The Impact of Exogenous Shocks on the Sustainability of Supply Chain Relationships: Evidence from the COVID-19 Pandemic

Abstract

In recent years, supply chain risks and stability have become a focal point of public attention. However, there is no consensus on how exogenous shocks affect the sustainability of supply chain relationships, nor a clear mechanism of influence. This study uses data from all A-share listed companies in China from Q2 2018 to Q4 2021, constructing a “supplier–quarter–customer” relationship dataset, with the COVID-19 pandemic serving as an exogenous shock. The results show that after experiencing exogenous shocks, the sustainability of supply chain relationships actually strengthens. This suggests that companies may take measures to enhance supply chain stability and maintain existing relationships to ensure sustainability. Channel analysis reveal that trade credit serves as a channel for the impact of exogenous shocks on the sustainability of supply chain relationships, with companies adjusting trade credit supply to downstream customers to maintain and strengthen stability. Additionally, the impact of exogenous shocks on the sustainability of supply chain relationships varies with market concentration, product input heterogeneity, and firms’ ownership type. Therefore, companies should enhance supply chain relationship management, utilize trade credit as a risk buffer, and optimize the supply chain structure to reduce risk transmission and maintain sustainability.

1. Introduction

Nowadays, the global economy is in an unstable state, with frequent exogenous shocks such as financial crises, credit crises, climate change, natural disasters, and unexpected public health events. These exogenous shocks have significant impacts on overall economic development, the operation of supply chains, and even the production of individual firms. Geopolitical events and uncertainties in the economic and social environment can put pressure on corporate supply chain networks, which then transmit these shocks to upstream and downstream companies. For instance, during the pandemic control phase in 2019, due to the possibility of regional lock downs and fluctuations in labor supply, production and sales activities of companies were interrupted, subsequently affecting the production and sales of upstream and downstream companies. In recent years, as the ties between companies in terms of production, division of labor, and sales have become increasingly closer, the strength of supply chain relationships has also grown. If one company encounters issues, it will inevitably affect the normal operation of other companies in the supply chain, thus impacting the stability of the entire supply chain. Therefore, it is necessary to examine the impact of supply chain risks caused by exogenous shocks on supply chains.

Supply chain relationships are a double-edged sword, capable of bringing cooperative mutual benefits and win–win situations but also potentially generating risks and vulnerabilities due to excessive dependence. As emphasized by resource dependence theory, interdependence in the supply chain between firms and their partners can lead to resource sharing and efficiency gains but can also expose firms to external risks [1]. In highly coordinated supply chains, a single party’s mistake or crisis can quickly spread and affect the stability of the entire supply chain. This risk is particularly pronounced in global supply chains, where any disruption in a single link can cause ripples across the entire chain [2]. However, from another perspective, close collaboration in supply chain relationships can enhance the supply chain’s responsiveness and resilience. When faced with exogenous shocks or market fluctuations, cooperation and trust among firms can strengthen the overall response speed and flexibility of the supply chain [3].

The sustainability of supply chain relationships is a direct manifestation of firms’ behavioral choices in the face of risk, but the impact of supply chain risks caused by exogenous shocks on the sustainability of these relationships remains unclear. On the one hand, supply chain relationships facilitate the transmission of exogenous shocks and related risks to firms, resulting in risk interconnection. When companies face severe risks, they cannot remain unaffected. To prevent the transmission of risk, firms are motivated to sever existing supply chain relationships. On the other hand, supply chain relationships can promote resource exchange between firms and their customers, achieving complementary advantages. Maintaining supply chain relationships is a key factor in obtaining competitive advantages and sustaining growth. Therefore, both parties in the supply chain are likely to reach a consensus on cooperation, risk sharing, and mutual benefits. To reap the benefits of continued supply chain relationships, firms are subjectively inclined to offer assistance. This raises the question: How do exogenous shocks affect the sustainability of supply chain relationships, and what mechanisms can enhance the sustainability of these relationships? Addressing these questions can provide insights into how companies can mitigate disruptions in supply chain relationships and contribute to the stability of supply chains.

Based on this, this paper uses data from all A-share listed companies in China from the second quarter of 2018 to the fourth quarter of 2021. A “supplier–quarter–customer” relationship dataset is constructed, with the unexpected COVID-19 pandemic serving as an exogenous shock experiment. First, we explores how the sustainability of supply chain relationships changes and the mechanisms behind these changes after suppliers in the supply chain experience exogenous shocks. The results show that after experiencing exogenous shocks, the sustainability of supply chain relationships actually strengthens. This suggests that companies may take measures to enhance supply chain stability and maintain existing relationships to ensure sustainability. Next, we conduct a channel analysis that reveals that trade credit serves as a channel for the impact of exogenous shocks on the sustainability of the supply chain, with companies adjusting the credit supply to downstream customers to maintain and strengthen stability. Finally, we further analyze the differential impact on the sustainability of supply chain relationships when firms are subjected to exogenous shocks under varying contexts with respect to product market concentration, input product heterogeneity, and differences in ownership nature; the results reveal that the sustainability of supply chain relationships tends to strengthen following an exogenous shock when companies face a more competitive market environment, have lower product heterogeneity, or are state-owned enterprises. Additionally, it is important to note that the reason for using the COVID-19 pandemic as an exogenous shock experiment in this study lies in the following: (1) The occurrence of this public health event is completely unrelated to any financial supply shocks (e.g., financial crises, credit crises, etc.), providing a unique opportunity to analyze the impact of a purely negative exogenous shock on corporate supply chains. Unlike financial shocks, the impact of COVID-19 is more directly reflected in corporate operations, supply chain management, and consumer behavior, rather than being indirectly transmitted through financial markets or credit channels. Therefore, this event allows for a clearer isolation of the independent effect of exogenous shocks on the sustainability of supply chain relationships, avoiding interference from financial factors in the research results. (2) This sudden event is characterized by its large scale and strong abruptness, with its impact covering a wide range of regions and involving almost all industries and types of enterprises. The intensity and universality of this shock provide a research scenario and data support for the empirical analysis in this paper. At the same time, due to the suddenness and unpredictability of public health events, firms are unable to make adequate preparations in advance, which makes the research results more reflective of the real impact of exogenous shocks on enterprises.

