More than a Band-Aid: The Alleviating Effect and Channels of the Industry–Finance Cooperation Pilot Policy on Corporate Financing Constraints

Abstract

As a major policy initiative, China’s Industry–Finance Cooperation (IFC) Pilot Program aims to address the enduring difficulties enterprises face in securing affordable financing. Despite its intent, the policy’s actual efficacy in alleviating corporate financing constraints remains ambiguous. Based on panel data of Chinese A-share listed firms (2011–2023, 19,742 observations), this paper adopts a difference-in-differences (DID) estimator to investigate the effect of China’s IFC Pilot Policy on corporate financing constraints. The results demonstrate that the IFC Pilot Policy significantly alleviates such constraints. Further mechanism and heterogeneity analyses reveal that it operates primarily by reducing earning management, lowering financing costs, and mitigating business risks. This study contributes to the field by establishing the finance-easing effect and risk management mechanism of industry–finance cooperation, offering valuable guidance for policymakers seeking to refine and optimize similar supply-side financial reform measures.

1. Introduction

As China advances toward high-quality economic growth and deepens supply-side structural reforms, the longstanding problem of expensive and inaccessible financing continues to hinder firms in the real economy, restricting their investment, innovation, and long-term development (Brown et al., 2009; Luo et al., 2025; Wen et al., 2021; Yu et al., 2021; T. Zhang, 2023). To address this challenge, the Chinese government is actively exploring new pathways for deep integration between finance and industry. Since 2016, the Ministry of Industry and Information Technology (MIIT), the People’s Bank of China (PBC), the Ministry of Finance (MOF), and other ministries have jointly launched the “Industry–Finance Cooperation (IFC) Pilot Program”. This policy aims to establish a regularized platform for the government, banks, and enterprises to enhance coordination between industrial and financial policies, thereby guiding financial resources to flow more precisely and effectively toward key sectors and weak links in the real economy. This institutional innovation with Chinese characteristics provides a crucial practical platform for alleviating corporate financing constraints at the policy level. However, a crucial and as yet unanswered question remains: What are the actual micro-level effects of these IFC Pilot Policies emphasizing “targeted support”? Can they truly help to alleviate financing constraints faced by real-sector firms? What are their underlying channels of action and the boundaries of their external impacts?

The issues addressed in this study possess profound practical significance. While the existing literature provides a foundational basis for understanding corporate financing frictions, it also reveals critical gaps that this paper aims to fill. Our analytical framework is rooted primarily in Information Asymmetry Theory, which posits that information gaps between insiders and outsiders are the fundamental drivers of external financing premiums and constraints (Almeida et al., 2004; Healy & Palepu, 2001; Lambert et al., 2007). Information is the lifeblood of capital markets; just as third-party disclosures and media coverage enhance transparency and performance (Dagestani et al., 2024; Qing et al., 2024), the IFC pilot program—through its mandatory information exchange platforms and deepened bank–firm ties—is designed to serve as a structural mechanism to bridge this divide.

Building on this theoretical foundation, we construct a specific transmission mechanism grounded in three distinct theoretical stages. First, based on Agency Theory and Signaling Theory, the IFC Pilot Policy reduces information asymmetry, thereby curtailing earnings management. In the presence of information frictions, managers engage in opportunistic earnings manipulation to signal quality or mask distress (Meckling & Jensen, 1976). The IFC Pilot Policy, by acting as an institutionalized monitor, increases the detection probability of such practices. Consequently, the policy reduces the necessity and space for earnings management, leading to a substantial improvement in financial reporting quality. Second, drawing on Real Earnings Management Theory and Risk Transmission Theory, the suppression of earnings management enhances corporate solvency (Z-score; Distance to Default, DD). We argue that earnings management is not merely a cosmetic accounting issue but often involves real distortions (e.g., cutting R&D or overproduction) that erode firm value (Cheng et al., 2025). By forcing firms to abandon these short-sighted behaviors, the IFC Pilot Policy promotes operational efficiency and stabilizes cash flows. This fundamental shift in corporate behavior lowers the objective risk level, manifesting as an increase in the Altman Z-score and Distance to Default (DD). Third, consistent with Risk Pricing Theory and Credit Rationing Theory, the improved risk profile leads to lower financing costs. As the “lemons problem” is alleviated, creditors adjust their required compensation (Stiglitz & Weiss, 1981). A higher probability of default distance directly translates into a lower risk premium demanded by financial institutions. This leads to a direct reduction in debt financing costs, ultimately culminating in the alleviation of financing constraints (proxied by the SA index and FC index). Therefore, we hypothesize that the IFC Pilot Policy acts as an exogenous shock that alleviates financing constraints through the following chain: IFC Pilot Policy → improved information transparency (reduced earnings management) → lower firm risk (higher Z-score/Distance to Default) → reduced financing costs → mitigated financing constraints.

While theories of financial development and government credit enhancement (Borisova & Megginson, 2011) offer optimistic expectations regarding policy efficacy, the empirical evidence remains fragmented. Much of the existing literature focuses on macro-level financial deepening or traditional policy tools, largely overlooking the micro-level causal identification of novel collaborative policies like the IFC Pilot Policy. Specifically, the literature lacks rigorous causal inference regarding whether such policies truly function by improving information environments rather than merely providing subsidies (S. Li et al., 2025; Qiu et al., 2024; S. Zhang et al., 2025). To define the scope of this study and clarify the central research question it aims to address, we present the following central hypothesis:

H₁. 

The IFC Pilot Policy helps alleviate corporate financing constraints.

To address this research gap, this study employs Chinese A-share listed real-economy sector enterprises as its sample. Treating the phased rollout of the IFC Pilot Policy as a quasi-natural experiment, it constructs a multi-period difference-in-differences (DID) model to systematically evaluate the causal effects of this policy on alleviating corporate financing constraints. In research designing, we first employ the mainstream SA Index to represent corporate financing constraints, with FC Index as an alternative dependent variable for robustness testing. In addition, we conducted confounding policy effect analysis, a parallel trend test, a placebo test and Bacon decomposition for additional robustness. According to the benchmark results, the IFC Pilot Policy generated significantly reduced financing constraints for enterprises. Second, based on theoretical logic, we separately tested three potential mechanical channels by a formal mediation model: the level of earnings management firms engage in for financing purposes, business risks, and debt financing costs. Third, we employed a 500-simulation percentile bootstrap approach to validate the chain-mediated pathway linking the IFC Pilot Policy to financing constraints, with earnings management and corporate risk serving as sequential mediators. The results suggest that the IFC Pilot Policy alleviates financing constraints through a sequential mechanism: it curbs earnings management, thereby mitigating business risks and ultimately reducing firms’ debt financing costs. Finally, to enhance the targeted nature of policy insights, we conduct a heterogeneity analysis across two key dimensions—level of marketization and industry characteristics—to examine whether systematic differences exist in policy outcomes between firms in provinces with different marketization, as well as between manufacturing and non-manufacturing enterprises. The results indicate that financing constraints are alleviated more significantly when the enterprises are located in provinces with low levels of marketization or operate in the manufacturing sector.

