Intelligent Transformation: The Invisible Shield Against Corporate Credit Risk

Abstract

In the context of a progressively intricate and uncertain global economic landscape, the credit risk businesses encounter is intensifying. This study seeks to analyze whether intelligent transformation, a significant trend in current organization development, might serve as a novel method for mitigating credit risk. We employ panel data from 1533 listed enterprises in China’s manufacturing sector to investigate how intelligent transformation influences credit risk empirically. This research indicates that intelligent transformation can mitigate business credit risk. The production, management, and financing effects are the primary mechanisms via which intelligent transformation mitigates credit risk. Heterogeneity analysis indicated that the credit risk reduction effect of the intelligent transformation of traditional manufacturing firms surpassed that of intelligent manufacturing enterprises. In contrast to high-growth firms, low-growth enterprises exhibited more robust credit risk mitigation benefits from intelligent transformation. Subsequent analysis indicated that enhancing supply chain finance can facilitate intelligent transformation and, hence, more effectively mitigate credit risk.

1. Introduction

Businesses’ credit risk has increased in a more complicated and unpredictable global economic climate. Economic swings and market rivalry uncertainties have heightened credit risk exposure for enterprises figure [1]. Credit risk is intricately linked to a firm’s financing capacity and influences its reputation [2] and supply chain alliances [3], impacting its sustainable development. Current research indicates that corporate governance [4], fintech [5], ESG [6], and managerial disparities [7] influence company credit risk. In recent years, corporate credit risk has become a significant issue due to economic downturns and market volatility, particularly for traditional manufacturing firms, which urgently require innovative strategies to mitigate credit risk.

In 2015, the Chinese government introduced “Made in China 2025”, fostering the robust advancement of intelligent enterprise transformation, transitioning from “Made in China” to “Intelligent manufacturing in China”, and achieving the strategic objective of ascending the global value chain. Intelligent transformation, a pivotal trend in contemporary firm development, can enhance production efficiency, management quality, and market responsiveness, diminishing credit risk. Initially, implementing advanced technologies and automated processes can improve the production operations of enterprises, minimize human errors, and elevate production efficiency [8,9], thereby augmenting enterprises’ profitability and financial stability. Secondly, intelligent transformation allows organizations to attain more precise cost management and resource distribution, better overall operational efficiency, and mitigate operational risk, hence improving the solvency of enterprises [10]. Thirdly, intelligent transformation enhances the construction of enterprise informatization, facilitating improved data analysis and decision-making support. This advancement increases the scientific rigor and foresight of decisions, mitigating risks associated with decision-making errors and optimizing internal control. Moreover, intelligent transformation aids firms in augmenting their market competitiveness [11], enhances the innovative capacity of items and services, and further fortifies their resilience to risks and creditworthiness. Consequently, intelligent transformation augments organizations’ operational capacity and fundamentally mitigates their credit risk by strengthening their financial standing and bolstering their market competitiveness.

Research on intelligent transformation and enterprise credit risk is limited, with the majority focusing on the correlation between digital transformation and enterprise risk. This study collectively refers to digital transformation and intelligent transformation as intelligent transformation, given that intelligent transformation represents an advanced stage of digital transformation. The correlation between intelligent transformation and enterprise risk is a contention among scholars. The majority of scholars assert that intelligent transformation can mitigate enterprise risk. For instance, it mitigates the risk of stock market crashes [12,13,14,15] and corporate insolvency [16], enhances corporate risk-taking [17,18,19], and alleviates financial distress [20]. As intelligent transformation progresses to the phase of “comprehensive reinvention”, multinational organizations confront the predicament of “squeezed transformation”, characterized by a limited timeframe and significant challenges. Certain researchers contend that this problem could elicit the adverse effects of intelligent transformation, including corporate non-compliance due to heightened operational complexity and diminished information quality [21]. Certain researchers contend that the correlation between intelligent transformation and danger is non-linear. Huang et al. (2023) posited that the correlation between intelligent transformation and idiosyncratic risk exhibits a U-shaped relationship [22]. Ai et al. (2023) contended that the impact of intelligent transformation on the likelihood of stock price crashes exhibits an inverted U-shaped pattern characterized by initial exacerbation followed by subsequent suppression [23]. What is the precise relationship between intelligent transformation and company risk, particularly credit risk? This subject requires further empirical investigation.

This study empirically investigates the impact of intelligent transformation on reducing corporate credit risk, utilizing a sample of publicly listed Chinese manufacturing companies. It presents potential innovations: first, there is a lack of consensus regarding the effect of intelligent transformation on company risk, whether beneficial or detrimental. Furthermore, current research does not investigate the correlation between intelligent transformation and credit risk. Our research addresses this gap and indicates that intelligent transformation can mitigate corporate credit risk, offering new evidence for the theoretical viability of intelligent transformation. Secondly, we develop the theoretical framework of “intelligent transformation-production effect/management effect/financing effect-credit risk”, which enhances the research concept of “intelligent transformation-credit risk” and offers novel perspectives and frameworks for future theoretical investigations and policy development in related domains. The heterogeneity study indicates that intelligent transformation implementation’s credit risk mitigation effect differs among various enterprise types and across distinct periods of their life cycles. This also offers pragmatic advice on enhancing the regulation of intensity and velocity throughout the implementation of intelligent transformation in organizations. Third, we incorporate supply chain finance into analyzing the relationship between intelligent transformation and credit risk to investigate the facilitative impact of novel financial instruments on credit risk mitigation. This groundbreaking discovery offers a novel theoretical foundation for integrating supply chain financing with the intelligent transformation of firms and holds substantial importance in informing policymakers and corporate practices.

The subsequent sections of this document are structured as follows: The subsequent section encompasses the theoretical analysis and research hypotheses. The third part delineates the research design. The fourth part comprises the research analysis. The final part constitutes the conclusion and discussion.