The potential marginal contributions of this paper are as follows: First, the study on the impact of exogenous shocks on supply chain relationship sustainability supplements existing research on the impact of exogenous shocks on firms and expands the empirical research on the sustainability of supply chain relationships. Second, although there are studies on the spillover effects of exogenous shocks on supply chains and their impact on supply chain relationships, little is known about the specific mechanisms at play. This paper investigates the channels through which exogenous shocks affect the sustainability of supply chain relationships, thereby extending the scope of the existing research. Third, this study explores the impact of exogenous shocks on the sustainability of firms’ supply chain relationships using the Jaccard similarity coefficient to measure supply chain relationship sustainability, which expands the measurements for supply chain relationships. Fourth, in the channel analysis, this paper examines trade credit as a channel through which exogenous shocks affect the sustainability of supply chain relationship. This perspective contrasts with the prevailing view in most of the literature, which regards trade credit as an alternative financing channel. By doing so, this study not only highlights the critical role of trade credit in supply chains but also expands the existing literature on trade credit. Specifically, trade credit can be used by firms as a tool to mitigate supply chain risks, thereby affecting the stability and efficiency of the entire supply chain.

This paper is organized as follows. Section 2 aims to review and analyze existing research on the impact of exogenous shocks on firms, supply chain relationships and their economic consequences, the spillover effects experienced by firms following such shocks, and the sustainability of supply chain relationships. It also identifies gaps in the current literature. Building on these studies, a theoretical analysis is conducted to derive the hypothesis that the sustainability of supply chain relationships may strengthen after firms experience exogenous shocks. Section 3 describes the sample selection, data sources, the main variables, and the empirical models employed in the study. Section 4 presents the empirical results, including baseline regression findings, robustness checks, channel analysis, and heterogeneity analysis. Section 5 outlines potential directions for future research. Section 6 provides our conclusions and policy suggestions based on this paper.

2. Literature Review and Hypothesis Development

2.1. The Impact of Exogenous Shocks on Firms

The impact of exogenous shocks at the firm level is predominantly negative. Previous literature has found that fluctuations at the firm level often amplify through product networks, becoming risks that affect the entire business cycle [4,5]. The influence of negative exogenous shocks on firms is random, and such shocks can stem from financial disturbances, such as reductions in bank credit supply, or from adverse events like natural disasters or terrorist attacks [6]. Existing studies on the impact of exogenous shocks on firms are primarily focused on three aspects: the firm’s fundamentals, capital market performance, and financing capabilities.

Exogenous shocks can directly disrupt a company’s production by affecting its operational fundamentals. For instance, Barrot and Sauvagnat (2016) found that firms impacted by natural disasters experience a decline in both revenue and sales growth, and their capital market performance also suffers, reflecting investors’ sensitivity to supply chain risks [5]. Pankratz and Schiller (2024) demonstrated that climate change affects both the firms and their related supply chain partner’s performance [7]. Noth and Rehbein (2019) observed that firms affected by floods suffer performance declines [8]. During the COVID-19 pandemic, large-scale shutdowns and production halts significantly undermined the production capabilities of businesses [9]. Disruptions in the supply chain further exacerbated this issue, particularly in industries with a high degree of reliance on globalized supply chains [10]. Moreover, exogenous shocks are frequently accompanied by a significant decline in market demand. Given the rigidity of fixed costs, enterprises face substantial pressure on profit margins [11]. In terms of capital market performance, exogenous shocks primarily affect corporate capital markets by triggering market panic and subsequent sharp declines in stock prices. For example, Baker et al. (2020) demonstrated that during the COVID-19 pandemic, global stock markets experienced an unprecedented rapid bear market cycle, resulting in a significant decline in corporate valuations [12]. Ramelli et al. (2020) also indicated that responses to this crisis varied significantly across sectors, with the technology sector exhibiting relatively strong resilience, while the aviation and tourism industries experienced severe disruptions [13]. The impact of exogenous shocks on a company’s financing capability is mainly attributed to the increased risk of default, which, in turn, leads to higher financing costs. Javadi and Masum (2021) indicated that firms impacted by climate-related negative effects face higher bank lending costs [14]. Berg and Schrader (2012) showed that firms affected by volcanic eruptions encounter increased difficulties in applying for new bank loans [15]. Carletti et al. (2020) found that the credit risk premium in financial markets significantly increased during the crisis, thereby further exacerbating the difficulty for companies to secure financing [16].

The existing literature primarily focuses on the direct negative effects of exogenous shocks on firms themselves, with little consideration of the indirect impacts on their suppliers or customers. However, we explore how the sustainability of supply chain relationships is affected when suppliers are hit by exogenous shocks in this paper.

2.2. Supply Chain Relationships and Firm Economic Outcomes

Research on supplier–customer relationships has consistently been a prominent area of study both domestically and internationally. The existing literature has primarily explored the economic consequences of primary supplier–customer relationships on firms, encompassing various aspects such as financing capabilities, performance, cash holding levels, innovation activities, and performance in the capital markets, etc.

For example, in the realm of financial financing capabilities, studies show that closer supply chain relationships can influence a firm’s bank borrowing costs, credit spreads on secondary market bonds, cash holdings, and tax policies [17,18,19]. Regarding innovation, existing research indicates that strong customer relationships facilitate breakthrough innovations within firms [20,21]. However, the uncertainty of the external environment often poses substantial risks to innovation investments, stable supply chain relationships can mitigate these risks through contractual binding and trust mechanisms [22]. Furthermore, the stability, depth of collaboration, and information-sharing capabilities within supply chain relationships have a significant impact on firm performance. Previous studies indicate that close collaboration in supply chains, such as joint research and development and coordinated planning, can strengthen a firm’s innovation capabilities, indirectly boosting performance [23]. Additionally, information sharing, as a critical component of supply chain relationships, has been shown to significantly improve a firm’s market responsiveness and operational efficiency [24]. However, over-reliance on a single supply chain relationship may increase a firm’s risk exposure, particularly in environments with high external uncertainty [25]. In summary, stable collaborative relationships and information sharing can significantly enhance firm performance, but excessive dependence on a single supply chain may introduce potential risks. Moreover, the stability, transparency, and depth of collaboration in supply chain relationships have a significant impact on a firm’s stock price performance, financing costs, and investor confidence. Stable supply chain relationships can reduce a firm’s operational risks, thereby enhancing investor confidence and boosting stock price performance [26].

Current studies primarily focus on the business ties between upstream and downstream firms within the supply chain. In summary, the literature on the economic consequences of supply chain relationships highlights both the advantages and disadvantages of close relationships, often creating a community in which success and failure are shared.

