0.361 ***	−0.014	0.028 ***	−0.008
	(−0.669)	(−69.359)	(7.999)	(−0.669)	−2.849	(−0.401)
Growth	−0.037	−0.002 *	−0.015	−0.037	−0.049	−0.047
	(−0.591)	(−1.672)	(−0.251)	(−0.591)	(−1.618)	(−0.762)
Ownershare	−0.068	0.039 ***	−0.452	−0.068	−1.437 ***	−0.368
	(−0.121)	(2.708)	(−0.818)	(−0.121)	(−5.411)	(−0.684)
Duality	0.011	0.002	−0.009	0.011	−0.052	0.000
	(0.090)	(1.004)	(−0.076)	(0.090)	(−0.984)	(0.001)
Independence	0.016 *	0.000	0.016 *	0.016 *	−0.007	0.014 *
	(1.800)	(0.120)	(1.761)	(1.800)	(−1.644)	(1.662)
Audit	0.245	−0.003	0.276	0.245	−0.563 ***	0.1272
	(0.989)	(−0.793)	(1.125)	(0.989)	(−5.511)	(0.521)
GDP_Growth	−0.080 ***	0.001	−0.085 ***	−0.080 ***	0.054 ***	−0.068 **
	(−2.698)	(0.756)	(−2.917)	(−2.698)	−4.084	(−2.352)
Constant	39.368 ***	−3.609 ***	74.937 ***	39.368 ***	−8.144 ***	37.662 ***
	(15.857)	(−43.018)	(14.348)	(15.857)	(−6.667)	(15.149)
Firm fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
N	18,423	18,423	18,423	18,423	18,423	18,423
Adj R2	0.143	0.864	0.167	0.143	0.205	0.152

This table summarizes the results of the analysis examining how ESG performance affects bankruptcy risk by easing firms’ financial constraints. Detailed definitions of all variables are provided in the Appendix A. Standard errors are adjusted for heteroskedasticity using the Huber–White robust estimator and are clustered at the firm level. The t-statistics appear in parentheses beneath the coefficient estimates. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively.

. Columns (1)–(3) present the estimation results of Equations (4), (7), and (8), respectively, using the SA index as the proxy for financing constraints. Columns (4)–(6) display the corresponding results when the KZ index is used as the alternative proxy.

The significant coefficients of ESG performance in Columns (2) and (5) indicate that stronger ESG performance helps alleviate firms’ financing constraints, consistent with prior studies (Xie & Lv, 2022; Fang & Hu, 2023; Luo et al., 2023; Mao & Wang, 2023). Furthermore, the results suggest that the reduction in financing constraints partially mediates the relationship between ESG performance and bankruptcy risk. Overall, these findings support our conjecture that ESG performance lowers bankruptcy risk by easing firms’ financing constraints.

5.1.3. Improving Operating Performance

Third, we hypothesize that ESG performance may reduce a firm’s bankruptcy risk via enhancing its operating performance. Firms with strong ESG performance often exhibit more effective corporate governance, which can help reduce operational costs and enhance efficiency (Wang et al., 2022; Xue et al., 2022). Moreover, effective ESG practices foster stronger relationships with key stakeholders, enabling firms to gain additional support and access to valuable resources (J. Lin et al., 2020). Enhanced operating performance and sound governance, in turn, contribute to a lower likelihood of bankruptcy (Darrat et al., 2016).

To test this mechanism, we use return on assets (ROA) and earnings per share (EPS) as proxies for operating performance. Consistent with the previous section, we employ a mediation model to examine this hypothesis. The equations used in the model are presented below:

$${Risk}_{i,t+1}={\alpha }_0+{\alpha }_1{ESG}_{i,t}+{\alpha }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

$${Performance}_{i,t+1}={\beta }_0+{\beta }_1{ESG}_{i,t}+{\beta }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(9)

$${Risk}_{i,t+1}={\gamma }_0+{\gamma }_1{ESG}_{i,t}+{\gamma }_2{Performance}_{i,t+1}+{\gamma }_3{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(10)

The analytical procedure follows the same approach described in Section 5.1.1, and the results are presented in

Table 8. Analysis on ESG performance’s impact on bankruptcy risk through improving operating performance.