2. Theoretical Analysis and Research Hypothesis

The intelligent transformation of manufacturing enterprises entails the comprehensive integration of advanced digital technologies into all facets of production, management, and financing, thereby effectively mitigating credit risk by enhancing production efficiency, optimizing management processes, and refining financing methods. The analysis is detailed as follows:

2.1. Production Effect

The intelligent transformation of manufacturing enterprises enhances production efficiency. Intelligent transformation enhances production efficiency by developing intelligent equipment, production lines, and workshops. The production assets in traditional manufacturing industries are typically proprietary [24]. When products are discontinued, companies must either idle the existing assets or acquire new ones. This results in the wastage of production resources and elevates the costs associated with repeated acquisitions. Intelligent transformation effectively integrates various modules by connecting assets with digital technology, leveraging self-learning capabilities and adaptability. Modifying the production line to accommodate new product requirements enhances production efficiency [8,9]. Enhancing production efficiency reduces production costs, increases product value, increases operating income, and diminishes default risk [25]. A positive correlation between productivity and stock returns in the subsequent year has also been demonstrated [26]. An increase in stock return enhances a firm’s market value, decreasing the risk of default. Conversely, intelligent transformation employs digital technology to enhance production capacity. Enterprises can enhance production optimization by utilizing real-time data analysis, sensors, and IoT technology to monitor processes, adjust production plans promptly, and minimize resource waste. This enhances production efficiency and increases the scale of production. The expansion of production scale enables enterprises to amortize fixed costs and enhance unit production capacity via the scale effect [27]. The fixed cost per product unit decreases as production increases, enhancing the firm’s profitability. Increased profitability enhances a firm’s capacity to manage market risks and financial pressures, thereby mitigating credit risk associated with these pressures.

2.2. Management Effect

The intelligent transformation of manufacturing companies can enhance the management efficiency of organizations. Intelligent transformation enhances internal control. By integrating digital technology and information systems, organizations can achieve transparent administration of all departments and business processes, guaranteeing that all operations are traceable and documented. Information transparency enhances the dependability of an organization’s internal control [28] and diminishes friction among management tiers. An efficient internal control system enables organizations to recognize potential financial, operational, and market risks. By enhancing the oversight and administration of diverse risks, companies can proactively implement strategies to mitigate credit crises resulting from risk events [29,30]. Conversely, intelligent transformation allows enterprises to leverage technologies like big data and artificial intelligence for expedited market demand analysis, rapid response, and product and service innovation implementation. The capacity for fast responses to market demand gives enterprises a competitive advantage, facilitating the launch of innovative products that more effectively address consumer needs, thereby enhancing market share and dominance [11]. Enhanced market power can amplify the influence of enterprises within the supply chain, augment the volume of commercial credit [31], lower credit costs, and consequently mitigate credit risk.

2.3. Financial Effect

The intelligent transformation of manufacturing businesses can improve the efficiency of corporate financing. Intelligent transformation can diminish information asymmetry between organizations and stakeholders [32]. Businesses can disseminate information regarding inventory, production, transportation, and other aspects to suppliers and customers using data-sharing platforms to improve supply chain transparency. This bolsters trust among supply chain enterprises and boosts engagement efficiency with external partners. When every link in the supply chain can exchange real-time information, organizations can more promptly identify possible sources of risk. Production disruptions, shipping delays, and quality issues can be swiftly identified through information sharing, enabling rapid adjustments to mitigate the adverse effects of these risks on corporate credit. Conversely, intelligent transformation can mitigate financial restrictions. Intelligent manufacturing technologies can profoundly influence a company’s financial standing and resource distribution by enhancing operational efficiency and productivity, potentially leading to improved financial performance, reputation, and fewer financing limitations [33]. Moreover, intelligent transformation standardizes, clarifies, and optimizes corporate operations, increasing the confidence of external investors and financial institutions. The augmentation of trust enables firms to secure more advantageous financing terms via the stock market or bank loans. Mitigating funding limitations can significantly diminish credit risk [34].

Figure 1 presents the conceptual framework diagram for this study.

3. Research Methodology

3.1. Data and Sample

This paper focuses on A-share listed companies within the manufacturing sector. In 2015, China introduced “Made in China 2025” to actively advance the intelligent transformation of its manufacturing sector. The sample period is established as 2015–2023 according to this policy. The sample is treated as follows: (1) The sample is restricted to Chinese A-share listed companies, excluding non-listed entities and those listed after 31 December 2015, to ensure data availability for calculating the distance to default; (2) the focus is solely on the manufacturing industry; and (3) to mitigate the influence of outliers, all continuous variables are adjusted using shrink-tailing at the 1% and 99% thresholds. Firms that were delisted during the observation period were excluded from the sample. The outcome was 1533 listed companies within the manufacturing sector. This study utilized the Wind, CSMAR, and EPS databases, with Baidu as the search engine.

3.2. Variable Construction

3.2.1. Dependent Variable

Credit risk ($DD$): The dependent variable in this study is corporate credit risk, which is proxied by the distance to default ($DD$) calculated using the KMV model. A greater distance to default indicates lower credit risk. The KMV (Kealhofer, McQuown, and Vasicek) model is a market-value-based credit risk measurement approach developed by KMV Corporation, later acquired by Moody’s. It extends the structural credit risk model proposed by Merton (1974) and is widely used in corporate credit risk assessment and default probability prediction.

The core concept of the KMV model is that a firm’s default is determined by the relationship between its market value of assets and its debt obligations. Default occurs when the market value of a firm’s assets ($V_A$) falls below the book value of the company’s debt ($D$). The critical threshold for default is reached when the company’s asset value equals its debt value. The KMV model quantifies the proximity of a firm’s asset market value to the default threshold using the distance to default ($DD$), where a larger $DD$ indicates a lower probability of default, and a smaller $DD$ suggests a higher risk of default. In this model, corporate equity is viewed as a European call option on the firm’s assets, with the debt value serving as the strike price. The calculation of the distance to default involves two steps:

First, based on the option pricing model, the firm’s asset market value ($V_A$) and asset volatility (${\sigma }_A$) are estimated using the firm’s equity value ($V_E$), equity volatility (${\sigma }_E$), risk-free interest rate ($r$), book value of debt ($D$), and the remaining time to debt maturity ($T$). These calculations were performed using MATLAB 2024a with the following equations:

$$V_E=V_{A}N(d_1)-D{e}^{-rT}N(d_2)$$

(1)

$$d_1={\displaystyle \frac{ln(\frac{V_A}{D})+(r+0.5{\sigma }_A^2)T}{{\sigma }_A\sqrt{T}}}$$

(2)

$$d_2=d_1-{\sigma }_A\sqrt{T}$$

(3)

$${\sigma }_E={\displaystyle \frac{N(d_1)V_A{\sigma }_A}{V_E}}$$

(4)

In the above equations, $N(d)$ represents the standard cumulative normal distribution function. $V_E$ denotes the firm’s equity value, $D$ represents the book value of the firm’s debt, and $V_A$ refers to the market value of the firm’s assets. The debt maturity period $T$ is set to one year. ${\sigma }_A$ represents the volatility of the firm’s asset value, while $r$ is the risk-free interest rate, which is derived from the Chinese government bond yield published by the Ministry of Finance, calculated as the average of daily one-year yields. ${\sigma }_E$ represents the annualized volatility of the firm’s equity value. The above data were from the Wind database.