2.3. Spillover Effects of Exogenous Shocks in Supply Chains

The risk spillover effects of exogenous shocks within supply chains have long been a topic of scholarly attention. When a firm within a supply chain is affected by an exogenous shock, it inevitably impacts other firms in the supply chain, suggesting that supply chain risks spread both upstream and downstream [5,6]. This occurs because the supply chain relationships between firms create interdependencies in production, division of labor, and sales such that even if a firm itself is not directly impacted by the exogenous shock, it may still be indirectly affected through its customers or suppliers, which is known as the supply chain risk spillover effect. Regarding the spillover effects of exogenous shocks within supply chains, existing studies indicate that these effects manifest primarily in three dimensions: between supply chain firms, in external financial capital markets, and across the entire business cycle.

First, the spillover effects among supply chain firms are particularly significant. Previous studies have shown that events such as natural disasters and pandemics often lead to production disruptions, logistical delays, and raw material shortages, thereby affecting downstream firms and causing supply chain interruptions, inventory build-ups, or demand delays [27]. Pankratz and Schiller (2024) [7] conducted a study using global supply chain companies as a sample to investigate the impact of climate change on supply chain networks. Their research revealed that the effects of climate disasters on supplier firms propagate to downstream customer firms. Specifically, both the directly affected supplier firms and the indirectly affected customer firms experience negative impacts on their operational performance [7]. Additionally, since corporate investment and financing activities are closely intertwined with financial capital markets, this means that the impact of exogenous shocks can also permeate into financial capital markets beyond the supply chain within firms. Supply chain disruptions can lead to declining firm profits, increased stock price volatility, and ultimately trigger instability in capital markets. Typically, firms’ stock prices typically react quickly to supply chain issues, with investors experiencing uncertainty regarding these firms’ future prospects [2].

Supply chain disruptions not only affect the production and sales of individual firms but may also trigger industrial restructuring, changes in consumption patterns, and adjustments in economic policies [28]. Similarly, the existing literature reveals that firm-level fluctuations are often amplified through their product networks, becoming risks that affect the entire business cycle and have broader macroeconomic impacts, as noted by Acemoglu et al. (2012) [4], Di Giovanni et al. (2018) [29], and others.

2.4. The Sustainability of Supply Chain Relationships

The sustainability of supply chain relationships is the opposite of supply chain relationship disruption. A small number of studies have found that low-quality internal controls within firms can directly affect the reliability of the information they provide, thereby exacerbating the risks associated with specialized investments in the supply chain [30]. This risk is not only reflected in information asymmetry but may also lead to a decline in the stability of supply chain relationships [31]. Specifically, low-quality internal controls make it difficult for firms to effectively manage critical aspects of the supply chain, increasing distrust among partners and consequently undermining the continuity of supply chain relationships. However, alumni relationships between firms also contribute to the sustainability of supply chain relationships [32]. Regarding the disruption of supply chain relationships, there are risks associated with inefficient investments in the supply chain and opportunistic behavior by members [33], or when supply chain members strategically conceal negative information, other members may choose to interrupt the relationship to mitigate contractual risks [34]. Existing research also reveals that the sustainability of supply chain relationships involves not only trust, commitment, and cooperation but also closely relates to strategic choices in supply chain management, resource dependence, and dynamic capabilities. First, trust is widely recognized as a fundamental basis for the sustainability of supply chain relationships [35]. Trust reduces transaction costs, mitigates information asymmetry, and encourages long-term stable cooperation. Second, effective collaboration can enhance the overall performance of the supply chain, foster innovation, and facilitate information sharing, thereby strengthening mutual dependence and reciprocity between cooperative partners. Studies show that in long-term relationships, firms are generally more willing to invest resources to support the long-term development of the supply chain [36]. Simultaneously, resource dependence theory provides an important perspective for understanding the sustainability of supply chain relationships [1]. According to resource dependence theory, firms in the supply chain continue to maintain collaborative relationships due to their mutual reliance on resources, thus reducing uncertainty in the external environment. This dependency encourages the parties in the supply chain to maintain stable relationships during long-term cooperation.

In summary, both the positive and negative economic outcomes induced by supply chain relationships have received support from certain scholars, indicating the complexity of interactions between firms and customers. However, existing research on how firms respond to exogenous shocks is still insufficient. A limitation of the current literature is its focus solely on the passive impact of exogenous shocks on firms, with little exploration into whether firms possess proactive awareness in addressing such shocks. First, from the perspective of supply chain risk transmission, existing studies have mainly focused on the phenomenon of risk spillover, while how supply chain members respond remains a “black box”. The sustainability of supply chain relationships offers observable behavioral choices, which creates space for this study. Second, there is insufficient attention to the sustainability of supply chain relationships in the existing literature, particularly the lack of exploration into the specific contexts of firms maintaining supply chain relationships. This study aims to clarify whether the sustainability of supply chain relationships changes when firms face exogenous shocks and to identify the specific mechanisms underlying this change.

2.5. Theoretical Analysis and Hypothesis Development

Since the supply chain involves an entire production chain encompassing “suppliers-firms-customers”, firms are not only connected through economic activities such as production and sales but also engage in fund lending through trade credit transactions. This dual connection strengthens the interdependence among firms, leading to shared interests among supply chain members, fostering collaboration, and generating positive economic benefits for both parties. However, supply chain member firms also face risk interdependence. When one firm suffers from an exogenous shock, there is a high likelihood of production disruption, which may lead to supply chain interruptions. In other words, the impact of exogenous shocks can spread through the supply chain to upstream suppliers and downstream customers, thereby creating potential risks to supply chain relationships. This paper examines the decision-making and choices of supply chain relationships when the supply chain is affected by exogenous shocks from the perspectives of risk interdependence and shared interests.