	(1)	(2)	(3)	(4)	(5)	(6)
	Risk_Zscore	ROA	Risk_Zscore	Risk_Zscore	EPS	Risk_Zscore
ESG	0.134 ***	0.180 ***	0.108 ***	0.135 ***	0.024 ***	0.118 ***
	(4.020)	(3.229)	(3.327)	(3.788)	(3.107)	(3.345)
ROA			0.147 ***
			(19.672)
EPS						0.723 ***
						(9.037)
Size	−1.599 ***	−1.895 ***	−1.320 ***	−1.400 ***	−0.006	−1.396 ***
	(−13.985)	(−12.497)	(−12.025)	(−11.497)	(−0.157)	(−11.852)
ROE	0.036 ***	0.109 ***	0.020 ***	0.034 ***	0.011 ***	0.026 ***
	(11.539)	(14.097)	(7.862)	(9.476)	(5.708)	(8.090)
Age	−0.014	0.002	−0.014	0.001	0.023 ***	−0.016
	(−0.669)	(0.062)	(−0.708)	(0.035)	(3.762)	(−0.495)
Growth	−0.037	0.646 ***	−0.132 **	−0.022	0.064 ***	−0.068
	(−0.591)	(6.335)	(−2.219)	(−0.411)	(5.106)	(−1.296)
Ownershare	−0.068	4.159 ***	−0.680	−0.442	0.207	−0.592
	(−0.121)	(5.497)	(−1.290)	(−0.770)	(1.280)	(−1.061)
Duality	0.011	0.236	−0.024	−0.06	−0.010	−0.053
	(0.090)	(1.253)	(−0.199)	(−0.466)	(−0.234)	(−0.425)
Independence	0.016 *	0.018	0.013	0.013	0.008	0.008
	(1.800)	(1.321)	(1.605)	(1.410)	(1.458)	(0.857)
Audit	0.245	1.966 ***	−0.044	0.615 **	0.113 **	0.533 **
	(0.989)	(4.093)	(−0.188)	(2.199)	(1.994)	(1.963)
GDP_Growth	−0.080 ***	−0.067 *	−0.070 **	−0.052	0.008	−0.058 *
	(−2.698)	(−1.680)	(−2.464)	(−1.570)	(1.152)	(−1.780)
Constant	39.368 ***	41.429 ***	33.271 ***	34.496 ***	−0.298	34.712 ***
	(15.857)	(12.339)	(13.987)	(13.239)	(−0.297)	(13.857)
Firm fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
N	18,423	18,423	18,423	14,344	14,344	14,344
Adj R2	0.143	0.12	0.188	0.134	0.05	0.152

This table presents the regression outcomes assessing how ESG performance influences bankruptcy risk through enhancements in firms’ operating performance. Definitions of all variables are provided in the Appendix A. Standard errors are adjusted for heteroskedasticity using the Huber–White robust estimation method and are clustered at the firm level. The t-statistics are shown in parentheses below the coefficients. The symbols ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

. Columns (1)–(3) present the estimation results of Equations (4), (9), and (10), respectively, using ROA as the proxy for operating performance. Columns (4)–(6) replicate the analysis using EPS as the alternative proxy.

The results in Columns (2) and (5) show that ESG performance is significantly and positively related with both ROA and EPS, indicating that stronger ESG performance enhances firms’ operating performance (Gillan et al., 2021). Furthermore, the findings suggest that operating performance partially mediates the relationship between ESG performance and bankruptcy risk. Overall, these results support the hypothesis that improved operating performance is one mechanism through which ESG performance reduces firms’ bankruptcy risk.

5.1.4. Lowering the Leverage Ratio

Finally, we posit that ESG performance may reduce a firm’s bankruptcy risk by lowering its leverage ratio. As analyzed earlier, strong ESG performance often enhances firm value and decreases the need for external financing (Edmans, 2011; Servaes & Tamayo, 2013; Lins et al., 2017). It can also lower the cost of capital (Qiu & Yin, 2019; Eliwa et al., 2021), thereby reducing firms’ dependence on debt financing. Furthermore, improved ESG practices help mitigate information asymmetry, enabling firms to raise equity capital more efficiently (Lemma et al., 2022) and consequently lessen their reliance on debt. Since lower leverage has been shown to correspond with reduced bankruptcy risk (Ellul & Pagano, 2019), this mechanism provides a plausible channel through which ESG performance enhances financial stability.