This study makes a primary theoretical contribution by redefining the role of the IFC Pilot Policy within the Information Asymmetry framework. Moving beyond the conventional view that treats policy interventions merely as capital injections or subsidies, we position the IFC Pilot Policy as an institutionalized “information intermediary” designed to mitigate the “lemons problem” in credit markets (Akerlof, 1978). Drawing on Information Asymmetry Theory (Healy & Palepu, 2001; Lambert et al., 2007), we argue that the core friction in emerging markets is not solely the scarcity of capital but the opacity of corporate information. By establishing mandatory information-sharing platforms, the policy actively reduces the screening and monitoring costs for financial institutions, thereby validating that structured institutional arrangements, rather than just market forces, can effectively bridge the information gap between firms and creditors (Dagestani et al., 2024). This perspective enriches the application of information theory by demonstrating how policy-driven transparency can correct market failures at the micro level.

Our second contribution lies in elucidating a specific micro-transmission channel that links corporate governance to solvency dynamics. While the prior literature often examines earnings management and firm risk as separate outcomes, we integrate them into a sequential causal chain: the IFC Pilot Policy helps to alleviate financing constraints by first suppressing opportunistic earnings management and subsequently enhancing corporate solvency. Grounded in Agency Theory (Meckling & Jensen, 1976) and Risk Management Theory (Nocco & Stulz, 2006), the mechanism analysis hints that the policy-induced reduction in information asymmetry curbs managerial discretion in financial reporting, forcing firms to rely on genuine operational performance. This behavioral shift stabilizes cash flows and improves fundamental credit metrics, as captured by the Altman Z-score and DD-KMV models (Altman, 1968; Bharath & Shumway, 2008). This finding provides a granular explanation of how macro-level policies recalibrate micro-level agency costs and risk-taking capacities (Cheng et al., 2025; Whited & Wu, 2006).

Thirdly, this study contributes to the broader discourse on the “Government–Market” relationship in transitioning economies by offering a synthetic perspective that moves beyond the substitution-versus-distortion dichotomy. While traditional views often depict government intervention as either crowding out private investment or correcting market failures (Cull et al., 2015), our findings align with New Structural Economics (Lin, 2011) to show that the IFC Pilot Policy acts as a structural facilitator of synergy. We provide empirical evidence that government-guided institutional innovation—specifically, the creation of an information infrastructure—stimulates market vitality by lowering transaction costs without displacing market mechanisms. This theoretical extension clarifies how an “active government” and an “efficient market” can achieve organic integration (Y. Xu et al., 2026b), offering a nuanced understanding of institutional complementarity in developing economies.

Finally, at the practical level, this research prompts a strategic pivot in policy design from being a “direct capital provider” to an “information quality supervisor”. Based on the empirical validation that financing constraints are more tightly linked to information opacity (earnings management) than to the absolute scarcity of funds, we recommend that policymakers prioritize the construction of digitized information-sharing ecosystems. Rather than adopting a one-size-fits-all approach, the government should leverage the IFC framework to integrate multi-dimensional data—such as taxation, intellectual property, and supply chain logistics—to cross-verify corporate financial health (Bushman & Smith, 2001). This shift would enable the government to reduce regulatory costs and ensure that resources are allocated efficiently to higher-quality projects, ultimately fostering a resilient financial ecosystem where risk mitigation and transparency drive sustainable economic growth.

In summary, the contributions of this paper are precisely positioned within three streams of the literature. First, in the field of financial development, this study enriches the literature on how state-led institutional design shapes regional financial environments. While the traditional financial development literature heavily emphasizes market-driven deregulation and price mechanisms, our study highlights the role of localized information infrastructure. Our results indicate that government-orchestrated information-sharing platforms can act as an effective institutional remedy to improve capital allocation efficiency, particularly in emerging markets where formal market mechanisms are still developing. Second, in the domain of corporate finance, this paper extends the research on corporate financing constraints and information asymmetry. By identifying earnings management as a critical channel, we provide direct micro-level evidence of how reducing external information friction translates into eased financing constraints for firms. This offers a more nuanced understanding of the interaction between corporate financial behavior and external information environments. Third, with respect to policy evaluation, this paper contributes to the growing body of empirical literature evaluating credit-targeting policies. Utilizing a rigorous DID design, we provide clean, causal evidence of the real-world effects of China’s IFC Pilot Policy.

2. Institutional Background

China’s financial system has long exhibited features of financial repression, under which credit is channeled mainly to state-owned enterprises (SOEs) and infrastructure initiatives, thereby marginalizing private firms and the manufacturing industry (Allen et al., 2005). This structural imbalance has led to severe financing constraints for the real economy, primarily due to acute information asymmetry between financial institutions and industrial enterprises (Liu et al., 2023). Banks often lack sufficient data to assess the creditworthiness of non-SOEs, while high-quality manufacturing firms lack effective channels to signal their value (Ju et al., 2015).

To address these inefficiencies and mitigate the mismatch between financial supply and industrial demand, the Chinese central government initiated the strategy of “Industry–Finance Cooperation”. Unlike general monetary easing, this policy aims to build a specific bridging mechanism to channel financial resources directly into the real economy, thereby reducing transaction costs and agency problems.

A key step in executing this strategy occurred in December 2016, when the MIIT, MOF, and PBC jointly released the Guidance on Enhancing Information Sharing to Advance Industry– Finance Cooperation. Following this, the government launched the IFC Pilot Program to test the efficacy of proximity-based governance in financial resource allocation. The policy was rolled out in phases, creating a quasi-natural experimental setting. Figure 1 shows the timeline of policy implementation and numbers of pilot cities, districts, and counties.

As shown in Figure 1, two rounds of the IFC Pilot Policy have been implemented as of 2023. In the first batch (2016), 37 cities and counties were selected as pilot zones. In the second batch (2020), the scope was expanded to 51 cities or counties to deepen the integration. Of these, 18 of the pilot projects in the second batch overlap with those in the first batch.