On one hand, supply chain relationships create potential channels for risk contagion among firms. When a firm is affected by a negative exogenous shock, it can also generate negative risks for its downstream customer firms. The main reasons are as follows: (1) When a firm suffers from a negative exogenous shock, it may lead to production interruptions, making it unable to provide sufficient raw materials to downstream customers on time. This can result in resource shortages and production halts for the customers. (2) Negative exogenous shocks can lead to rising production costs for the firm (e.g., increased raw material prices, higher logistics costs, etc.). These costs may be transferred to downstream customers through price transmission mechanisms, leading to higher procurement costs for the customers, reduced profit margins, and increased operational risks. (3) The risk status of suppliers can also affect the financing capabilities of downstream customers. Supply chain relationships serve as an information screening mechanism and are often considered by banks and other fund providers. Negative shocks experienced by suppliers can be seen as unfavorable information for customers, potentially reducing their external financing capabilities and thereby exacerbating their risks. From this, it is evident that supply chain relationships provide a channel for risk contagion. To mitigate the risk of supply chain contagion, firms may reassess their relationships with supply chain partners, potentially leading to the breakdown or adjustment of these partnerships. In such cases, an intuitive outcome is that when a firm is affected by a negative exogenous shock, the sustainability of its supply chain relationships is significantly reduced. Existing research has also revealed such phenomena. For example, exogenous shocks, such as sudden political events or public health crises, can disrupt the existing supply chain system, resulting in long-term production interruptions, demand fluctuations, and market uncertainty, thereby affecting the normal operation of the supply chain [37]. Even a minor shock to a firm may lead to the disruption of existing relationships. If the firm decides to terminate these relationships and seek new customers and suppliers as a response, it could ultimately affect the firm’s comparative advantage and the region’s long-term growth [38].

On the other hand, from the perspective of shared interests, the sustainability of supply chain relationships is of great significance to firms. According to resource dependence theory, there exists an interdependent relationship between firms and other organizations in their environment. To acquire resources such as capital, technology, and raw materials, firms must establish cooperative relationships with external suppliers, customers, governments, and other entities [39]. This theory suggests that firms rely on external resources to sustain their survival and development, and supply chain partners are one of the key sources for firms to obtain resources (e.g., revenue, market information, technical feedback, etc.). Therefore, firms strive to maintain their existing supply chain relationships with partners. For example, maintaining supply chain relationships with customers ensures stable cash flow and order volumes, which helps firms plan long-term development strategies and reduce operational uncertainty. At the same time, long-term and stable supply chain relationships with customers foster synergistic effects, enabling joint development of new products or service optimization, while business continuity also reduces the firm’s sensitivity to market fluctuations. Moreover, the disruption of existing supply chain relationships can impose significant costs on enterprises. When existing supply chain relationships are interrupted, firms need to seek new partners. According to transaction cost theory, the costs incurred during the process of searching for customers, negotiating and signing contracts, and subsequently monitoring contract performance are all classified as transaction costs. This theory emphasizes that the formation and operation of firms are driven by the need to minimize transaction costs [40]. Therefore, to effectively reduce transaction costs and improve operational efficiency, firms tend to establish long-term and stable partnerships with customers. Additionally, when there are substantial specific investments between firms and their existing supply chain partners, such as customizing machinery and equipment for specific customer transactions, the proprietary and irreversible nature of these investments means that losing important customers can result in high sunk costs.

In summary, considering the long-term resource exchange and shared interests that maintaining existing supply chain relationships can bring, as well as the significant sunk costs that may arise from the disruption of these relationships, rational economic firms, even when facing negative exogenous shocks, will strive to maintain their existing supply chain relationships. It can be said that firms possess an active awareness of preserving their current supply chain relationships.

Of course, we also consider whether firms can truly adopt a series of measures to mitigate the impact of supply chain risks caused by exogenous shocks on the sustainability of supply chain relationships? In fact, early research has shown that enterprise supply chains can maintain stability through certain resilience and adaptability mechanisms after encountering exogenous shocks. Resilience theory suggests that supply chains can withstand external shocks through appropriate adjustments. For example, by implementing supplier diversification, inventory strategies, and establishing information-sharing mechanisms, the supply chain’s responsiveness can be enhanced [2]. This ability to respond to external shocks provides a solid foundation for the sustainability of supply chain relationships. Additionally, flexible supply chain configurations, rapid response, and recovery capabilities are considered critical factors in achieving supply chain stability [3]. For instance, in the face of shocks from natural disasters such as floods and earthquakes, some supply chains can quickly stabilize through promptly activated recovery plans, thereby mitigating long-term impacts on operations [41,42]. It can be argued that firms inherently possess the motivation to stabilize their supply chains and will mitigate any long-term damage caused by exogenous shocks [43], thereby maintaining the existing supply chain relationships. According to resource dependence theory, when firms face exogenous shocks, the interdependence between them and their supply chain partners may increase. In response to the uncertainty and resource risks posed by these shocks, firms are often compelled to collaborate more closely with other links in the supply chain to ensure the stable supply of resources and the flow of information [1]. This enhanced collaboration not only helps firms cope with external pressures but also strengthens mutual dependencies, thereby fostering long-term trust and stable relationships. Additionally, the trust-commitment theory posits that when facing exogenous shocks, the commitment between firms and their suppliers or customers intensifies. Shocks force the partners to confront external challenges together, which drives further cooperation in areas such as information sharing, risk-sharing, and resource allocation [35]. Particularly in long-term relationships, firms tend to deepen their commitment to supply chain partners in response to sudden events, working together to navigate crises. This strengthened commitment and trust provides a robust foundation for the sustainability of supply chain relationships. Furthermore, dynamic capabilities theory emphasizes that in a dynamic market environment, enterprises must continually adjust and enhance their capabilities to adapt to external changes [44]. When exogenous shocks occur, firms leverage their dynamic capabilities to optimize resource allocation and enhance supply chain coordination, thereby improving overall adaptability. This adaptive approach, which leads to innovative forms of collaboration, strengthens long-term cooperation within the supply chain, promoting stability and continuity in relationships.

Based on the analysis above, this paper proposes the following hypothesis:

Hypothesis. 

After experiencing exogenous shocks, the sustainability of supply chain relationships is, in fact, strengthened.

3. Research Design

3.1. Sample Selection and Data Sources

The data used in this paper primarily consist of three components: supply chain information (customer-supplier relationship data), the impact of the COVID-19 pandemic in China, and other firm-level control variables. All data are sourced from the CSMAR database. First, drawing on international supply chain studies, such as Barrot and Sauvagnat (2016) [5] and Gofman and Wu (2022) [45], this study constructs a “supplier–year–customer” dataset. Specifically, the paper extracts data on the top five suppliers and customers of Chinese listed companies from the CSMAR database. Given the selective nature of corporate disclosures, for example, Company A may report Company B as one of its top five customers, while Company B does not disclose Company A as one of its top suppliers, we merge the two data tables, and duplicate observations are removed to ensure unique “supplier–year–customer” observations. Moreover, as it may be challenging to obtain financial information for non-listed firms, only listed customers and suppliers are retained in the dataset. Furthermore, this study assumes that the supply chain relationship remains constant within each year, thus constructing a “supplier–quarter–customer” dataset. Second, this study retrieves data on COVID-19 cases in China from the CSMAR database. This dataset includes daily data on regions affected by the pandemic, cumulative confirmed cases, cumulative deaths, and more. Considering that the data related to the COVID-19 pandemic in China begin in 2020, and we found that the first date on which sample companies could be identified as being impacted by the pandemic fell within the first quarter of 2020, we therefore chose a time span of two years before the first impact year (from the first quarter of 2018 to the fourth quarter of 2019) and two years after the first impact year (from the first quarter of 2020 to the fourth quarter of 2021) as the research scope. To mitigate endogeneity issues in the empirical process, this study lags the control variables by one period, resulting in a final sample time span from the second quarter of 2018 to the fourth quarter of 2021. This way corresponds to the multi-period difference-in-differences (DID) empirical research method, as noted in similar literature, such as Jin et al. (2022), Apergis et al. (2023), and Athira et al. (2024) [46,47,48].