To test this mechanism, we use the debt-to-total-asset ratio as a proxy for leverage. Consistent with the previous sections, we employ a mediation model to examine this hypothesis.

The equations used in the model are presented below:

$${Risk}_{i,t+1}={\alpha }_0+{\alpha }_1{ESG}_{i,t}+{\alpha }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

$${Lev}_{i,t}={\beta }_0+{\beta }_1{ESG}_{i,t}+{\beta }_2{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(11)

$${Risk}_{i,t+1}={\gamma }_0+{\gamma }_1{ESG}_{i,t}+{\gamma }_2{Lev}_{i,t}+{\gamma }_3{Control}_{i,t}+{\phi }_i+{\mu }_t+{\epsilon }_{i,t}$$

(12)

The analytical procedure follows the same approach described in Section 5.1.1, and the results are shown in

Table 9. Analysis on ESG performance’s impact on bankruptcy risk through lowering the leverage ratio.

	(1)	(2)	(3)
	Risk_Zscore	Lev	Risk_Zscore
ESG	0.134 ***	−0.011 ***	0.058 *
	(4.020)	(−8.508)	(1.804)
Leverage			−7.188 ***
			(−16.671)
Size	−1.599 ***	0.076 ***	−1.051 ***
	(−13.985)	(16.518)	(−9.821)
ROE	0.036 ***	−0.002 ***	0.020 ***
	(11.539)	(−19.008)	(6.423)
Age	−0.014	−0.007 ***	−0.063 ***
	(−0.669)	(−8.909)	(−3.195)
Growth	−0.037	0.015 ***	0.069
	(−0.591)	(6.862)	(1.146)
Ownershare	−0.068	−0.024	−0.241
	(−0.121)	(−1.145)	(−0.463)
Duality	0.011	−0.001	0.007
	(0.090)	(−0.132)	(0.065)
Independence	0.016 *	−0.000	0.014 *
	(1.800)	(−0.780)	(1.664)
Audit	0.245	−0.029 ***	0.040
	(0.989)	(−2.985)	(0.167)
GDP_Growth	−0.080 ***	0.003 ***	−0.059 **
	(−2.698)	(2.786)	(−2.144)
Constant	39.368 ***	−1.085 ***	31.572 ***
	(15.857)	(−10.640)	(13.834)
Firm fixed effects	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes
N	18,423	18,423	18,423
Adj R2	0.143	0.16	0.183

This table displays the regression results examining how ESG performance influences bankruptcy risk by reducing firms’ leverage ratios. Detailed definitions of all variables are provided in the Appendix A. Standard errors are clustered at the firm level and corrected for heteroskedasticity using the Huber–White robust estimator. The t-statistics are shown in parentheses beneath the coefficients. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively.

. Columns (1)–(3) present the estimation results of Equations (4), (10), and (12), respectively. The coefficient of ESG performance in Column (2) is −0.011 and statistically significant at the 1% level, indicating that ESG performance is negatively associated with leverage. This finding is consistent with prior research (Lemma et al., 2022; Asimakopoulos et al., 2023; Asimakopoulos et al., 2024).

Furthermore, the results suggest that leverage partially mediates the relationship between ESG performance and bankruptcy risk. Overall, the evidence in Table 9 supports our conjecture that stronger ESG performance decreases bankruptcy risk by lowering firms’ leverage ratios.

5.2. Heterogeneity Analysis

To gain deeper insight into the connection between ESG performance and bankruptcy risk, this section explores whether the relationship differs across various firm characteristics. The heterogeneity results suggest that certain firm-specific factors significantly influence the magnitude and direction of this relationship.