The implementation of the IFC Pilot Policy has brought about many changes for Chinese enterprises. Pilot cities have generally established integrated “comprehensive information service platforms for IFC”. The government shares corporate data on tax payments, social security contributions, customs declarations, and other administrative matters with financial institutions, thereby establishing a “white list” system for businesses. This enables financial institutions to create credit profiles based on multidimensional, real-world data, significantly reducing pre-loan due diligence costs and post-loan risks. As a result, the financing constraints faced by businesses have been alleviated to some extent. In addition, precisely because the cost of debt financing for enterprises in pilot cities has fallen significantly and the proportion of long-term loans has risen, financial institutions have shifted their focus from the traditional emphasis on collateral for short-term loans to a focus on corporate growth potential. This has enabled a large number of new materials and high-end equipment manufacturing companies—which lack traditional collateral—to secure significant credit support.

3. Data and Identification Strategy

3.1. Data

To accurately assess the impact of the IFC Pilot Policy, this study focuses on A-share listed firms in the manufacturing sector, green industries, and advanced manufacturing clusters. This selection is based on the fact that the IFC Pilot Policy primarily influences enterprises in these sectors (Qiu et al., 2024; Y. Xu et al., 2026a; S. Zhang et al., 2025). This study retained enterprises in the following industry categories based on Industry Classification Guidelines from the Securities Regulatory Commission of China: Category C (Manufacturing), along with Category I (Information Technology), Category M (Research), Category D (Energy/Utilities), and Category N (Environmental Protection), representing the technology and green sectors. Our sample consists of listed companies covering the period from 2011 to 2023, with the underlying data extracted from the CSMAR database.

The two batches of lists for the IFC Pilot Policy data originate from notifications issued by MIIT1. Correspondingly, companies located in cities affected by the IFC Pilot Policy were assigned to the treatment group, while the remaining companies were assigned to the control group. In addition, the data on the Sci-tech Finance Pilot Policy and Innovative Industrial Clusters (IIC) Policy is primarily derived from two documents publicized by the Ministry of Science and Technology of China about the Advancing of the Cooperation of Science, Technology, and Finance2 and the Innovative Industrial Cluster Pilots3.

In this paper, we construct the SA index as the core proxy and FC index as an alternative proxy for financing constraints using data obtained from the CSMAR database. Unlike the KZ and WW indices, which incorporate potentially endogenous financial variables (e.g., cash flow, leverage, and Tobin’s Q) that are jointly determined with financing constraints, the SA index is exclusively derived from firm size and age. These two variables are highly exogenous and immune to the reverse causality problem, providing a cleaner and more credible measure of financing constraints (Hadlock & Pierce, 2010). Specifically, the SA index is calculated based on company size and age:

$$SA=-0.737\times Size+0.043\times {size}^2-0.04\times Age$$

(1)

The variable $size$ is the natural logarithm of a company’s total assets divided by 1,000,000. Since the original SA indices are all negative, and a smaller SA index indicates greater financing constraints, we have inverted the original SA indices to facilitate the interpretation of the results. A higher adjusted SA index indicates that the enterprise may face more tremendous financing constraints.

Following Hadlock and Pierce (2010), we construct a Financial Constraint (FC) Index based on a Logit model to measure the degree of corporate financing constraints. First, we classify firms into high and low financial constraint groups to generate the dependent variable. We standardize firm size, firm age, and cash dividend payout ratio annually. Listed companies are then ranked in ascending order based on the mean value of these standardized variables. Using the top and bottom tertiles as cutoffs, we define the financial constraint dummy variable (QUFC). Specifically, companies below the 33rd percentile are classified as the high-financial-constraint group (QUFC = 1), while those above the 66th percentile are classified as the low-financial-constraint group (QUFC = 0). Second, using the grouped sample, we estimate the following Logit model:

$$Z_{i,t}={\alpha }_0+{\alpha }_1{ \mathrm{size}}_{i,t}+{\alpha }_2{ \mathrm{lev}}_{i,t}+{\alpha }_3{\left({\displaystyle \frac{ \mathrm{CashDiv} }{ta}}\right)}_{i,t}+{\alpha }_4 M{B}_{i,t}+{\alpha }_5{\left({\displaystyle \frac{NWC}{ta}}\right)}_{i,t}+{\alpha }_6{\left({\displaystyle \frac{EBIT}{ta}}\right)}_{i,t}$$

(2)

The linear predictor $Z_{i,t}$ is subsequently transformed into a probability value ranging from 0 to 1, which serves as our FC index:

$$F{C}_{i,t}=P\left(QUFC=1\mid Z_{i,t}\right)={\displaystyle \frac{e^{Z_{i,t}}}{1+e^{Z_{i,t}}}}$$

(3)

where ${\mathrm{size}}_{i,t}$ is the natural logarithm of total assets; ${\mathrm{lev}}_{i,t}$ is the financial leverage ratio (total liabilities divided by total assets); $\mathrm{CashDiv}$ represents cash dividends scaled by total assets; $M{B}_{i,t}$ is the market-to-book ratio (market value divided by book value); ${\left({\displaystyle \frac{NWC}{ta}}\right)}_{i,t}$ is net working capital (working capital minus cash and cash equivalents minus short-term investments) scaled by total assets; and ${\left({\displaystyle \frac{EBIT}{ta}}\right)}_{i,t}$ is earnings before interest and taxes scaled by total assets. A higher FC value indicates more severe financial constraints faced by the firm.

The variable $DA$ represents the level of earnings management, measured as the absolute value of discretionary accruals scaled by total assets, where a higher $DA$ value indicates more aggressive earnings manipulation. The variable $Cost$ denotes the cost of debt financing, calculated as interest expense divided by the average of short-term and long-term debt. Business risk is captured by $BR$, represented by the Altman Z-Score, which is computed via a linear combination of five financial ratios: (Current Assets − Current Liabilities)/Total Assets, Retained Earnings/Total Assets, EBIT/Total Assets, Market Value of Equity/Total Liabilities, and Revenue/Total Assets, with a lower Z-Score indicating higher bankruptcy risk. Credit risk, denoted as $BDT$, is proxied by the probability of default derived from the Merton/KMV model, which applies the Black–Scholes option pricing framework to solve for asset value and volatility, thereby estimating the default distance. Finally, $Market$ and $MS$ are dummy variables indicating regional marketization and industry affiliation, respectively; $Market$ equals 1 for cities with high-level marketization and 0 otherwise, where $MS$ equals 1 if the firm belongs to the manufacturing sector and 0 otherwise.