Second, after obtaining the firm’s asset market value ($V_A$) and its volatility (${\sigma }_A$), the distance to default ($DD$) was calculated as follows:

$$DD={\displaystyle \frac{E(V_A)-DP}{E(V_A){\sigma }_A}}$$

(5)

where $E(V_A)$ represents the expected value of the company’s asset value, $DP$ denotes the default point, and $DP$ is computed as follows:

$$DP=SD+0.5LD$$

(6)

where $SD$ represents the enterprise’s short-term debt, whereas $LD$ denotes the enterprise’s long-term debt.

3.2.2. Independent Variable

Intelligent transformation (${lnIM}_{-}wf$): Based on the research of Wu et al.(2024) [35], we performed a word frequency analysis of business annual reports focusing on terms associated with “intelligent transformation”. The “intelligent transformation” terminology encompasses networking, digitization, informatization, intelligence, intelligentization, information technology, digital technology, Internet, Internet of Things, cloud computing, big data, artificial intelligence, information resources, e-commerce, intelligent manufacturing, smart manufacturing, digital platform, digital transformation, ERP, information system, information technology, information management, MES, automation, robotics, digital twin, digital delivery, and virtual prototyping. The intelligent transformation (${lnIM}_{-}wf$) is the natural logarithm of the annual report word frequency incremented by one.

3.2.3. Control Variables

This study identifies the control variables: growth ($Growth$), cash flow from operating activities ($CFO$), return on net assets ($ROE$), equity concentration ($Top10$), percentage of independent directors ($Indep$), dual positions ($Dual$), percentage of fixed assets ($PPE$), type of audit opinion ($Audit$), net interest rate on total assets ($ROA$), and GDP ($GDP$).

3.3. Model Specification

To test the central hypothesis of this paper, we constructed the following regression model:

$$D{D}_{i,t}={\alpha }_0+{\alpha }_1{lnIM}_{-}wf+{\alpha }_{2}Control{s}_{i,t}+{\displaystyle \sum Industry}+{\displaystyle \sum Year}+{\epsilon }_{i,t}$$

(7)

where $D{D}_{i,t}$ is the distance to default for firm $i$ in year $t$. A longer distance to default correlates with reduced credit risk. ${lnIM}_{-}w{f}_{i,t}$ represents the level of intelligent transformation of firm $i$ the year $t$. $Control{s}_{i,t}$ represents the control variables. Refer to

Table 1. Variable definitions.

Variable Type	Variable Name	Codes	Proxy Measures
Dependent Variable	Credit Risk	$DD$	The KMV model is used to calculate the distance to default.
Independent Variable	Intelligent transformation	${lnIM}_{-}wf$	Calculate the total word frequency using text mining, add 1 to the obtained result, and then take the logarithm.
Control variables	Growth	$Growth$	Rate of revenue growth.
	Cash flow from operating activities	$CFO$	Net cash flows from operating activities/total assets from the preceding year.
	Return on net assets	$ROE$	Net profit/average shareholders’ equity.
	Equity concentration	$Top10$	Proportion of equity held by the 10 largest shareholders.
	Percentage of independent directors	$Indep$	Number of independent directors/number of board of directors.
	Dual positions	$Dual$	1 for both the chairman and the general manager, 0 otherwise.
	Percentage of fixed assets	$PPE$	Net fixed assets/total assets.
	Type of audit opinion	$Audit$	1 denotes a standard unqualified opinion, whereas 0 signifies alternatives.
	Net interest rate on total assets	$ROA$	Net profit/total assets.
	GDP	$GDP$	Provincial GDP per capita is expressed in tens of thousands of RMB.

for detailed definitions and measurements of the variables. ${\displaystyle \sum Industry}$ and ${\displaystyle \sum Year}$ denote the fixed effects for industry and year, respectively, whereas e represents the residual. The standard errors for each regression are aggregated at the firm level.

4. Empirical Results and Analyses

4.1. Descriptive Statistics

Table 2. Descriptive statistics.

Variable	N	Mean	Max	Min	SD
$DD$	13,797	2.520	7.194	0.854	0.902
${lnIM}_{-}wf$	13,797	3.048	5.992	0	1.117
$Growth$	13,797	0.115	2.199	−0.609	0.298
$CFO$	13,797	0.0583	0.346	−0.217	0.0755
$ROE$	13,797	0.0470	0.596	−1.119	0.165
$Top10$	13,797	0.541	0.878	0.198	0.143
$Indep$	13,797	0.376	0.600	0.333	0.0546
$Dual$	13,797	0.269	1	0	0.443
$PPE$	13,797	0.227	0.663	0.0107	0.135
$Audit$	13,797	0.956	1	0	0.206
$ROA$	13,797	0.0339	0.296	−0.398	0.0742
$GDP$	13,797	8.856	20.03	2.881	3.626

presents the findings of the descriptive statistics. The average value of $DD$ is 2.520, the maximum value is 7.194, and the minimum value is 0.854, suggesting significant variability in credit risk across manufacturing businesses. The average value of ${lnIM}_{-}wf$ is 3.048, with a high of 5.992 and a low of 0, signifying considerable variability in the degree of intelligent transformation among manufacturing businesses.

4.2. Baseline Results

Table 3. Regression analysis.

	(1)	(2)
Variable	$DD$	$DD$
${lnIM}_{-}wf$	0.042 ***	0.046 ***
	(4.73)	(5.46)
$Growth$		−0.265 ***
		(−11.15)
$CFO$		−0.417 ***
		(−4.23)
$ROE$		0.202 **
		(2.23)
$Top10$		0.298 ***
		(4.14)
$Indep$		0.128
		(0.76)
$Dual$		−0.061 ***
		(−3.23)
$PPE$		0.404 ***
		(4.95)
$Audit$		0.107 ***
		(3.15)
$ROA$		0.257
		(1.23)
$GDP$		0.002
		(0.71)
$Constant$	2.393 ***	2.007 ***
	(89.01)	(21.79)
Observations	13,797	13,797
Industry FE	YES	YES
Year FE	YES	YES
Adjusted R-squared	0.416	0.430

Notes: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively (This also applies to the tables below).

presents the outcomes of the baseline regression. Column (1) displays the regression results devoid of control variables, revealing a coefficient of 0.042, significant at the 1% level. Column (2) illustrates the regression results inclusive of control variables, with an ${lnIM}_{-}wf$ coefficient of 0.046, also significant at the 1% level. This suggests that intelligent transformation can mitigate firms’ default risk.

4.3. Endogeneity Test

4.3.1. Instrumental Variable Method

This paper addresses the endogeneity problem arising from omitted variables by constructing instrumental variables for analysis using the share–shift method (Bartik). The results are presented in Columns (1) and (2) of

Table 4. Endogeneity test.