5.2.1. Ownership Structure

Previous research has highlighted notable distinctions between state-owned enterprises (SOEs) and non-state-owned enterprises (non-SOEs) in aspects such as governance structure, political affiliation, and ESG-related behavior (J. Lin et al., 2020; Chen et al., 2018). In the Chinese context, SOEs are typically obligated to implement ESG initiatives as part of broader governmental directives. In contrast, non-SOEs tend to pursue ESG activities more voluntarily and with greater flexibility, since they operate with limited government intervention or support. Consequently, robust ESG performance in non-SOEs may act as a credible signal of financial stability and operational soundness, enhancing their appeal to investors and other stakeholders.

Non-SOEs generally encounter higher financing constraints and competitive pressures, increasing their exposure to financial distress. Conversely, SOEs often benefit from government backing, which can cushion them against bankruptcy risk. Based on these distinctions, we propose that ESG performance plays a stronger role in mitigating bankruptcy risk among non-SOEs than among SOEs.

To empirically evaluate this hypothesis, the sample is divided into two groups—SOEs and non-SOEs—and the baseline regression model is estimated separately for each subset. The results, reported in Columns (1) and (2) of

Table 10. Heterogeneous analysis based on sample firms categorized by ownership structure and geographical locations.

	Ownership Structure		Geographic Locations
	(1) SOE	(2) Non_SOE	(3) East	(4) Central and Western
ESG	0.083 **	0.137 **	0.152 ***	0.095 *
	(2.457)	(2.557)	(3.460)	(1.932)
Size	−1.230 ***	−1.670 ***	−1.696 ***	−1.376 ***
	(−10.482)	(−8.738)	(−11.612)	(−7.173)
ROE	0.031 ***	0.037 ***	0.036 ***	0.035 ***
	(9.823)	(7.564)	(9.609)	(6.381)
Age	−0.005	−0.038	−0.013	−0.010
	(−0.233)	(−0.972)	(−0.450)	(−0.343)
Growth	−0.050	−0.064	−0.041	−0.046
	(−0.869)	(−0.666)	(−0.512)	(−0.461)
Ownershare	−1.038 *	−0.347	0.039	−0.706
	(−1.714)	(−0.386)	(0.055)	(−0.753)
Duality	0.055	0.004	−0.026	0.111
	(0.431)	(0.026)	(−0.172)	(0.543)
Independence	0.007	0.027 *	0.020 *	0.010
	−0.753	−1.706	−1.669	−0.812
Audit	0.521 **	0.011	0.302	0.147
	(2.296)	(0.030)	(1.043)	(0.311)
GDP_Growth	−0.016	−0.153 ***	−0.065 *	−0.062
	(−0.525)	(−2.710)	(−1.654)	(−1.495)
Constant	30.945 ***	41.616 ***	41.186 ***	34.610 ***
	(12.386)	(10.067)	(12.746)	(8.543)
Firm fixed effects	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes
N	8226	10,197	12,237	6186
Adj R2	0.116	0.165	0.149	0.132

This table presents the regression results for subsamples grouped according to firms’ ownership structure and geographic region. Standard errors are clustered at the firm level and corrected for heteroskedasticity using the Huber–White robust estimation method. The t-statistics are shown in parentheses beneath the coefficients. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively.

, show that the ESG performance coefficients are 0.083 for SOEs and 0.137 for non-SOEs, both statistically significant at the 5% level. The comparatively larger coefficient for non-SOEs suggests that ESG performance has a more pronounced effect in lowering bankruptcy risk for these firms. Overall, the evidence supports the notion that ESG engagement functions as a more effective protective mechanism for non-SOEs relative to SOEs.

5.2.2. Geographical Locations

China’s regions display notable disparities in economic development. The eastern provinces generally possess higher per capita GDP and more mature financial systems than the central and western areas. In addition, local authorities in the east tend to prioritize environmental, social, and governance (ESG) initiatives, encouraging firms to strengthen their ESG practices. Prior studies have also shown that geographic location can play an important role in shaping firms’ financial and operational outcomes (Becker, 2007; Arena & Dewally, 2012). Based on this context, we posit that ESG performance should have a more pronounced effect in reducing bankruptcy risk for firms operating in the eastern region relative to those located in the central and western regions.