All variables described above are obtained from the CSMAR database. In addition, data on city-level characteristics (e.g., city GDP and population density) are drawn from the City Statistical Yearbook, while firm-level controls—including employment, leverage, ROA, ROE, revenue growth, and total assets—are retrieved from the CSMAR database.

Table 1. Descriptive statistics of core variables.

	n	Mean	SD	Min	Max
Panel A: Dependent variables.
$SA$	19,742	3.831	0.233	3.139	4.468
$FC$	19,489	0.521	0.276	$2.00\times {10}^{-5}$	0.992
Panel B: Pre-determined and control variables.
$GDP$	14,670	$1.273\times {10}^8$	$1.195\times {10}^8$	613,500	$5.140\times {10}^8$
$City population density$	14,823	986.6	1034	0.586	4341
$Industrial$	13,345	0.336	0.132	0	1.000
$Labor$	13,642	5120	12,225	28	570,060
$Lev$	13,730	0.388	0.186	0.00752	1.056
$ROA$	13,730	0.0449	0.0649	−0.662	1.285
$ROE$	13,730	0.0681	0.129	−4.857	1.536
$Growth$	14,576	0.170	0.909	−0.902	84.99
$Total assets$	14,030	$1.127\times {10}^{10}$	$2.941\times {10}^{10}$	$3.030\times {10}^8$	$6.795\times {10}^{11}$
Panel C: Mechanical and heterogeneous variables.
$DA$	19,601	0.0484	0.0438	0.000643	0.233
$Cost$	18,177	0.017	0.014	−0.0006	0.059
$BR$	19,653	5.7985	9.2049	−4.655	419.818
$BDT$	19,653	1.361	1.554	−40.688	11.843
$Market$	19,737	0.507	0.500	0	1
$MS$	19,742	0.848	0.359	0	1

Notes: $SA$ and $FC$ denote the corporate financing constraint indices. To account for the heterogeneous economic landscapes across regions, we incorporate the following city-level controls: City GDP, which proxies for regional economic magnitude; Population Density, reflecting the intensity of local economic activities and agglomeration; and the Industrialization Ratio, computed as the contribution of the secondary industry to the overall city GDP. $Labor$ means the number of employees in the company. $Lev$ represents the company’s debt-to-asset ratio. $ROA$ indicates the return on assets, while $ROE$ signifies the return on equity. $Growth$ reflects the company’s revenue growth rate. $DA$ is the absolute value of corporate earnings management levels. $Cost$ represents the cost of debt financing. $BR$ is a measure of business risks. $BDT$ means corporate credit risks. $Market$ is a dummy variable equal to 1 when the province has high-level marketization and 0 otherwise. $MS$ is also a dummy variable equal to 1 when the firm belongs to the manufacturing sector and 0 otherwise.

presents statistics results for all core variables: core dependent variables ($SA$, $FC$), control and pre-determined variables ($GDP$, $City population density$, $Labor$, $Lev$, $ROA$, $ROE$, $Growth$, $Total assets$), mechanical variables ($DA$, $Cost$), and heterogeneous variables ($market$, $MS$). Notably, the average value for $SA$ is 3.831, and the average value for $FC$ is 0.521; the distance from these values to the maximum and minimum is nearly equal, this indicates that the average level of financing constraints faced by firms in the sample is moderate. The mean value of $MS$ is 0.848, it means approximately 84.8% of the enterprises in the sample belong to the manufacturing sector, representing a relatively large proportion.

3.2. Identification Strategy

Using existing datasets, this study constructs the following identification strategy equations to find the influences of the IFC Pilot Policy:

$${SA}_{i,t}={\beta }_0+{\beta }_1{Policy}_{i,t}^{IFC}+{\mu }_i+{\lambda }_t+{\epsilon }_{i,t}$$

(4)

In Equation (4), ${SA}_{i,t}$ means corporate SA index of firm $i$ in year $t$. ${Policy}_{i,t}^{IFC}$ is a representation of IFC Pilot Policy. When it is equal to 1 for enterprise $i$ in year $t$ if its governing local authority (city or county) had enacted the IFC Pilot Policy by that year, and 0 is opposite. ${\beta }_1$ captures the coefficient, or the influence of, the IFC Pilot Policy.

In addition, noting that the IFC Pilot Policy was implemented at two different stages during the period from 2011 to 2023, we can derive Equation (5) based on Equation (4):

$${SA}_{i,t}={\alpha }_0+{\alpha }_1{Policy}_{i,t}^{Phase-1}+{\alpha }_2{Policy}_{i,t}^{Phase-2}+{\mu }_i+{\lambda }_t+{\epsilon }_{i,t}$$

(5)

Among these, ${Policy}_{i,t}^{Phase-1}$ represents the first phase of the IFC Pilot Policy. The first phase of the IFC Pilot Policy designated 37 pilot cities and counties and began in December 2016. Therefore, this study sets the starting date as 2017. In December 2020, the government approved and implemented the second phase of Pilot Policies, namely ${Policy}_{i,t}^{Presult-2}$. It has designated 51 pilot cities and counties. Therefore, this study sets the end date for the first phase of IFC Pilot Policy as 2020 and the start date for the second phase as 2021. Details regarding the establishment of these phases can also be found in Figure 1.

To analyze whether the baseline findings are credent, we incorporated confounding policies into Equation (4). In this study, the confounding policies consists of the following policies: the Sci-tech Finance Pilot Policy and the IIC Policy4. From this, we obtain Equation (6):

$$\begin{array}{c}{SA}_{i,t}={\beta }_0+{\beta }_1{Policy}_{i,t}^{IFC}+{\beta }_2{Policy}_{i,t}^{Sci-tech}+{\beta }_3{Policy}_{i,t}^{IIC}\\ +{\mu }_i+{\lambda }_t+{\epsilon }_{i,t}\end{array}$$

(6)

${Policy}_{i,t}^{Sci-Tech}$ refers to the Sci-tech Pilot Policy, and ${Policy}_{i,t}^{IIC}$ refers to the IIC policy. Once these two policies are implemented, it will be possible to assess ${\beta }_1$ to explore the actual influences of the IFC Pilot Policy.

To verify the similarity between the treatment and control groups before the coming into effect of the IFC Pilot Policy, one parallel trend test is needed. Therefore, Equation (7) is necessary.

$${SA}_{i,t}={\alpha }_0+{\displaystyle \sum _{k=-6}^{6, k\ne 0}}{\mathsf{\Phi }}_k{D}_{i,t}^k+{\mu }_i+{\lambda }_t+{\epsilon }_{i,t}$$

(7)

The sample covers data from 2011 to 2023, with $k$ representing the year relative to 2017. To avoid collinearity, the research lets $D_{i,t}^{-1}$ = 0. When enterprise $i$ is in the treatment group in year $k$, $D_{i,t}^k=$1; otherwise, $D_{i,t}^k$ = 0.