Finally, additional firm-level control variable data are obtained from CSMAR, which include firm age, financial leverage, firm size, and so on. During the sample cleaning process, ST, ST*, PT, and delisted firms are excluded. Furthermore, financial firms and companies with missing relevant variables are also removed. Continuous variables are winsorized at the 1% level on both tails to mitigate the influence of extreme values.

3.2. Definition of Variables

3.2.1. The Sustainability of Supply Chain Relationship (Duration)

The sustainability of supply chain relationships and the interruption of supply chain relationships represent two distinct choices for firms. Specifically, using the concept of the Jaccard similarity coefficient, the similarity between the firm’s current and previous customers is used to measure supply chain sustainability. For example, if the firm’s customer set in the previous period is A, and the customer set in the current period is B, the similarity between the two periods’ customers is J(A,B) = |A∩B|/|A∪B|. This study assumes that if the similarity exceeds 50%, the firm’s supply chain relationship is considered sustainable, and a dummy variable, Duration, is set to 1; otherwise, it is set to 0.

3.2.2. Impact of the Shock on the Firm (Affected)

Regarding the determination of whether companies were impacted by the COVID-19 pandemic, we followed approaches similar to those used in studies such as Chebbi et al. (2021) [49], Jin et al. (2022), and Apergis et al. (2023) [46,47], utilizing data on confirmed COVID-19 cases as a measure. Compared to internal corporate financial data or the subjective judgments of management, city-level epidemic data are more readily obtainable and verifiable, thereby reducing measurement errors in the research.

We determined whether a firm was affected by the COVID-19 pandemic by examining the cumulative number of confirmed cases in the city where the firm’s office is located, using percentile thresholds. Specifically, the firm is considered to have been impacted by the pandemic if the cumulative number of confirmed cases in its city exceeded the 25th percentile value during the respective quarter, with a dummy variable taking the value of 1; otherwise, the variable is set to 0. In robustness tests, a 50th percentile value is used as the threshold for assessing whether a firm is affected by the pandemic in a given quarter.

3.2.3. Other Control Variables

To ensure the objectivity and comprehensiveness of this study, a series of control variables are included that may affect the sustainability of supply chain relationships. Following similar studies (Costello, 2020; Garcia-Appendini and Montoriol-Garriga, 2013) [50,51], the control variables at the firm level include firm size (Size), firm age (ListedAge), financial leverage (Lev), return on equity (Roe), firm growth (Growth), ownership structure (SOE), cash flow ratio (CashFlow), and asset tangibility (PPE). The definitions of the specific variables are presented in

Table 1. The definitions of variables.

Variable	Definition
Duration	The dummy variable is set to 1 if the customer similarity between two consecutive periods for a firm exceeds 50%; otherwise, it is set to 0
Affected	The dummy variable equals 1 if the firm experiences a shock in the current period; otherwise, it is set to 0
Size	Natural logarithm of total assets in the end of the period
Lev	Net profit divided by the average of net assets at the beginning and end of the period
ListedAge	The natural logarithm of the firm’s years since listing plus one
Roe	Net profit divided by the average of net assets at the beginning and end of the period
Growth	The difference in operating revenue between two consecutive periods divided by operating revenue from the same period in the previous year
SOE	The dummy variable equals 1 if the firm is state-owned; otherwise, it is set to 0
CashFlow	Net cash flow from operating activities divided by total assets
PPE	Net fixed assets divided by total assets
TC1	(Accounts receivable−accounts payable)/total assets
TC2	(Accounts receivable + notes receivable + prepayments−advance receipts−accounts payable−notes payable)/total assets
TC3	(Accounts receivable + notes receivable−accounts payable−notes payable)/total assets.

.

3.2.4. Trade Credit

This paper posits that trade credit serves as the channel through which exogenous shocks affect the sustainability of supply chain relationships. Trade credit is reflected in financial statements as accounts payable, accounts receivable, notes payable, notes receivable, advance receipts, and advance payments. Specifically, “accounts payable”, “notes payable”, and “advance receipts” represent a “receive first, pay later” procurement or supply model, reflecting the scale of funds occupied by the firm. Conversely, “accounts receivable”, “notes receivable”, and “advance payments” represent a “pay first, receive later” supply or procurement model, reflecting the scale of funds occupied by others from the firm. Following previous research on trade credit (Amberg et al., 2021; Lai et al., 2022) [52,53], the measure for TC is denoted by three different proxies: TC1, TC2, and TC3. The specific measurement methods are presented in Table 1.

3.3. Empirical Methodology

The research question of our paper focuses on the impact of exogenous shocks on the sustainability of supply chain relationships, using the COVID-19 pandemic as an exogenous shock experiment. Given that different firms are affected by exogenous shocks at varying time points, similar to previous studies [53,54,55], we use the difference-in-differences (DID) model as the baseline regression model to test the hypothesis of this study. The rationale for using the multi-period DID model lies in its ability to effectively identify the causal impact of exogenous shock events on the sustainability of supply chain relationships by comparing the treated group (affected firms) with the control group (unaffected firms). At the same time, this method helps to mitigate endogeneity issues to a certain extent. The model is as follows:

$$\begin{array}{ll}{\mathrm{Duration}}_{\mathrm{i},\mathrm{t}}=& {\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{Affected}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_3{\mathrm{Controls}}_{\mathrm{i},\mathrm{t}-1}\\ & + {\mathsf{\eta }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{t}}+{\epsilon }_{\mathrm{i},\mathrm{t}}\end{array}$$

(1)

where the dependent variable, Duration_i,t, represents the sustainability of supply chain relationships. Affected_i,t is the treatment variable in the DID regression model, indicating whether the firm is impacted by an exogenous shock, thereby dividing the sample into a control group and a treatment group. Controls_i,t-1 is the set of control variables for the firm, lagged by one period. The model controls for firm-fixed effects and time-fixed effects. All empirical tests in this study employ robust standard errors.