Following previous literature, the sample firms are divided into two groups according to geographic region: one comprising firms based in the eastern region and the other including those from the central and western regions. The baseline regression model is then estimated separately for each group. The results, presented in Columns (3) and (4) of Table 10, indicate that the coefficient for ESG performance is 0.152 for eastern-region firms—significant at the 1% level—whereas the corresponding coefficient for firms in the central and western regions is 0.095 and significant only at the 10% level.

These results imply that the impact of ESG performance on lowering bankruptcy risk is more substantial among firms located in the economically advanced eastern provinces than among those operating in China’s central and western regions.

5.2.3. Industry Competition

Prior studies suggest that varying levels of industry competition influence firms’ operations, governance, and performance (Nickell, 1996). A more competitive environment tends to mitigate information asymmetry (Armstrong et al., 2011), and firms operating under intense competition often exhibit stronger ESG performance (Lopez-Gamero et al., 2009). In highly competitive markets, firms may leverage ESG engagement to differentiate themselves, strengthen their competitive advantage, and gain greater support from stakeholders. Therefore, we expect the relationship between ESG performance and bankruptcy risk to differ depending on the level of industry competition.

Following the existing literature, we use the Herfindahl–Hirschman Index (HHI) to measure industry competition. Based on the median value of the HHI, we classify firms into highly competitive industries and less competitive industries. We then estimate the baseline regression model separately for each group. The results, presented in Columns (1) and (2) of

Table 11. Heterogeneous analysis based on firms categorized by levels of competition and business life cycles.

	Level of Competition		Business Life Cycle
	(1) Low	(2) High	(3) Growth	(4) Mature	(5) Decline
ESG	0.079 *	0.172 ***	0.076 **	0.203 ***	0.123
	(1.773)	(3.240)	(2.227)	(3.262)	(0.988)
Size	−1.491 ***	−1.705 ***	−1.264 ***	−1.854 ***	−1.917 ***
	(−8.720)	(−11.175)	(−9.323)	(−8.678)	(−6.572)
ROE	0.036 ***	0.033 ***	0.031 ***	0.045 ***	0.016 **
	(8.782)	(7.222)	(6.515)	(7.170)	(2.382)
Age	0.000	−0.049 *	0.023	−0.034	−0.030
	(0.002)	(−1.692)	(0.832)	(−0.978)	(−0.446)
Growth	0.094	−0.214 **	−0.093	0.108	0.405 *
	(1.089)	(−2.384)	(−1.500)	(0.750)	(1.691)
Ownershare	−0.818	0.288	−0.497	0.407	0.614
	(−1.057)	(0.353)	(−0.905)	(0.460)	(0.337)
Duality	−0.136	0.272	0.175	0.003	−0.113
	(−0.811)	(1.570)	(1.231)	(0.013)	(−0.303)
Independence	0.022 *	0.006	0.025 ***	−0.002	−0.001
	(1.688)	(0.488)	(2.578)	(−0.164)	(−0.037)
Audit	−0.207	0.480	0.528 **	0.521	0.087
	(−0.541)	(1.587)	(1.977)	(1.365)	(0.133)
GDP_Growth	−0.023	−0.148 ***	0.009	−0.109 *	−0.118
	(−0.547)	(−3.402)	(0.294)	(−1.929)	(−1.202)
Constant	36.811 ***	42.718 ***	29.667 ***	46.093 ***	47.948 ***
	(9.892)	(12.903)	(10.359)	(10.073)	(7.643)
Firm fixed effects	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes
N	9007	9416	8631	6754	2958
Adj R2	0.134	0.148	0.165	0.131	0.169

This table presents the regression outcomes for subsamples classified according to firms’ market competition levels and the phase of the business cycle. Standard errors are clustered at the firm level and corrected for heteroskedasticity using the Huber–White robust estimation method. The t-statistics are shown in parentheses beneath the coefficients. The symbols ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

, show that the coefficient of ESG performance is 0.172 for firms in highly competitive industries—significant at the 1% level—whereas the coefficient for firms in less competitive industries is 0.079, significant at the 10% level.

These findings indicate that ESG performance has a stronger mitigating effect on bankruptcy risk for firms operating in very competitive markets, suggesting that competition amplifies the benefits of strong ESG engagement.