4. Empirical Results and Analysis

4.1. Baseline Results

This section incorporates the basic influences of the IFC Pilot Policy on corporate financing constraints.

Table 2. Implication of the IFC Pilot Policy on corporate financial constraints.

	$\mathit{S}\mathit{A}$
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.013 ***	−0.017 ***	−0.017 ***	−0.018 ***
	(0.005)	(0.006)	(0.005)	(0.005)
Constant	√	√	√	√
Firm FE; Year FE	√	√	√	√
Pre-determined Variables		√
Province Trend; Industry Trend			√
Province–Year; Industry–Year FE				√
Observations	19,742	19,742	19,722	19,714
$R^2$	0.973	0.973	0.976	0.977

Notes: Robust standard errors are in parentheses. *** represents significant levels of 1%, respectively. Pre-determined variables include $City population density$, City $Industrialization$ and City $GDP$. The note “√” means this regression includes the fixed effects or variables.

presents baseline results reflecting the impact of the IFC Pilot Policy. The results include not only the direct DID regression results (Column (1)), but also models that incorporate pre-determined variables (Column (2)), province trends and industry trends (Column (3)), and fixed effects for province–year and industry–year interactions (Column (4)). Firm fixed effects are controlled to absorb all time-invariant, unobserved, firm-specific characteristics, while year fixed effects are controlled to absorb aggregate macroeconomic shocks common to all firms in a given year. Pre-determined variables are constructed by interacting the pre-policy baseline mean values of city population density, city industrialization (share of secondary industry GDP), and city GDP with year fixed effects. From Column (1), it reports a significant regression coefficient of −0.013; it could be calculated that the IFC Pilot Policy reduced the SA index by about 0.34% (calculated as −0.013/3.831). The coefficients for all other columns are significantly negative, which indicate that the IFC Pilot Policy has indeed significantly alleviated corporate financing constraints. The findings provide strong support for the previously mentioned Information Asymmetry Theory and Financial Development Theory.

4.2. Robustness Tests

4.2.1. Alternative Dependent Variables

The SA index is not the only measure of corporate financing constraints (Hadlock & Pierce, 2010). To demonstrate the reliability of the baseline findings, this study introduces the equally widely used FC index. The FC index is a number ranging from 0 to 1, and the higher the value, the more severe the corporate financing constraints. In this study, we utilize the variable $FC$ in place of $SA$, and the results are in

Table 3. Substitution of the other dependent variables.

	$\mathit{F}\mathit{C}$
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.030 ***	−0.030 ***	−0.027 ***	−0.021 **
	(0.008)	(0.009)	(0.008)	(0.009)
Constant	√	√	√	√
Firm FE; Year FE	√	√	√	√
Pre-determined Variables		√
Province Trend; Industry Trend			√
Province–Year; Industry–Year FE				√
Observations	19,487	19,487	19,467	19,459
$R^2$	0.806	0.806	0.822	0.830

Notes: Robust standard errors are in parentheses. *** and ** represent significant levels of 1% and 5%, respectively. Pre-determined variables include $City population density$, City $Industrialization$ and City $GDP$. The note “√” means this regression includes the fixed effects or variables.

. As shown in Column (1), the IFC Pilot Policy caused the corporate FC index to decline by approximately 5.76% (−0.030/0.521). The coefficients for all four columns remain significantly negative, once again confirming the IFC Pilot Policy could mitigate corporate financing constraints.

4.2.2. Confounding Policies

During the same period that the IFC Pilot Policy being implemented, China was also implementing other industrial and financial policies, such as the Sci-tech Finance Pilot Policy and Innovative Industrial Clusters (IIC) Policy. These policies share some similarities with the IFC Pilot Policy (The detailed introduction of these two policies is presented in the Appendix A). Therefore, it is necessary to consider whether potential confounding effects of these policies exist (Callaway & Sant’Anna, 2021). Based on Equation (6), this study treats these two policy variables as control variables.

Table 4. Controlling for confounding policies.

	$\mathit{S}\mathit{A}$
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.013 ***	−0.017 ***	−0.017 ***	−0.018 ***
	(0.005)	(0.006)	(0.005)	(0.005)
${Policy}^{Sci-tech}$	√	√	√	√
${Policy}^{IIC}$	√	√	√	√
Constant	√	√	√	√
Firm FE; Year FE	√	√	√	√
Pre-determined Variables		√
Province Trend; Industry Trend			√
Province–Year; Industry–Year FE				√
Observations	17,178	17,178	17,158	17,134
$R^2$	0.972	0.972	0.976	0.977

Notes: Robust standard errors are in parentheses. *** represents significant levels of 1%, respectively. ${Policy}^{Sci-tech}$ represents the Sci-tech Pilot Policy. ${Policy}^{IIC}$ represents the Innovative Industrial Cluster (IIC) Policy. Pre-determined variables include $City population density$, City $Industrialization$ and City $GDP$. The note “√” means this regression includes the fixed effects or variables.

presents the results obtained by incorporating two confounding policies into the model described in Table 1. The coefficients and their significance levels of ${Policy}^{IFC}$ remained virtually unchanged, confirming the confidence of the results in Table 1, namely that the IFC Pilot Policy has a genuine mitigating effect.

It is worth noting that although the Sci-tech Finance and Innovative Industrial Cluster (IIC) policies overlap temporally with the IFC policy, they do not constitute a process of broad financial liberalization (e.g., interest rate deregulation). Unlike systemic reforms aimed at removing economy-wide distortions, these constitute targeted industrial interventions. Specifically, the Sci-tech Finance policy addresses collateral constraints arising from the “asset-light, high-risk” nature of high-tech firms by promoting the monetization of intellectual property. Concurrently, the IIC policy mitigates information asymmetry for SMEs through cluster-based “soft information” derived from geographical proximity. Since these policies target specific market failures—namely R&D monetization and supply chain collaboration—rather than systematically freeing capital flows, their concurrent implementation reflects a multi-pronged industrial strategy rather than a general trend of financial liberalization.