	Instrumental Variable Method		Heckman Method		Entropy Balance Method
	(1)	(2)	(3)	(4)	(5)
	First	Second	First	Second
Variable	${lnIM}_{-}wf$	$DD$	${lnIM}_{-}wf$	$DD$	$DD$
${IV}_{-}Bartik$	5.323 ***
	(0.0683)
${IV}_{-}{industry}_{-}other$			0.835 ***
			(0.0786)
${lnIM}_{-}wf$		0.0820 ***		0.0397 ***
		(0.0118)		(0.00743)
$IMR$				−0.0207
				(0.0323)
${IM}_{-}median$					0.0940 ***
					(0.0178)
$Growth$	−0.00965	−0.313 ***	−0.195 *	−0.178 ***	−0.251 ***
	(0.0248)	(0.0264)	(0.111)	(0.0268)	(0.0288)
$CFO$	0.375 ***	−0.462 ***	1.168 **	−0.602 ***	−0.319 ***
	(0.110)	(0.108)	(0.527)	(0.118)	(0.109)
$ROE$	0.293 ***	0.155	−0.129	0.275 ***	0.180
	(0.102)	(0.0952)	(0.384)	(0.0935)	(0.117)
$Top10$	0.0394	0.380 ***	0.247	0.0797	0.346 ***
	(0.0764)	(0.0796)	(0.261)	(0.0544)	(0.0790)
$Indep$	−0.158	0.142	1.717 **	0.239 *	0.134
	(0.170)	(0.182)	(0.771)	(0.140)	(0.178)
$Dual$	0.0176	−0.0641 ***	−0.195 **	−0.108 ***	−0.0690 ***
	(0.0217)	(0.0209)	(0.0793)	(0.0173)	(0.0199)
$PPE$	−0.535 ***	0.468 ***	−0.240	0.737 ***	0.330 ***
	(0.0903)	(0.0901)	(0.256)	(0.0608)	(0.0895)
$Audit$	0.125 ***	0.113 ***	0.316 **	0.0752 *	0.121 ***
	(0.0456)	(0.0352)	(0.135)	(0.0391)	(0.0348)
$ROA$	−0.375	0.447 **	0.939	0.181	0.200
	(0.232)	(0.218)	(0.955)	(0.223)	(0.255)
$GDP$	0.00658 *	0.00272	0.0249 **	0.0304 ***	0.00238
	(0.00338)	(0.00348)	(0.0124)	(0.00216)	(0.00347)
$Constant$	-	-	−1.148 ***	1.823 ***	2.083 ***
	-	-	(0.390)	(0.0748)	(0.0957)
Industry FE	YES	YES	NO	NO	NO
Year FE	YES	YES	NO	NO	NO
Observations	12,264	12,264	13,796	13,796	13,797
R-squared	0.4731	0.032	-	0.032	0.426
Cragg–Donald Wald F statistic	10,970.59		-	-	-
Kleibergen–Paap rk Wald F statistic	6076.42		-	-	-
Stock–Yogo weak ID test critical values: 10% maximal IV size	16.38		-	-	-
Kleibergen–Paap rk LM statistic	594.8		-	-	-

Notes: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively (This also applies to the tables below)

, where the coefficient from the first-stage regression is 5.323, significantly positive at the 1% level, indicating that the instrumental variables meet the correlation requirement. Furthermore, both Cragg-Donald’s Wald F statistic (10,970.59) and Kleibergen–Paap rk’s Wald F statistic (6076.42) exceed the critical value (16.38) at the 10% significance level, leading to the rejection of the weak instrument hypothesis. The coefficient from the second-stage regression is 0.0820, which is significantly positive at the 1% level and aligns with the baseline regression result. The LM statistic of Kleibergen–Paap rk (594.8) is significant at the 1% level and successfully meets the criteria for non-identifiability testing.

4.3.2. Heckman Method

This study measured this through textual analysis; however, self-selection bias may pose an issue, as intelligent transformation is not mandatory for disclosure. This study employs the Heckman selection model for testing purposes. In the initial stage of the model, “intelligent transformation” serves as the explanatory variable (${lnIM}_{-}wf$). In contrast, “the mean value of intelligent transformation of other enterprises in the same industry in the same year” is the instrumental variable (${IV}_{-}{industry}_{-}other$). The control variables align with those in the Equation (7), and the analysis is conducted using the Probit model. Subsequently, the value obtained in the initial step is utilized in the second step. The $IMR$ computed in the initial step was subsequently regressed into the second-stage model to account for self-selection bias. The regression results are presented in Columns (3) and (4) of Table 4. The coefficient from the first-stage regression is 0.835, while the second-stage regression coefficient is 0.0397. Both stages yield significantly positive results at the 1% level, corroborating the prior study’s findings.

4.3.3. Entropy Balance Method

This study employs the entropy balancing method to mitigate sample selection bias. The entropy balancing method effectively manages the multidimensional balance of sample covariates across treatment and control groups. Unlike propensity score matching, it retains all observations, even unbalanced ones. The regression results are presented in Column (5) of Table 4, indicating that the coefficient of ${IM}_{-}median$ is 0.094, which is statistically significant at the 1% level.

4.3.4. Placebo Test

This paper may be influenced by additional unobserved or random factors impacting corporate credit risk. This study employs the placebo method for testing to eliminate this possibility. Throughout the sample period, random assignments were conducted for the intelligent transitions to produce the associated pseudo-intelligent transitions, and Equation (7) was utilized to repeat the regression 1000 times. If the coefficients of the pseudo-intelligent transformations remain significant following random assignment, it suggests the presence of unobserved factors influencing the firms’ credit risk. Conversely, the intelligent transformations significantly enhance the firms’ credit risk if they are insignificant.. Figure 2 shows the kernel density plot of the pseudo-intelligent transformation coefficients under credit risk. As can be seen from Figure 2, the coefficients of pseudo-intelligent transformation are mostly concentrated near 0 and are significantly smaller than the coefficients of the actual intelligent transformation (0.046) in Column (2) of Table 3. This indicates that the research conclusion that intelligent transformation significantly increases corporate credit risk is robust and not a result of random regression.

4.4. Robustness Test

4.4.1. Alternative Dependent Variable

The credit rating of the debt issuer is utilized as a proxy for the default distance ($Creditrating$). A higher credit rating correlates with lower credit risk. Based on the People’s Bank of China’s document titled “Credit Rating Elements, Marks and Meanings”, which outlines the classification and requirements for enterprise credit ratings, we utilize quantitative methods from existing literature [6,36] to systematically assign values to the credit ratings of intelligent manufacturing enterprises. The AAA grade is assigned a value of 6, the AA grade (AA+, AA, AA−) is assigned a value of 5, the A grade (A+, A, A−) is assigned a value of 4, the BBB grade (BBB+, BBB, BBB−) is assigned a value of 3, the BB grade (BB+, BB−) is assigned a value of 2, and the B grade (B+, B−) is assigned a value of 1. The regression results presented in Column (1) of

Table 5. Robustness test.