5.2.4. Business Life Cycle

In accordance with the firm life cycle theory (Mueller, 1972), organizational characteristics, operational conditions, and capital allocation efficiency vary across different stages of a firm’s life cycle, resulting in differing levels of bankruptcy risk. To implement ESG policies effectively, firms must tailor their ESG strategies to their specific life cycle stages. Firms in the growth or mature stages typically experience stable profits and cash flows, whereas those in the decline stage often face shrinking market share and profitability, leading to tighter cash flows. For firms in the declining stage, prioritizing non-financial objectives such as ESG investments may further weaken operational performance (Gao et al., 2021). Therefore, we hypothesize that ESG performance tends to reduce bankruptcy risk for firms in the growth and mature stages than for those in decline.

Consistent with the classification method of Dickinson (2011), firms are divided into three categories—growth, mature, and declining—according to their respective cash flow patterns. The baseline regression model is then estimated separately for each category. As shown in Columns (3)–(5) of Table 11, ESG performance exhibits a significant negative association with bankruptcy risk for firms in the growth and mature stages, whereas the coefficient for firms in the declining stage is not statistically significant.

These findings are consistent with our expectations and suggest that the effectiveness of ESG practices in mitigating bankruptcy risk varies across the business life cycle. In particular, firms in growth and maturity stages benefit more from ESG engagement, while those in decline may face limited advantages.

5.3. Analysis Based on the Three Pillars of ESG Performance

Building upon the previous analysis of aggregate ESG performance, this section explores how each of the three ESG components—environmental (E-Rating), social (S-Rating), and governance (G-Rating)—individually relates to bankruptcy risk. This approach follows prior research by Kim et al. (2018) and Bernardi and Stark (2018). To conduct this analysis, the baseline regression model (Equation (1)) is re-estimated using each ESG dimension separately and then all three simultaneously. Data for these component scores are sourced from the Sino-Securities Index ESG rating database.

The environmental (E) dimension is expected to reduce bankruptcy risk by enhancing firms’ compliance with environmental regulations, improving resource efficiency, and lowering potential costs related to environmental liabilities and policy sanctions. The social (S) dimension mitigates bankruptcy risk by strengthening stakeholder relationships, improving employee morale, and enhancing corporate reputation, which collectively stabilize firm performance and reduce operational uncertainty. The governance (G) dimension is theorized to have the strongest effect, as effective governance reduces agency costs, promotes transparency, and ensures sound managerial decision-making, thereby directly improving financial stability and lowering bankruptcy risk. Together, these theoretical expectations suggest that while all three ESG dimensions contribute to reducing bankruptcy risk, the strength and significance of their effects may differ.

The following analysis empirically tests these relationships, with the results reported in

Table 12. Analysis on the relation between the three pillars of ESG and bankruptcy risk.

	(1)	(2)	(3)	(4)
	Risk_Zscore	Risk_Zscore	Risk_Zscore	Risk_Zscore
ERating	0.015			0.008
	(0.418)			(0.206)
SRating		0.085 ***		−0.025
		(2.832)		(−0.613)
GRating			0.152 ***	0.163 ***
			(5.116)	(4.416)
Size	−1.574 ***	−1.588 ***	−1.578 ***	−1.575 ***
	(−13.859)	(−13.947)	(−13.978)	(−13.948)
ROE	0.036 ***	0.036 ***	0.036 ***	0.036 ***
	(11.612)	(11.551)	(11.482)	(11.505)
Age	−0.016	−0.018	−0.003	−0.002
	(−0.769)	(−0.860)	(−0.153)	(−0.092)
Growth	−0.046	−0.042	−0.037	−0.037
	(−0.739)	(−0.672)	(−0.595)	(−0.601)
Ownershare	−0.027	−0.047	−0.177	−0.180
	(−0.049)	(−0.084)	(−0.318)	(−0.324)
Duality	0.004	0.008	0.008	0.007
	(0.028)	(0.069)	(0.062)	(0.054)
Independence	0.018 **	0.017 *	0.013	0.013
	(2.043)	(1.909)	(1.516)	(1.516)
Audit	0.283	0.265	0.205	0.204
	(1.141)	(1.065)	(0.831)	(0.830)
GDP_Growth	−0.079 ***	−0.079 ***	−0.079 ***	−0.079 ***
	(−2.658)	(−2.673)	(−2.677)	(−2.673)
Constant	39.223 ***	39.311 ***	38.648 ***	38.588 ***
	(15.819)	(15.843)	(15.780)	(15.737)
Firm fixed effects	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes
N	18,423	18,423	18,423	18,423
Adj R2	0.142	0.143	0.145	0.144