4.2.3. Parallel Trend Assumption Check

To verify the robustness of the empirical results, this study also examines whether the parallel trend assumption must be satisfied. This study uses 2017, the implementation year of IFC Pilot Policy as the base period and sets the observation window to the six years preceding and six years following the IFC Pilot Policy’s implementation. Based on Equation (5), it analyzes the dynamic influences of the IFC Pilot Policy on financing constraints, with results shown in Figure 2. Figure 2 connects the regression coefficients for different time points with solid lines, while the dashed lines extending above and below each point represent the confidence intervals. In the period preceding the policy’s introduction, the coefficients remained close to zero across all years, and the 95% confidence intervals consistently included zero. This indicates that prior to the implementation of the IFC Pilot Policy, no systematic divergence existed in the trend of greenwashing behavior between treated and control firms. This result strongly supports the parallel trends hypothesis. This confirms the validity of the DID model estimation results. Additionally, in the year of policy implementation, the regression coefficient began to take on a significantly negative value. This demonstrates that following the IFC Pilot Policy, corporate financing constraints were immediately alleviated to some extent. In subsequent years, the coefficient remained significantly negative, demonstrating the IFC Pilot Policy had a sustained mitigating effect on corporate financing constraints.

4.2.4. Placebo Test

To exclude the possibility that the main results arise from unobserved factors or random fluctuations, we also performed a placebo test as an additional robustness check. Specifically, a set of firms was randomly selected from the initial sample, with the sample size matched to that of the actual treatment group, and was then artificially designated as a fictitious “treatment group”. At the same time, the timing of policy implementation is also randomly allocated to construct a pseudo-policy-shock variable. The results of this placebo test are presented in Figure 3. In the figure, the vertical dashed line represents the baseline coefficient, which is set to 0; the horizontal dashed line represents the significance threshold, i.e., p = 0.1. The horizontal axis reports the estimated regression coefficients, while the blue scatter points represent the associated p values. As can be observed, the 500 coefficients generated from the random simulations are approximately normally distributed around zero. The overwhelming majority of these estimates are concentrated near zero, and most corresponding p values exceed 0.1, lying above the horizontal dashed line. These results indicate that the randomly assigned “pseudo-policies” do not produce a significant effect on corporate financing constraints. Hence, the observed effect of the IFC Pilot Policy is unlikely to be attributable to random factors or general macroeconomic fluctuations. Put differently, absent the true effect of IFC Pilot Policy, it would be difficult for random assignment alone to generate estimation results of comparable significance and robustness. Overall, the placebo test provides further support for the credibility of this study’s findings, mitigates concerns regarding omitted-variable bias, and reinforces the causal interpretation that the IFC Pilot Policy contributes to the mitigation of financing constraints.

4.2.5. Bacon Decomposition

Table 5. The results of Bacon decomposition.

	$\mathit{S}\mathit{A}$
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.013 ***	−0.017 ***	−0.013 ***	−0.018 ***
	(0.005)	(0.006)	(0.005)	(0.005)
Firm FE; Year FE	√	√	√	√
Pre-determined Variables		√		√
Control Variables			√
Province–Year; Industry–Year FE				√
Observations	19,742	19,742	19,742	19,742

Notes: Robust standard errors are in parentheses. *** represents significant levels of 1%, respectively. Pre-determined variables include $City population density$, City $Industrialization$ and City $GDP$. Control variables include $Labor$, $Lev$, $ROA$, $ROE$, $Growth$, and $Total assets$. The note “√” means this regression includes the fixed effects or variables.

presents the results of the baseline regression and decomposition of the influences of the IFC Pilot Policy on the financing constraints. In Columns (1) through Column (4), regardless of whether we include pre-determined variables, control variables, and high-dimensional fixed effects for province–year and industry–year interactions, the estimated coefficients of $Polic{y}^{IFC}$ are all significantly negative. It is suggested that the IFC Pilot Policy has significantly alleviated financing constraints. To further verify the robustness of this multi-period DID estimator and eliminate estimation biases that may arise from heterogeneity in processing times, this paper employs the Goodman-Bacon decomposition to decompose the total effect (de Chaisemartin & D’Haultfoeuille, 2020; Goodman-Bacon, 2021).

Figure 4 illustrates the specific composition of the Bacon decomposition. The results show that within the total effect of the multi-period DID, the comparison between the treated and never-treated groups accounts for the vast majority of the weight (over 80%, as indicated by the blue scatter points on the right). The coefficients for this component are centered around −0.025; it supports the baseline finding and constitutes the primary source of the estimated effect. Furthermore, although there are comparisons of the “late treatment group” and the “early treatment group” (represented by the red and green data points in the figure), their weights are extremely low (close to 0) and do not alter the overall coefficient. This indicates that the multi-period DID estimates in this paper are not significantly distorted by “bad weights” or treatment timing heterogeneity, the baseline estimator primarily captures the net effect of policy implementation and exhibits strong robustness.

4.3. Mechanical Analyses

4.3.1. One Potential Channel: Earning Management

Information asymmetry is one of the root causes of financing constraints faced by businesses. There is a natural information barrier between real-economy enterprises and financial institutions; financial institutions find it difficult to gain a thorough understanding of a company’s actual operations and can only extend credit with the utmost caution, thereby creating financing constraints. To verify the conclusions drawn earlier, based on the Information Asymmetry Theory, this study uses corporate earnings management—which represents the degree of information asymmetry—as a mechanical variable for testing (Chen & Hung, 2021; Sun et al., 2024).

Table 6. The mechanical effect of corporate earnings management level.

	$\mathit{D}\mathit{A}$
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.006 ***	−0.009 ***	−0.006 ***	−0.007 ***
	(0.002)	(0.002)	(0.002)	(0.002)
Constant	√	√	√	√
Firm FE; Year FE	√	√	√	√
Pre-determined Variables		√
Province Trend; Industry Trend			√
Province–Year; Industry–Year FE				√
Observations	19,584	19,584	19,564	19,556
$R^2$	0.253	0.253	0.282	0.298

Notes: Robust standard errors are in parentheses. *** represent significant levels of 1%, respectively. Pre-determined variables include $City population density$, City $Industrialization$ and City $GDP$. The note “√” means this regression includes the fixed effects or variables.

shows one of the results from mechanical analysis. It is presented in Column (1) that the IFC Pilot Policy reduced corporate earnings management by approximately 12.40% (−0.006/0.0484). Furthermore, all regression coefficients were significant on the 1% level. It clearly indicates that information asymmetry among firms has been mitigated by the implementation of the IFC Pilot Policy, thereby validating the findings discussed earlier.