	(1)	(2)	(3)	(4)	(5)
Variable	$Creditrating$	$DD$	$DD$	$DD$	$DD$
${lnIM}_{-}wf$	0.296 ***	0.041 ***	0.044 ***		0.028 ***
	(6.18)	(4.89)	(6.23)		(3.30)
${L.lnIM}_{-}wf$				0.050 ***
				(5.49)
$Growth$	−0.132 *	−0.230 ***	−0.265 ***	−0.311 ***	−0.154 ***
	(−1.78)	(−9.66)	(−8.57)	(−11.87)	(−5.29)
$CFO$	0.374	−0.247 **	−0.416 ***	−0.445 ***	−0.171
	(0.93)	(−2.57)	(−3.82)	(−4.12)	(−1.46)
$ROE$	1.600 ***	0.258 ***	0.203 **	0.171 *	0.075
	(4.51)	(2.68)	(2.20)	(1.79)	(0.63)
$Top10$	−0.884 **	0.273 ***	0.299 ***	0.388 ***	−0.021
	(−2.28)	(3.74)	(5.10)	(4.87)	(−0.28)
$Indep$	1.512 *	0.109	0.128	0.133	0.220
	(1.76)	(0.65)	(1.02)	(0.73)	(1.16)
$Dual$	−0.084	−0.065 ***	−0.061 ***	−0.063 ***	−0.038 *
	(−0.88)	(−3.38)	(−4.50)	(−3.00)	(−1.86)
$PPE$	1.586 ***	0.417 ***	0.402 ***	0.430 ***	0.386 ***
	(3.66)	(5.09)	(5.80)	(4.82)	(4.53)
$Audit$	0.373 **	0.078 **	0.107 ***	0.124 ***	0.077 *
	(2.28)	(2.25)	(3.16)	(3.55)	(1.91)
$ROA$	−3.480 ***	0.107	0.256	0.429 **	0.585 **
	(−4.09)	(0.50)	(1.28)	(1.97)	(2.27)
$GDP$	−0.009	2.079 ***	0.002	0.003	0.003
	(−0.54)	(24.50)	(1.24)	(0.86)	(0.88)
$Constant$	0.243	2.079 ***	2.013 ***	2.054 ***	2.114 ***
	(0.52)	(24.50)	(26.37)	(20.70)	(21.91)
Industry FE	YES	NO	YES	YES	YES
Year FE	YES	NO	YES	YES	YES
Industry*year FE	NO	YES	NO	NO	NO
Province*year FE	NO	YES	NO	NO	NO
Observations	13,797	13,796	13,797	12,264	7665
Adjusted R-squared	0.0829	0.475	0.430	0.333	0.495

Notes: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively (This also applies to the tables below)

indicate that the coefficient of ${lnIM}_{-}wf$ is 0.296, remaining significantly positive at the 1% level.

4.4.2. Higher Dimensional Fixed Effect

To enhance control over the potential effects arising from industry and provincial differences and to improve the model’s accuracy and explanatory power, this paper substitutes the industry and year fixed effects with high-dimensional fixed effects that account for the interaction of industry and year, as well as province and year. This approach allows for better managing potential omitted variables by including more nuanced interaction fixed effects. The findings are presented in Column (2) of Table 5, indicating that the coefficient on ${lnIM}_{-}wf$ is 0.041, which remains significant at the 1% level.

4.4.3. Reclustering

Modify the clustering level from the company to the industry over time. Company-level data may exhibit significant fluctuations and noise. Clustering at the industry level mitigates the influence of individual company outliers, thereby enhancing model robustness and the capacity to generalize findings. Including the time dimension facilitates the analysis of the industry’s dynamic characteristics, enabling the identification of cyclical changes, trend shifts, and the effects of atypical events. This is essential for forecasting future trends and formulating long-term strategies. The results are presented in Column (3) of Table 5, indicating that the coefficient of ${lnIM}_{-}wf$ is 0.044, which is significant at the 1% level.

4.4.4. Lagging One Period Behind

Given that the risk-mitigating effect of intelligent transformation requires time for verification, we introduce a one-period lag for the explanatory variables. The findings are presented in Column (4) of Table 5, where the coefficient of ${lnIM}_{-}wf$ is 0.050, indicating a significant positive relationship at the 1% level. This suggests that the conclusion remains valid when considering the temporal effects of the explanatory variables, thereby minimizing the influence of endogeneity in the time series data or other external factors within the model.

4.4.5. Consideration of Exogenous Shocks

The external environment may influence the relationship between intelligent transformation and corporate credit risk. As 2020 marked the onset of the New Crown epidemic, samples from 2020 to 2023 were excluded, and the regression analysis is conducted anew. Column (5) of Table 5 indicates that the coefficient of ${lnIM}_{-}wf$ is 0.028, remaining significant at the 1% level.

4.5. Mechanism Analysis

4.5.1. Production Effect

According to the study by Li et al. (2018) [37], total factor productivity was selected as the initial proxy variable for the production effect (${TFP}_{-}OP$). The OP method was utilized to assess total factor productivity. The results are presented in Column (1) of

Table 6. Mechanism analysis.