4. Empirical Results

4.1. Sample Statistics

The summary statistics for the main variables in this study are presented in

Table 2. Summary statistics. This table presents the summary statistics of main variables.

Variable	N	Mean	SD	Min	p50	Max
Duration	9500	0.408	0.492	0	0	1
Affected	9500	0.311	0.463	0	0	1
Size	9500	22.030	1.704	18.580	21.950	26.550
Lev	9500	0.429	0.207	0.059	0.423	0.931
ListedAge	8200	2.290	0.970	0	2.485	3.367
Roe	9000	0.046	0.092	−0.434	0.037	0.322
Growth	8400	0.434	0.805	−0.913	0.481	3.213
SOE	8000	0.431	0.495	0	0	1
CashFlow	9500	0.021	0.068	−0.177	0.016	0.225
PPE	9500	0.206	0.166	0.002	0.167	0.669

. As shown in the table, the mean of the sustainability of supply chain relationships (Duration) is 0.408, which is higher than the median of 0, with a standard deviation of 0.492. This indicates considerable variation in the sustainability of the supply chain relationships of Chinese listed firms. The mean of the key explanatory variable (Affected) is 0.311, also higher than the median, with a standard deviation of 0.463, suggesting that most firms are impacted by the shock, with significant variation across firms. Regarding the control variables, the median and mean of firm size (Size) are quite close, indicating a relatively even distribution of firm sizes across the sample. The age of firms (Age) is concentrated around 9 years, with the longest listing duration reaching 28 years. The mean return on equity (Roe) and firm growth (Growth) are 0.046 and 0.434, respectively, both relatively low, indicating that the profitability of Chinese listed firms during the sample period was not optimistic. The mean of the state-owned firm dummy (SOE) is 0.431, which is greater than the median of 0, suggesting that the majority of firms are state-owned, consistent with the ownership structure of listed firms in China. The mean cash flow ratio (CashFlow) is 0.021, with a maximum value of 0.225 and a minimum of −0.177, indicating that the operational performance of Chinese listed firms was generally poor during the sample period, with overall low operational efficiency. The tangible asset ratio (PPE) has a maximum of 0.669, a minimum of 0.167, and a standard deviation of 0.206, reflecting significant variation in the fixed asset investments of listed firms in China.

4.2. Baseline Results: The Impact of Exogenous Shock on the Sustainability of Supply Chain

To investigate whether the sustainability of supply chain relationships changes following an exogenous shock, the regression is conducted using model (1), and the results are presented in

Table 3. The result of baseline regression.

	(1)	(2)	(3)	(4)
	Duration	Duration	Duration	Duration
Affected	0.0495 **	0.0425 **	0.0526 **	0.0413 *
	(0.0212)	(0.0214)	(0.0247)	(0.0249)
Size	−0.0658 **	−0.0877 ***	−0.0905 **	−0.1069 ***
	(0.0328)	(0.0334)	(0.0374)	(0.0382)
Lev	−0.1051	−0.0603	−0.0288	−0.0014
	(0.0858)	(0.0886)	(0.0981)	(0.1003)
ListedAge	−0.2039 ***	−0.1952 ***	−0.2158 ***	−0.2020 ***
	(0.0610)	(0.0631)	(0.0674)	(0.0695)
Roe	0.0075	0.0560	0.0629	0.1190
	(0.0768)	(0.0763)	(0.0962)	(0.0963)
Growth	−0.0091	−0.0089	−0.0080	−0.0079
	(0.0118)	(0.0118)	(0.0130)	(0.0130)
SOE	−0.0396	−0.0514	−0.0769	−0.0858 *
	(0.0412)	(0.0415)	(0.0487)	(0.0492)
CashFlow	−0.2059 *	−0.2141 *	−0.2986 **	−0.3019 **
	(0.1202)	(0.1216)	(0.1358)	(0.1384)
PPE	0.4722 ***	0.4597 ***	0.2928	0.2710
	(0.1562)	(0.1571)	(0.1789)	(0.1788)
N	6162	6162	5988	5988
firm	YES	YES	YES	YES
quarter	YES	YES	NO	NO
city	NO	YES	NO	YES
industry_quarter	NO	NO	YES	YES
Adj R-Square	0.571	0.574	0.586	0.590

Note: t statistics in parentheses. * p < 0.10; ** p < 0.05; *** p < 0.001.

. Column (1) controls for both firm-fixed effects and time-fixed effects. The coefficient for Affected is significantly positive at the 5% level, indicating that after upstream suppliers are affected by the COVID-19 pandemic, the sustainability of their supply chain relationships actually strengthens. Given that the impact of the COVID-19 Pandemic varies across different regions, column (2) incorporates fixed effects for the city where the firm’s office is located. The results remain consistent with those in column (1). Columns (3) and (4) further control for firm-fixed effects while adding fixed effects for industry–quarter interactions and city-level fixed effects. The coefficient for Affected remains significantly positive at the 5% and 10% levels, respectively, confirming that the conclusion holds and validating the hypothesis of this study.

4.3. Robustness Tests

4.3.1. Parallel Trend Test and Dynamic Effect Analysis

This section conducts a dynamic effect test using a multiple-period difference-in-differences (DID) approach. Drawing upon the study by Li et al. (2016) [56] and Cao et al. (2023) [57], this paper employs an event study method and constructs the following model to analyze the dynamic effects:

$${\mathrm{Duration}}_{\mathrm{i},\mathrm{t}}={\alpha }_0+{\displaystyle \sum _{\mathrm{j}=1}^4{\beta }_{-\mathrm{j}}}{\mathrm{Pre}}_{\mathrm{i},\mathrm{j}}+{\displaystyle \sum _{\mathrm{k}=0}^{1+}{\gamma }_{\mathrm{k}}}{\mathrm{Post}}_{\mathrm{i},\mathrm{k}}+\delta {\mathrm{X}}_{\mathrm{i},\mathrm{t}}+{\eta }_{\mathrm{i}}+{\mu }_{\mathrm{t}}+{\epsilon }_{\mathrm{i},\mathrm{t}}$$

(2)

where Duration_i,j is the explained variable; η_i and μ_t represent the firm-fixed effect and time-fixed effect, respectively; ε_i,j represents the error term; and X_i,t represents a combination of a series of control variable. This paper uses standard robust error to control for possible sequence correlation and heteroscedastic problems. Pre_i,j and Post_i,j represent the j-th period prior to the shock (denoted as −j) and the k-th period after the shock (denoted as k), respectively. To ensure the simplicity of the regression results, only the four periods preceding the shock are considered, and any exogenous shock time after the first period is combined (denoted as 1+) to estimate the average effect of the exogenous shock in the first period and subsequent periods. To avoid multicollinearity issues, the −4 period is excluded as the baseline period. The estimated coefficients (β) for each period prior to the exogenous shock are used to verify whether the control and treatment groups satisfy the parallel trends assumption before the shock, while the estimated coefficients (γ) for each period after the shock describe the distribution of the treatment effects on firms following the exogenous shock.