This table presents the regression results where the Environmental rating, Social rating, and Governance rating are each used as the dependent variable. Standard errors are clustered at the firm level and corrected for heteroskedasticity using the Huber–White robust estimation method. The t-statistics are shown in parentheses beneath the coefficient estimates. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively.

. Columns (1)–(3) report the results for the environmental, social, and governance ratings, respectively, while Column (4) presents the estimates when all three components are included simultaneously. When analyzed individually, both the S-Rating and G-Rating are positively associated with lower bankruptcy risk at the 1% significance level, whereas the E-Rating coefficient is statistically insignificant. In Column (4), when all three components are included together, the G-Rating remains significant with a coefficient of 0.163 (at the 1% level), while the effects of the E-Rating and S-Rating lose statistical significance.

These results suggest that the mitigating effect of ESG performance on bankruptcy risk is primarily driven by the governance dimension, rather than by environmental or social factors. A plausible explanation is that an effective corporate governance system enhances operational efficiency, reduces agency costs, and aligns the interests of shareholders and other stakeholders. Strong governance also improves performance and innovation while promoting transparency in information disclosure, thereby lowering bankruptcy risk.

The limited impact of the environmental and social dimensions may be explained by two factors: First, excessive engagement in social or environmental initiatives may lead to resource misallocation and inefficiency, ultimately increasing bankruptcy risk (Friedman, 1970). Second, managers may pursue such initiatives for personal reputation building or to divert attention from prior underperformance, which can also heighten bankruptcy risk (Hemingway & Maclagan, 2004).

Taken together, the empirical evidence consistently demonstrates that strong ESG performance significantly reduces firms’ bankruptcy risk and that this relationship is robust across multiple model specifications, alternative measures, and identification strategies. The analysis of underlying mechanisms further reveals that ESG engagement mitigates bankruptcy risk primarily by enhancing information transparency, easing financing constraints, improving operating performance, and reducing leverage. Moreover, the heterogeneity tests show that the influence of ESG performance is more pronounced for non-SOEs, firms located in economically developed regions, firms in highly competitive industries, and those in the growth and maturity stages of their life cycles. Finally, the component analysis indicates that the governance pillar of ESG plays the most critical role in lowering bankruptcy risk. These findings provide important implications for policymakers, regulators, and corporate managers.

6. Policy Implications

The results of this study offer several policy implications that can help promote financial stability and sustainable corporate development in China.

First, policymakers and regulators should recognize the stabilizing role of ESG engagement in reducing corporate bankruptcy risk. Regulatory authorities may consider integrating ESG criteria into financial supervision frameworks and credit risk assessments. For instance, banks and other financial institutions could incorporate ESG ratings into their credit evaluation models to better capture firms’ default probabilities. Encouraging ESG disclosure through standardized reporting frameworks can also enhance market transparency and investor confidence.

Second, government agencies—especially those in less developed central and western regions—can promote regional economic resilience by incentivizing ESG adoption among local firms. Targeted subsidies, tax incentives, or preferential lending policies for firms with strong ESG records could help reduce financing constraints and improve overall credit quality in the corporate sector.

Third, corporate managers should view ESG investments not merely as a compliance obligation or reputational tool but as a strategic asset that enhances long-term financial stability. Strengthening corporate governance, in particular, can yield substantial benefits by improving operational efficiency, reducing information asymmetry, and mitigating bankruptcy risk. Managers should align ESG initiatives with firm-specific conditions, such as life cycle stage and industry competition, to maximize their effectiveness.

Finally, investors and stakeholders can use ESG performance—especially governance indicators—as an important signal of firms’ financial resilience. Incorporating ESG criteria into investment decisions not only aligns with sustainable finance principles but also helps investors manage portfolio risk more effectively.