	(1)	(2)	(3)	(4)	(5)	(6)
Variable	${TFP}_{-}OP$	$lnProduction$	$lnIC$	$Market$	$SA$	$ASY$
${lnIM}_{-}wf$	0.085 ***	0.267 ***	0.009 ***	0.005 ***	−0.012 **	−0.064 ***
	(6.15)	(10.28)	(5.17)	(3.59)	(−2.48)	(−7.72)
$Growth$	0.181 ***	0.359 ***	0.017 ***	0.036 ***	−0.034 ***	−0.119 ***
	(7.15)	(7.69)	(3.75)	(10.39)	(−4.56)	(−6.16)
$CFO$	1.259 ***	2.087 ***	0.086 ***	0.269 ***	−0.062	−1.020 ***
	(9.57)	(8.40)	(4.41)	(13.38)	(−1.49)	(−10.36)
$ROE$	0.746 ***	1.638 ***	0.005	−0.082 ***	0.001	−0.263 ***
	(6.43)	(6.60)	(0.25)	(−4.53)	(0.03)	(−3.37)
$Top10$	0.479 ***	1.277 ***	0.040 ***	0.008	−0.231 ***	0.460 ***
	(4.43)	(6.09)	(3.82)	(0.83)	(−6.32)	(7.16)
$Indep$	−0.054	−0.449	−0.003	−0.006	−0.253 ***	−0.259 *
	(−0.22)	(−0.91)	(−0.11)	(−0.25)	(−3.09)	(−1.66)
$Dual$	−0.144 ***	−0.255 ***	0.002	0.011 ***	−0.031 ***	0.031 *
	(−5.12)	(−4.76)	(0.85)	(3.88)	(−3.42)	(1.71)
$PPE$	−0.900 ***	0.786 ***	−0.036 **	−0.032 **	−0.011	0.046
	(−7.35)	(3.32)	(−2.46)	(−2.26)	(−0.28)	(0.61)
$Audit$	0.167 ***	0.461 ***	0.269 ***	0.020 **	−0.010	−0.078 ***
	(3.58)	(5.52)	(23.72)	(2.46)	(−0.62)	(−3.43)
$ROA$	0.001	−0.837	0.268 ***	1.030 ***	0.096	−0.781 ***
	(0.00)	(−1.48)	(5.87)	(21.89)	(1.13)	(−4.08)
$GDP$	0.017 ***	0.006	0.002 ***	0.001	−0.001	−0.001
	(3.59)	(0.70)	(3.93)	(1.59)	(−0.66)	(−0.21)
$Constant$	6.050 ***	19.354 ***	0.140 ***	0.021	4.214 ***	−0.133 *
	(45.18)	(73.58)	(7.57)	(1.48)	(93.58)	(−1.65)
Industry FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
Observations	13,625	13,111	13,646	13,797	13,797	13,791
Adjusted R-squared	0.330	0.510	0.326	0.553	0.243	0.245

Notes: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively (This also applies to the tables below)

. The coefficient of ${lnIM}_{-}wf$ is 0.085, indicating a significant positive relationship at the 1% level. Intelligent transformation enhances the total factor productivity of manufacturing enterprises, thereby reducing corporate credit risk.

According to the study by Zhou et al. (2024) [38], production size was selected as the second proxy variable for the production effect ($lnProduction$). It is measured by adding 1 to take the logarithm of the difference between the total inventory at the end of the period and the total inventory at the beginning of the period plus the income from the main business. The results are presented in Column (2) of Table 6. The coefficient of ${lnIM}_{-}wf$ is 0.267, indicating a significant positive relationship at the 1% level. Intelligent transformation enhances the production scale of manufacturing firms, thereby decreasing corporate credit risk.

4.5.2. Management Effect

According to Li et al. (2021) [39], the level of internal control was selected as the primary proxy variable for assessing management effects ($lnIC$). The measurement involves adding one to the Dibble internal control index and applying the logarithm. The findings are presented in Column (3) of Table 6. The coefficient of ${lnIM}_{-}wf$ is 0.009, indicating a significant positive relationship at the 1% level. Intelligent transformation enhances the internal control level of manufacturing firms, subsequently reducing corporate credit risk.

According to Zhang et al. (2024) [31], market power was selected as the second proxy variable for managerial effects ($Market$). Lerner’s index is utilized for measurement. The measure is defined as: “(Operating Revenue − Operating Costs − Selling Expenses − Administrative Expenses)/Operating Revenue”. The results are presented in Column (4) of Table 6, indicating that the coefficient of ${lnIM}_{-}wf$ is 0.005, which is significantly positive at the 1% level. Intelligent transformation can improve the market dominance of manufacturing enterprises, thereby decreasing corporate credit risk.

4.5.3. Financial Effect

According to D. Chen et al. (2024) [40], financing constraints serve as the primary proxy variable for assessing the financing effect ($SA$). The absolute value of the SA index determines the measurement. A greater absolute value indicates a more severe financing constraint. The results presented in Column (5) of Table 6 indicate that the coefficient of ${lnIM}_{-}wf$ is −0.012, demonstrating a statistically significant negative relationship at the 5% level. Intelligent transformation effectively alleviates financing constraints for manufacturing enterprises and reduces credit risk.

According to the research conducted by Yu et al. (2012) [41], information asymmetry is selected as the second proxy variable for the financing effect ($ASY$). The degree of information asymmetry is assessed using daily frequency trading data, employing the microstructure data of stock trading from Chinese listed companies to develop indicators for liquidity ratio ($L{R}_{i,t}$), illiquidity ratio ($IL{L}_{i,t}$), and yield reversal ($GA{M}_{i,t}$), which are defined as follows:

$$L{R}_{it}=-{\displaystyle \frac{1}{D_{it}}}{\displaystyle \sum _{k=1}^{D_{it}}\sqrt{\frac{V_{it}(k)}{\left|r_{it}(k)\right|}}}$$

(8)

$$IL{L}_{it}={\displaystyle \frac{1}{D}}{\displaystyle \sum _{k=1}^{D_{it}}\sqrt{\frac{\left|r_{it}(k)\right|}{V_{it}(k)}}}$$

(9)

$$GA{M}_{it}=\left|{\gamma }_{it}\right|$$

(10)

$$r_{it}^e(k)={\theta }_{it}+{\phi }_{it}{r}_{it}(k-1)+{\gamma }_{it}{V}_{it}(k-1)sign[r_{it}^e(k-1)]+{\epsilon }_{it}(k)$$

(11)

$$r_{it}^e(k)=r_{it}(k)-r_{mt}(k)$$

(12)

In this context, $r_{mt}(k)$ denotes the stock return for firm $i$ on the $kth$ trading day of year $t$; $V_{it}(k)$ indicates the daily turnover; $D_{it}$ signifies the total number of trading days within the year; and $r_{mt}(k)$ reflects the market return adjusted for market capitalization. This paper extracts the first principal component from the original indicators to identify elements associated with non-information symmetry, facilitating the construction of the information asymmetry indicator. The results are presented in Column (6) of Table 6, indicating that the coefficient of ${lnIM}_{-}wf$ is −0.064, which is significantly negative at the 1% level. Intelligent transformation effectively mitigates information asymmetry in manufacturing enterprises, decreasing credit risk.

4.6. Heterogeneity Analysis

4.6.1. Intelligent Manufacturing Enterprise

The sample enterprises were categorized based on their classification as intelligent manufacturing enterprises. The compilation of intelligent manufacturing enterprises is based on the intelligent manufacturing demonstration projects released by the Ministry of Industry and Information Technology of the People’s Republic of China between 2015 and 2023, totaling 34 projects. If the sample firm qualifies as an intelligent manufacturing enterprise, then ${IM}_{-}IF$ is equal to 1. In the case of a traditional manufacturing enterprise, the value of ${IM}_{-}IF$ is equal to 0.