4.3.2. One Potential Channel: Business Risks

Corporate managers, policymakers and particularly lenders, place great emphasis on business risk and credit risk (Mhlanga, 2021). Corporate managers and policymakers alike place great emphasis on business risk. If a company’s business and credit risks are high, financial institutions will become wary, leading them to proactively reduce loan amounts or raise interest rates, thereby creating financing constraints. Building on the earlier analysis, this study hypothesizes that the IFC Pilot Policy may lower firms’ operational and credit risks, thereby expanding their access to external financing. To test this hypothesis, this study employed two variables representing corporate risk: business risk ($BR$) and credit risk ($BDT$). Higher BR and BDT values indicate higher operational risk and lower default distance for the company. The results are shown in

Table 7. The mechanical effect of business risks.

	$\mathit{B}\mathit{R}$		$\mathit{B}\mathit{D}\mathit{T}$
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.759 ***	−0.314	−0.174 ***	−0.023
	(0.270)	(0.354)	(0.050)	(0.069)
Firm FE; Year FE	√	√	√	√
Pre-determined Variables		√		√
Observations	19,451	19,451	19,451	19,451
$R^2$	0.608	0.608	0.627	0.627

Notes: Robust standard errors are in parentheses. *** represent significant levels of 1%, respectively. Pre-determined variables include $City population density$,City $Industrialization$ and City $GDP$. The note “√” means this regression includes the fixed effects or variables.

. It can be observed that regardless of whether BR or BDT is used as the mechanical variable, the coefficients are significantly negative. Based on Columns (1) and (3), it can be calculated that the IFC Pilot Policy reduced corporate business risk by approximately 13.27% (−0.759/5.7186) and corporate credit risk by approximately 12.73% (−0.174/1.367). This finding highlights the risk prevention and mitigation effects of the IFC Polit Policy and also hints at the synergistic effects between industry and finance.

4.3.3. One Potential Channel: The Cost of Debt Financing

According to the Pecking Order Theory of finance, debt financing is the external financing channel on which businesses rely most heavily. Based on the theoretical arguments presented above, it can be concluded that IFC Pilot Policy has broken down information barriers and enabled the targeted allocation of credit resources, thereby reducing corporate debt financing costs and helping to alleviate financing constraints. To test the validity of this theory, this study conducted a regression analysis using corporate debt financing costs as a mechanical variable. The cost of debt financing is calculated as follows: the sum of corporate interest expense, fee costs, and other expenses, is divided by total liabilities at the end of the period.

Table 8. Mechanical effect of the cost of debt financing.

	$\mathit{C}\mathit{o}\mathit{s}\mathit{t}$
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.003 ***	−0.003 **	−0.003 ***	−0.003 ***
	(0.001)	(0.001)	(0.001)	(0.001)
Constant	√	√	√	√
Firm FE; Year FE	√	√	√	√
Pre-determined Variables		√
Province Trend; Industry Trend			√
Province–Year; Industry–Year FE				√
Observations	18,083	18,083	18,061	18,053
$R^2$	0.666	0.666	0.691	0.706

Notes: Robust standard errors are in parentheses. *** and ** represent significant levels of 1% and 5%, respectively. Pre-determined variables include $City population density$, City $Industrialization$ and City $GDP$. The note “√” means this regression includes the fixed effects or variables.

presents the results of this mechanical analysis. Column (1) of Table 8 illustrates that the IFC Pilot Policy reduced corporate debt financing costs by approximately 17.6% (−0.003/0.017). This is a significant development that validates the theory discussed earlier. The IFC Pilot Policy has effectively reversed the situation where financial institutions were reluctant to lend, thereby creating opportunities for debt financing for a wide range of enterprises.

4.4. Heterogeneity Tests

To explore how the IFC Pilot Policy affects enterprises differently across contexts, we conduct heterogeneity analyses along two theoretically motivated dimensions: regional marketization level and industry type (manufacturing vs. non-manufacturing). First, we use the dummy variable $Market$, derived from the provincial marketization index (Fan, 2016), to distinguish firms in regions with high versus low marketization. This aligns with the “institutional substitution” theory: government-led interventions (e.g., the IFC Pilot Policy) are expected to exert stronger marginal effects in regions with underdeveloped market mechanisms and severe information asymmetry, where market failures are most acute (Acemoglu et al., 2005). Second, regarding industry heterogeneity, we introduce the dummy variable $MS$ (Manufacturing Sector). Theoretically, this distinction is rooted in the collateral channel hypothesis (Chaney et al., 2012; Gan, 2007). Manufacturing enterprises are typically asset-specific and collateral-intensive, possessing substantial tangible fixed assets. In contrast, non-manufacturing firms (e.g., services, technology) often rely on intangible assets or human capital, which are harder to collateralize. Under credit market frictions, banks are more willing to lend to firms with high collateral values. However, during periods of tightening credit, the financing constraints of manufacturing firms are disproportionately affected due to their reliance on external funding for large-scale, irreversible investments. Therefore, the IFC Pilot Policy, designed to mitigate credit misallocation, is hypothesized to play a more significant role in alleviating the financing constraints of manufacturing firms, which is the very sector where the “collateral gap” is most critical for accessing bank loans.

Table 9. Heterogeneity effect of the level of marketization.

	$\mathit{S}\mathit{A}$
	High Marketization		Low Marketization
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.008	−0.012 **	−0.014 **	−0.019 ***
	(0.006)	(0.005)	(0.006)	(0.007)
Firm FE; Year FE	√	√	√	√
Province–Year; Industry–Year FE		√		√
Observations	16,348	16,318	16,371	16,345
$R^2$	0.975	0.980	0.976	0.980

Notes: Robust standard errors are in parentheses. *** and ** represent significant levels of 1% and 5%, respectively. The note “√” means this regression includes the fixed effects or variables.

presents heterogeneity test results of marketization. Columns (1)–(2) are results when provincial marketization indices are above the median, and Columns (3)–(4) are results when provincial marketization indices are less than the median. The results reveal that the financing-constraint-alleviating effect of the IFC Pilot Policy is more pronounced among firms situated in provinces with lower degrees of marketization. This also demonstrates that the IFC Pilot Policy can effectively address the shortcomings in financial development in regions with low levels of marketization, and that it has a clear inclusive and corrective effect. This serves as further evidence that the IFC Pilot Policy can help alleviate the problems caused by economic imbalances.

Table 10. Heterogeneity effect of manufacturing sector belonging.