Figure 1 presents the estimated coefficients for each period, obtained from the TWFE OLS estimation within a 90% confidence interval. As shown in the figure, compared to the reference baseline, there is no significant difference in the sustainability of supply chain relationships between the control and treatment groups prior to the exogenous shock, which satisfies the parallel trends assumption. After the exogenous shock, however, the sustainability of the firms’ supply chain relationships significantly increased, thereby confirming the hypothesis of this paper.

Moreover, recent studies have identified significant issues with the traditional two-way fixed effects (TWFE) methodology. Sun and Abraham (2021) highlight that in event study analyses, the TWFE regression coefficients not only capture the intended own effects but also incorporate causal effects from other periods within the model, as well as the causal effects from the excluded period (i.e., the baseline category); this leads to contamination bias [58]. Consequently, following the approach of Braghieri et al. (2022) [59], this paper adopts the methodology proposed by Sun and Abraham (2021) [58] to plot the estimated coefficients for each period within a 90% confidence interval, as illustrated in Figure 2. The results remain entirely consistent with our hypothesized conditions and align with the empirical findings presented in this study.

4.3.2. Placebo Tests

Given that the baseline regression model in this study employs a multi-period difference-in-differences (DID) approach, a placebo test was conducted to alleviate concerns about potential confounding factors that might influence the empirical results. Specifically, the treatment and control variables were randomly reassigned, and the regression of Model (1) was performed to generate the coefficients of the key explanatory variables. This process was repeated 500 times, resulting in the placebo test outcomes depicted in Figure 3. The findings further corroborate the robustness and validity of the multi-period DID regression results.

Furthermore, drawing on the methodology of Lai et al. (2022) [53], this study advances the shock timing by two periods and constructs a new key explanatory variable, Affected2, to re-run the model (1). As shown in

Table 4. The results of how the sustainability of the supply chain relationship will respond after the shock is advanced for two periods.

	(1)	(2)	(3)	(4)
	Duration	Duration	Duration	Duration
Affected2	0.0271	0.0232	0.0354	0.0284
	(0.0202)	(0.0204)	(0.0228)	(0.0230)
Controls	YES	YES	YES	YES
firm	YES	YES	YES	YES
quarter	YES	YES	NO	NO
city	NO	YES	NO	YES
industry_quarter	NO	NO	YES	YES
N	6162	6162	5988	5988
Adj R-Square	0.571	0.574	0.586	0.590

Note: t statistics in parentheses. The control variables remain consistent with those in the baseline regression model described earlier.

, the coefficient of Affected2 is not statistically significant and positive, which further corroborates the robustness of the multi-period difference-in-differences (DID) results presented earlier.

4.3.3. Replacement of Explained Variables

We alter the threshold of the Jaccard similarity coefficient to redefine the explained variable as Duration_rep, which is a dummy variable. We consider it to be 1 if the customer similarity of a company is 100% between two consecutive periods; otherwise, it is 0. Subsequently, we re-run the regression using model (1), and the regression results are presented in

Table 5. Results of baseline regression after changing the critical value of the Jaccard similarity coefficient.

	(1)	(2)	(3)	(4)
	Duration_rep	Duration_rep	Duration_rep	Duration_rep
Affected	0.0283 *	0.0274 *	0.0422 **	0.0390 **
	(0.0146)	(0.0147)	(0.0178)	(0.0181)
Controls	YES	YES	YES	YES
firm	YES	YES	YES	YES
quarter	YES	YES	NO	NO
city	NO	YES	NO	YES
industry_quarter	NO	NO	YES	YES
N	6162	6162	5988	5988
Adj R-square	0.6881	0.6903	0.7038	0.7071

Note: t statistics in parentheses. * p < 0.10; ** p < 0.05. The control variables remain consistent with those in the baseline regression model described earlier.

below. As can be seen from Table 5, the coefficient of the key explanatory variable, Affected, remains significantly positive across different fixed-effect control groups, which once again validates the conclusions of the baseline regression model and the hypothesis of this paper.

4.3.4. Replacement of Explanatory Variables

This study conducts a quantile test on the cumulative number of confirmed COVID-19 cases in the cities where firms are headquartered, sorted in ascending order. The value at the 50th percentile is selected as the threshold to determine whether a firm is affected by the COVID-19 shock. The variable Affected is redefined to indicate whether a firm was affected by the COVID-19 pandemic during the quarter. Model (1) is then re-run using this redefined variable. The regression results, presented in

Table 6. Results of baseline regression after replacing explanatory variables.

	(1)	(2)	(3)	(4)
	Duration	Duration	Duration	Duration
Affected	0.0926 ***	0.0843 ***	0.1116 ***	0.1017 ***
	(0.0215)	(0.0219)	(0.0249)	(0.0254)
Controls	YES	YES	YES	YES
firm	YES	YES	YES	YES
year	YES	YES	NO	NO
city	NO	YES	NO	YES
industry_quarter	NO	NO	YES	YES
N	6162	6162	5988	5988
Adj R-Square	0.5724	0.5752	0.5877	0.5916

Note: t statistics in parentheses. *** p < 0.001. The control variables remain consistent with those in the baseline regression model described earlier.

, show that the coefficients of the newly defined Affected is statistically significant and positive at the 1% level. This implies that firms experiencing exogenous shocks exhibit enhanced supply chain relationship persistence. These findings are consistent with the baseline regression results, further validating the robustness of the baseline regression and the reliability of the proposed hypothesis.