Table 7. Heterogeneity analysis.

	(1)	(2)	(3)	(4)
	${IM}_{-}IF=0$	${IM}_{-}IF=1$	$Growth=0$	$Growth=1$
Variable	$DD$	$DD$	$DD$	$DD$
${lnIM}_{-}wf$	0.028 ***	0.021	0.055 ***	0.032 ***
	(3.07)	(1.15)	(4.97)	(3.15)
$Growth$	−0.220 ***	−0.356 ***	0.033	−0.226 ***
	(−8.71)	(−6.35)	(0.48)	(−6.94)
$CFO$	−0.321 ***	−0.825 ***	−0.204	−0.320 ***
	(−2.84)	(−4.26)	(−1.41)	(−2.69)
$ROE$	0.142	0.236	0.187 **	0.110
	(1.32)	(1.50)	(2.07)	(0.62)
$Top10$	0.180 **	0.422 ***	0.281 ***	0.303 ***
	(2.19)	(3.14)	(3.15)	(3.68)
$Indep$	0.145	−0.117	0.090	0.257
	(0.78)	(−0.36)	(0.42)	(1.36)
$Dual$	−0.071 ***	−0.030	−0.058 **	−0.067 ***
	(−3.55)	(−0.74)	(−2.30)	(−3.15)
$PPE$	0.322 ***	0.341 **	0.485 ***	0.250 **
	(3.60)	(2.03)	(4.94)	(2.53)
$Audit$	0.098 ***	0.100	0.068 *	0.116 **
	(2.61)	(1.25)	(1.77)	(2.14)
$ROA$	0.205	0.566	0.704 ***	−0.484
	(0.88)	(1.30)	(3.14)	(−1.34)
$GDP$	0.002	0.009	0.006	−0.003
	(0.61)	(1.38)	(1.42)	(−0.71)
$Constant$	2.089 ***	2.204 ***	2.034 ***	2.069 ***
	(19.54)	(12.70)	(18.17)	(18.50)
Observations	9315	4482	6897	6900
Industry FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
Adjusted R-squared	0.434	0.441	0.443	0.429
p-value	0.008 ***		0.023 ***

Notes: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively (This also applies to the tables below)

presents the regression results in Columns (1) and (2). The coefficient for intelligent transformation in the traditional manufacturing industry is 0.028, which is significantly positive at the 1% level; however, the coefficient for intelligent transformation in intelligent manufacturing firms is insignificant. The p-value for the test of difference in coefficients between groups (Fisher’s test) is 0.008, indicating statistical significance at the 1% level.

Unlike intelligent manufacturing enterprises, traditional manufacturing enterprises can substantially reduce corporate credit risk by implementing intelligent transformation. The lower level of intelligence in traditional manufacturing may contribute to a more substantial enhancement of credit risk mitigation effects resulting from intelligent transformation. In intelligent manufacturing enterprises with advanced technological capabilities, the impact of transformation on credit risk mitigation is minimal, as these organizations typically exhibit higher production efficiency, enhanced financial transparency, and a more comprehensive risk management framework. Consequently, credit risk management enhancement through intelligent transformation is relatively constrained for them.

4.6.2. Growth

According to the life cycle theory, high-growth enterprises typically correspond to the growth stage of the corporate life cycle, whereas low-growth enterprises may fall into the maturity or decline stage. Therefore, we classify the sample firms into high-growth and low-growth enterprises based on the mean value of revenue growth rate. Firms with a revenue growth rate above the mean are categorized as high-growth enterprises (growth stage) and assigned a value of 1 ($Growth$ = 1), while those below the mean are classified as low-growth enterprises (maturity or decline stage) and assigned a value of 0 ($Growth$ = 0). The regression results, as shown in Columns (3) and (4) of Table 7, indicate that the coefficient of intelligent transformation for low-growth enterprises is 0.055, which is significantly positive at the 1% level. Similarly, the coefficient for high-growth enterprises is 0.032, also significantly positive at the 1% level. The inter-group coefficient difference test (Fisher test) yields a p-value of 0.023, which is significant at the 1% level. These results confirm that the impact of intelligent transformation on credit risk differs significantly between high-growth and low-growth enterprises. Specifically, compared with high-growth enterprises, intelligent transformation has a more pronounced effect in reducing the credit risk of low-growth enterprises. Moreover, this significant difference remains statistically robust at the 1% level.

High-growth enterprises in the growth stage are typically in a phase of market expansion, experiencing rapid business growth. However, due to their large-scale investments, their operating cash flows may not yet be stable, and their capital expenditures are relatively high, leading to persistent credit risk. Although intelligent transformation enhances production efficiency and market competitiveness, its risk-mitigating effect on these firms remains relatively limited, as they are still in the expansion phase. In contrast, low-growth enterprises in the maturity stage exhibit slower growth and more stable business models, but face challenges such as market saturation and declining profitability. Low-growth enterprises in the decline stage may even encounter operational difficulties, resulting in higher credit risk. At this juncture, intelligent transformation can help firms to optimize resource allocation, improve operational efficiency, and reduce costs, thereby significantly lowering credit risk.

4.6.3. Further Analysis

The intelligent transformation of Chinese manufacturing enterprises relies significantly on the availability of financial resources. Supply chain finance is an innovative financial service model for the real economy, distinguished by low financing interest rates, extended borrowing periods, minimal financing thresholds, and flexible financing schemes. It has emerged as the preferred mode of financial service [42]. Supply chain finance, grounded in actual trade contexts, utilizes the qualifications of core enterprises as a basis for financing. It offers a range of financial services, including financing, settlement, and cash management, aimed at maximizing the overall interests of the supply chain. Supply chain finance enhances capital utilization and efficiency, lowering overall transaction costs and credit risks within the supply chain [43,44]. Supply chain finance is a moderating variable empirically assessing its role in facilitating intelligent transformation to mitigate credit risk. In accordance with the studies by Pan et al. (2020) [43] and Gu et al. (2023) [44], we utilized a dummy variable to assess the implementation of supply chain finance by enterprises (${SCF}_{-}IF$). The measurement procedures are outlined as follows:

Initially, search engines are used. Utilize the Baidu search engine to query “listed company name + supply chain finance/supply chain finance/industry chain finance/industry chain financing” to ascertain if the core enterprise has adopted supply chain finance for upstream or downstream entities. To identify the earliest instance of supply chain finance activation within the observation period, the value of ${SCF}_{-}IF$ is assigned as 1 for the year of activation and the subsequent years, while it is 0 for all other years.