	$\mathit{S}\mathit{A}$
	Manufacturing Firm		Non-Manufacturing Firm
	(1)	(2)	(3)	(4)
${Policy}^{IFC}$	−0.013 **	−0.017 ***	−0.009	−0.021 **
	(0.005)	(0.006)	(0.009)	(0.009)
Firm FE; Year FE	√	√	√	√
Province–Year; Industry–Year FE		√		√
Observations	18,463	18,432	14,464	14,433
$R^2$	0.973	0.977	0.974	0.980

Notes: Robust standard errors are in parentheses. *** and ** represent significant levels of 1% and 5%, respectively. The note “√” means this regression includes the fixed effects or variables.

presents the results of another heterogeneity test. For the manufacturing sector subsample, both regression coefficients are significant, and their absolute values are larger. For the non-manufacturing firm subsample, only the coefficient in column (4) is significant. However, the coefficient in column (3) is not significant and has a smaller absolute value. This demonstrates that the IFC Pilot Policy is better suited to the financing characteristics of manufacturing firms and is more targeted in alleviating their financing constraints.

4.5. Synthesizing the Transmission Channels: An Integrated Framework

While the previous sections established the statistical significance of individual mechanisms, the core theoretical contribution lies in synthesizing these channels into a coherent causal narrative. To formally test the joint operation of these variables, we employ a chain mediation effect model (

Table 11. The chain channels.

	$\mathit{D}\mathit{A}$	$\mathit{B}\mathit{D}\mathit{T}$	$\mathit{C}\mathit{o}\mathit{s}\mathit{t}$	$\mathit{C}\mathit{o}\mathit{s}\mathit{t}$
${Policy}^{IFC}$	−0.010 *	0.463 ***	−0.007 ***	−0.007 ***
	(−1.926)	(3.619)	(−6.255)	(−6.061)
$DA$		−0.766 ***		−0.002
		(−3.757)		(−1.150)
$BDT$				−0.001 ***
				(−7.328)
Constant	Yes	Yes	Yes	Yes
Obs	17,686	17,686	17,686	17,686
R-squared	0.000	0.002	0.003	0.006
Adjust R-squared	−0.172	−0.171	−0.170	−0.166
F-Test	F = 3.709	F = 13.823	F = 39.128	F = 31.271
	p = 0.054	p = 0.000	p = 0.000	p = 0.000

Notes: Robust standard errors are in parentheses. *** and * represent significant levels of 1% and 10%, respectively. The note “√” means this regression includes the fixed effects or variables.

). This allows us to decompose the total effect of the IFC Pilot Policy into direct effects and indirect effects transmitted through specific pathways.

The estimation results reveal a sophisticated transmission architecture. Consistent with Information Asymmetry Theory, the IFC Pilot Policy first functions as an “Information Corrector”. We observe a significant indirect effect via the chain: IFC Pilot Policy → earnings management (DA) → business risk (BR) → financing costs (financing constraints).

Table 11 and Figure 5 outline a plausible chain of mechanisms consistent with our empirical patterns. First, the policy is associated with a reduction in earnings management activities (Path coefficient β₁). By constraining opportunistic reporting, the policy likely enhances the credibility of corporate financial disclosures, addressing a core friction in credit markets. Second, this decline in information asymmetry coincides with lower perceived business risk (β₂). As the IFC platform improves banks’ visibility into firm operations, uncertainty regarding solvency may diminish. Third, this shift in risk perception appears to translate into lower debt financing costs (β₃). As creditors update their beliefs about firm stability, the required risk premium tends to decrease, thereby alleviating financing constraints. Although the magnitude may appear modest relative to the direct policy effect, it is economically meaningful because it reflects a structural improvement in the information environment rather than a one-off liquidity injection.

Despite the non-parametric bootstrap method (500 replications), three limitations require caution. First, endogeneity persists. While the initial IFC shock is exogenous, reverse causality or omitted variable bias may affect the causal direction between mediators (e.g., business risk and earnings management). Coefficients should be interpreted as “associations consistent with theory” rather than definitive causal estimates. Second, the serial mediation chain yields small effect sizes, diluted by its multiplicative nature—a common feature in complex systems implying subtle behavioral adjustments, rather than structural shifts. Finally, unobservable factors (e.g., managerial ability, macro shocks) remain a “black box”, potentially biasing the mediated pathway.

In summary, the IFC Pilot Policy alleviates financing constraints through a sequential information-risk-cost correction. Policymakers should recognize that sustainable alleviation requires not just capital injection, but a holistic improvement in the information infrastructure.

5. Conclusions

Financing constraints represent a pervasive and critical challenge for enterprises. The IFC Pilot Policy is widely viewed as a representative policy tool designed to address such constraints. To investigate the actual impact of the IFC Pilot Policy on corporate financing constraints, this study employs a DID approach and a quasi-natural experiment to conduct a comprehensive and detailed empirical analysis of this issue.

The findings of this study indicate that the IFC Pilot Policy makes a rule of mitigation on corporate financing constraints. Furthermore, the mechanism analysis further proves that the IFC Pilot Policy is associated with the reduction in financing constraints by curbing earnings management, lowering the cost of debt financing, and decreasing business risks. Heterogeneity analysis also indicates that this effect is even more pronounced for manufacturing firms located in provinces with low levels of marketization.

The findings offer pragmatic yet context-dependent insights for policymakers. While the IFC Pilot Policy has proven effective, its expansion should be pursued with caution rather than as a blanket mandate; the success of such interventions is highly contingent on local governments’ administrative capacity, fiscal commitment, and regional industrial maturity. Consequently, the government should tailor strategies to mitigate information asymmetry—such as establishing corporate information-sharing platforms and mandatory disclosure systems—to align with local institutional readiness, avoiding a one-size-fits-all approach. Furthermore, policymakers should continue to leverage industry–finance cooperation to build comprehensive risk management systems, encouraging long-term information-sharing mechanisms between financial institutions and the real economy. By aligning policy intensity with local absorptive capacity, the government can effectively reduce corporate risk exposure at the source and ensure the sustainable implementation of financial reforms.

However, this study also has several limitations primarily stemming from sample selection. First, our sample is restricted to Chinese A-share listed firms. While this ensures data availability and quality, it introduces a potential sample bias: listed companies generally possess better access to formal financing channels and lower information asymmetry compared to non-listed entities (Allen et al., 2005; Cull et al., 2015). Consequently, the magnitude of the IFC Pilot Policy’s effect documented here may not be directly generalizable to SMEs or privately held firms, which often face more severe financing constraints and rely more heavily on informal financing. Whether the ‘information intermediary’ mechanism identified in this study holds for these underrepresented groups remains an open question for future research (Yu et al., 2021). Second, while we identify the main effect of the policy, we lack a comprehensive exploration of the moderating variables that might amplify or attenuate this impact (de Chaisemartin & D’Haultfoeuille, 2020; Whited & Wu, 2006). Future studies should investigate factors such as ownership structure, regional legal environments, or firm-level governance quality to provide more granular insights for policymakers (Cao, 2025).