Overall, these policy implications underscore the importance of promoting ESG integration at both the regulatory and corporate levels to foster a more resilient and sustainable financial system in China.

7. Conclusions

This study investigates the link between ESG performance and the likelihood of bankruptcy using a comprehensive dataset of firms listed on the Shenzhen and Shanghai Stock Exchanges from 2009 to 2022. By employing multiple empirical methods—including instrumental variable (IV) estimation, propensity score matching (PSM), and several robustness checks with alternative indicators—we consistently find that firms exhibiting stronger ESG performance experience a lower probability of bankruptcy.

The mechanism analysis indicates that ESG performance reduces bankruptcy risk through several channels: enhancing information transparency, easing financial constraints, improving operational efficiency, and decreasing leverage ratios. Furthermore, the heterogeneity analysis reveals that this effect is more significant for non-state-owned enterprises (non-SOEs), firms located in China’s economically advanced eastern regions, companies operating in highly competitive markets, and those in the growth or maturity phases of their business life cycle. When examining the individual ESG pillars, the findings suggest that corporate governance—rather than environmental or social dimensions—is the primary factor driving the observed reduction in bankruptcy risk.

Overall, the takeaways of this paper are summarized in

Table 13. Summary of Hypotheses and Findings.

Hypotheses
Hypothesis I	There is a negative relationship between ESG performance and bankruptcy risk; in other words, firms with higher ESG performance are expected to exhibit a lower likelihood of bankruptcy.
Hypothesis II	ESG performance mitigates bankruptcy risk through several channels, including enhanced information transparency, eased financial constraints, improved operational efficiency, and reduced leverage levels.
Hypothesis III	The three pillars of ESG performance—environmental, social, and governance—affect bankruptcy risk differently. Each component is anticipated to contribute to lowering bankruptcy risk, with the governance aspect exerting the strongest influence.
Channels via which ESG performance affect bankruptcy risk
Improving information asymmetry
Alleviating financing constraints
Improving operating performance
Lowering leverage ratio
Findings
Firms with stronger ESG performance exhibit a lower likelihood of bankruptcy. ESG performance helps reduce bankruptcy risk by enhancing corporate information transparency, relieving financial constraints, improving operational efficiency, and decreasing leverage levels. The mitigating effect of ESG performance on bankruptcy risk is more substantial among non-state-owned enterprises (non-SOEs), firms operating in less economically developed areas of China, those in highly competitive industries, and companies in the growth or maturity phases of their life cycle. Among the three ESG pillars, the governance pillar has the most significant impact in reducing bankruptcy risk, compared with environmental and social components.

, which makes it easier for readers to understand the main hypotheses and findings holistically. These results enrich the literature on ESG and corporate financial outcomes by offering robust evidence from the Chinese context—an emerging market where ESG integration is still developing. The findings highlight the critical role of governance quality and the need for ESG strategies tailored to firms’ structural characteristics and regional environments to enhance financial stability and resilience.

From a policy standpoint, the evidence implies that regulators, investors, and corporate leaders should regard ESG engagement as an essential mechanism for promoting long-term corporate sustainability and mitigating systemic financial risks. Strengthening ESG implementation—particularly through improved governance practices—can contribute to building a more resilient, transparent, and sustainable capital market in China.

Future research could further explore the dynamic interactions between ESG performance and other dimensions of financial risk, such as credit spreads, the cost of debt, and long-term investment efficiency. Investigating the evolving role of ESG disclosure regulations and investor activism may also provide valuable insights into how ESG integration shapes financial stability in emerging markets.

In addition, future research could further extend our analysis by conducting cross-country comparisons to examine whether the relationship between ESG performance and bankruptcy risk differs across emerging and developed markets. Such studies would help identify how variations in institutional quality, regulatory enforcement, and cultural norms influence the effectiveness of ESG practices in mitigating financial distress. Moreover, given the significant economic disruptions caused by the COVID-19 pandemic, future research could explore how the ESG–risk relationship evolves across different economic cycles, particularly by comparing pre- and post-pandemic periods. Investigating these temporal and cross-market dynamics would deepen the understanding of ESG’s role in enhancing corporate resilience under varying macroeconomic and institutional conditions.