Secondly, text analysis is conducted. This study summarizes findings from various related studies [44,45,46,47,48,49,50,51,52] to compile a list of 108 keywords pertinent to supply chain finance. Python 3 software was utilized to conduct word frequency analysis on corporate annual reports, with the natural logarithm applied after adding 1 to the frequency counts.

Thirdly, cross-validation is employed. The supply chain finance word frequency analysis results and search engine queries were cross-validated. In cases of inconsistency, annual reports were reviewed manually for verification, leading to corrections in the ${SCF}_{-}IF$. The following are the steps involved in cross-validation: Initially, if the news phrase indicates that the company has implemented supply chain finance but the annual report’s word frequency result is zero, the news phrase’s result is dominated, meaning that ${SCF}_{-}IF$ = 1; if the news phrase does not indicate that the company has implemented supply chain finance but the annual report’s word frequency result is zero, the corresponding annual report is manually read to find the keywords for verification. The result of ${SCF}_{-}IF$ changes from 0 to 1 if the verification result demonstrates that the business has actually adopted supply chain finance.

Table 8. Further analysis.

	(1)
Variable	$DD$
${lnIM}_{-}wf$	0.024 **
	(2.31)
${SCF}_{-}IF$	0.134 ***
	(7.59)
$IMSCF$	0.024 *
	(1.83)
$Growth$	−0.266 ***
	(−11.24)
$CFO$	−0.416 ***
	(−4.24)
$ROE$	0.172 *
	(1.91)
$Top10$	0.300 ***
	(4.25)
$Indep$	0.131
	(0.79)
$Dual$	−0.058 ***
	(−3.09)
$PPE$	0.414 ***
	(5.16)
$Audit$	0.113 ***
	(3.38)
$ROA$	0.364 *
	(1.78)
$GDP$	0.003
	(1.05)
$Constant$	1.966 ***
	(21.42)
Observations	13,797
Industry FE	YES
Year FE	YES
Adjusted R-squared	0.434

Notes: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively (This also applies to the tables below)

presents the results, indicating that the coefficient of ${lnIM}_{-}wf$ is 0.024, which is significantly positive at the 5% level. The coefficient of ${SCF}_{-}IF$ is 0.134, indicating a significant positive effect at the 1% level. The coefficient for the interaction term $IMSCF$ is 0.024, indicating a significant positive effect at the 10% level. Supply chain finance can facilitate intelligent transformation, thereby mitigating enterprise credit risk.

5. Conclusions and Discussion

5.1. Conclusions

We utilize panel data from 1533 publicly listed companies in China to experimentally examine how intelligent transformation influences credit risk. The research indicates that clever transformation can mitigate business credit risk. The production, management, and financing effects are the primary mechanisms via which intelligent transformation mitigates credit risk. Analysis of heterogeneity revealed that the intelligent transformation of traditional manufacturing firms has a more significant credit risk mitigation effect than that of intelligent manufacturing firms. The impact of intelligent transformation on reducing credit risk is more significant for low-growth businesses than for high-growth ones. Subsequent analysis shows that supply chain finance can facilitate intelligent transformation to minimize credit risk effectively.

5.2. Theoretical Contributions

On the one hand, studies have already been conducted to examine the connection between enterprise risk and intelligent transformation. Some scholars argue that intelligent transformation can mitigate enterprise risk [12,13,14,15], while others contend that it heightens the likelihood of enterprise violations [21]. Additionally, some researchers propose that the relationship is nonlinear [22,23]. In light of these disagreements, we comprehensively examine the relationship between the two. An empirical analysis of 1533 Chinese-listed manufacturing companies indicates that intelligent transformation can potentially mitigate corporate credit risk. The results contribute new evidence supporting the validity theory of intelligent transformation. Furthermore, we affirm that the production, management, and financing effects are the primary mechanisms by which intelligent transformation mitigates credit risk. This research enhances our understanding of the mechanisms underlying the economic consequences of smart transformation and elucidates the impact of intelligent transformation on credit risk.

On the other hand, this study broadens the understanding of the intelligent transformation mechanism regarding business risk by incorporating the moderating effect of supply chain finance inside the worldwide advocacy of financial services for the real economy. Prior research has predominantly focused on moderating variables related to organizational characteristics [15,21] and external environments [16,53], while insufficient emphasis has been placed on the facilitative impacts of financial components, particularly supply chain finance. While certain studies have examined financial indicators like cost stickiness [17] and financing constraints [54], they predominantly concentrate on the individual firm level and overlook the significance of financial factors from a holistic supply chain perspective. Consequently, by examining the moderating influence of supply chain finance within the framework of intelligent transformation, we enhance the contextual parameters of the smart transformation mechanism and deepen the comprehension of the economic implications of smart transformation through the lens of financial elements.

5.3. Practical Implications

First, the comprehensive intelligent transformation of Chinese manufacturing businesses remains in its first phase. Manufacturing businesses should prioritize the influence of credit risk mitigation resulting from intelligent transformation. Augment investment in intelligent assets to enhance production efficiency and broaden production capacity. Enhance digital oversight and refine internal control mechanisms. Concentrate on market demand and client customization, product innovation, and service innovation to strengthen market supremacy. Integrate digital technology into every segment of the supply chain to diminish information asymmetry and alleviate financial limitations.

Secondly, manufacturing firms must recognize the variations in the credit risk reduction impact of intelligent transformation based on the type of enterprise and its respective life cycle stage. Conventional manufacturing firms and low-growth organizations must expedite their intelligent transformation efforts. Furthermore, the intensity and pace of intelligent transformation must be promptly calibrated in response to the enterprise’s evolving circumstances.

Thirdly, firms should proactively use supply chain finance and fintech while cultivating robust supply chain alliances and fostering relationships with banks and other enterprises. They ought to effectively utilize data asset pledges, future cargo right pledges, patent pledges, and other synergistic strategies to mitigate credit risk arising from disruptions in the capital chain while leveraging financial instruments to facilitate intelligent transformation.

5.4. Research Limitations and Future Directions

Firstly, the research sample comprises Chinese manufacturing listed companies, which exemplify intelligent transformation; however, this specificity may restrict the generalizability of the findings. The sample may be extended to global emerging markets to examine intelligent transformation’s credit risk mitigation effect via sample diversity. Secondly, this study exclusively examines the mechanisms of the production effect, management effect, and financing effect. Future research may investigate the mechanism from alternative perspectives, including product effects, service effects, and technology effects. Third, this study is confined to examining the risk mitigation effects of intelligent transformation, specifically within manufacturing enterprises. Future research may investigate the impact of intelligent transformation on credit risk within supply chains or industrial chains. The connection between intelligent transformation and credit risk is examined from a global standpoint.