4.3.5. Heckman Two-Stage Regression

Since regulatory authorities typically encourage listed firms to disclose information about their top five customers and suppliers, the decision to disclose specific details about major suppliers and customers is ultimately at the discretion of the firms. This introduces a potential sample self-selection bias due to voluntary disclosure. To address this issue, this study employs the Heckman two-stage method. In the first stage of the Heckman test, the dependent variable Disclosure is defined as a dummy variable indicating whether a firm discloses specific information about its major suppliers (1 if disclosed; 0 otherwise). A probit model is used for this regression. Subsequently, the inverse Mills ratio obtained from the first-stage regression is incorporated into the baseline model for the second-stage regression. The results, presented in

Table 7. Results of Heckman two-stage regression.

	The First Stage	The Second Stage
	(1)	(2)
	Disclosure	Duration
Affected		0.0834 **
		(0.0381)
Size	−0.1944 **	−0.1448 *
	(0.0794)	(0.0816)
Lev	0.1306	−0.3161 **
	(0.2660)	(0.1444)
ListedAge	0.4433 ***	−0.3410 **
	(0.1387)	(0.1448)
Roe	0.2164	−0.0343
	(0.2635)	(0.0977)
Growth	0.0790 *	−0.0129
	(0.0467)	(0.0315)
SOE	−0.1536	−0.0015
	(0.1296)	(0.0642)
CashFlow	−0.4190	−0.2836
	(0.4009)	(0.1917)
PPE	−0.5971	0.6456 **
		(0.2736)
IMR		−0.0243
		(0.5049)
N	7848	3191
firm	YES	YES
quarter	YES	YES

Note: t statistics in parentheses. * p < 0.10; ** p < 0.05; *** p < 0.001.

, demonstrate that the coefficient of the key explanatory variable Affected remains statistically significant and positive even after adjusting for self-selection bias. This further confirms the robustness of the findings in this study.

4.4. Channel Analysis

Previous studies have indicated that when supply chains are subjected to exogenous shocks, it is inevitable that firms’ fundamental information, such as sales growth and operating income, may be directly negatively affected, thereby generating operational and liquidity risks [5,7]. Intuitively, when firms face liquidity constraints, they are likely to reduce the provision of trade credit to downstream firms. This has been extensively studied in the literature on the determinants of trade credit provision. For example, Luo (2022), using COVID-19 as an exogenous shock, found that firms affected by the pandemic reduced trade credit to their customers [60]. Similarly, Lin and Chou (2015) empirically demonstrated that firms reduced trade credit to downstream firms during the financial crisis due to liquidity constraints [61]. This aligns with the liquidity channel theory of trade credit provision [54]. On the other hand, firms affected by exogenous shocks may experience a decline in bargaining power due to potential negative impacts on their fundamentals, leading to increased pressure to provide more trade credit to downstream customers in a buyer-dominated market. This bargaining power or buyer market theory has also been widely analyzed in the literature [54]. In summary, while there is extensive literature on the determinants of trade credit provision, the conclusions remain complex and multifaceted.

The literature on trade credit is vast, with a significant focus on its origins and its role as an important alternative financing source [61,62]. However, recent studies have uncovered new insights into the role of trade credit, particularly its function in risk-sharing and as a competitive tool. The risk-sharing function suggests that trade credit can serve as a tool for firms to mitigate liquidity shocks. For example, suppliers experiencing liquidity shocks may provide less trade credit to downstream customers with better liquidity [4,50], reflecting trade credit as a vehicle for negative liquidity spillovers. The competitive tool function posits that trade credit can be used to maintain existing customer relationships. Firms motivated to stabilize their supply chains may provide more trade credit to downstream customers even after experiencing negative exogenous shocks, particularly to more important customers [50]. Additionally, firms with stronger external financing capabilities and enhanced market positions may reduce trade credit provision to downstream customers [54]. In essence, trade credit can act both as a buffer against supply chain risks and as a conduit for their transmission.

Existing research primarily focuses on how supply chain relationships influence firms’ use of trade credit, with limited attention paid to how trade credit usage influences the persistence of supply chain relationships. However, it is undeniable that trade credit usage can impact the sustainability of supply chain relationships. For example, Wu et al. (2014) emphasized the importance of trust in the use of trade credit in China [63]. Li et al. (2020) found that Confucian principles of honesty and trustworthiness reduced information collection costs, thereby increasing suppliers’ willingness to provide trade credit [64]. Considering the trust mechanism in trade credit usage, when firms offer more trade credit or extend accounts receivable periods, they are more likely to establish trust-based cooperative relationships with customers, thereby enhancing customer stability. Research shows that trade credit can alleviate financial pressures within supply chains, particularly between suppliers and retailers [65]. In supply chain management, trade credit significantly influences the relationship between suppliers and manufacturers. Studies indicate that firms providing trade credit can reduce liquidity pressure and enhance trust among supply chain partners by extending payment cycles [66].

In short, trade credit helps strengthen cooperative relationships within supply chains. By fostering trust, supply chain members can pursue mutual benefits, focus on collaborative innovation, and avoid short-term gains, thereby establishing more durable and stable supply chain relationships.

To test whether trade credit serves as a channel through which exogenous shocks affect supply chain stability, this study follows Dai et al. (2021) [67] and incorporates interaction terms to construct model (3):

$$\begin{array}{ll}{\mathrm{Duration}}_{\mathrm{i},\mathrm{t}}=& {\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{Affected}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{TC}}_{\mathrm{i},\mathrm{t}}+\mathrm{T}{\mathrm{C}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_3{\mathrm{Controls}}_{\mathrm{i},\mathrm{t}-1}\\ & + {\mathsf{\eta }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{t}}+{\epsilon }_{\mathrm{i},\mathrm{t}}\end{array}$$

(3)

where TC_i,t represents the net trade credit extended by firm i in quarter t. Following previous research on trade credit [52,53], the measure for TC is denoted by three different proxies: TC1, TC2, and TC3. The specific measurement methods are presented in Table 1. The model controls for firm-fixed effects and time-fixed effects. All empirical tests in this study employ robust standard errors.

The regression results, presented in

Table 8. Results of channel analysis.

	(1)	(2)	(3)
	Duration	Duration	Duration
Affected	0.0411 *	−0.0867	0.0585 *
	(0.0233)	(0.0545)	(0.0343)
TC1	0.1671
	(0.1230)
Affected * TC1	0.3029 **
	(0.1322)
TC2		0.2332 *
		(0.1381)
Affected * TC2		0.5965 **
		(0.2556)
TC3			0.2456 *
			(0.1370)
Affected * TC3			0.4264 **
			(0.1677)
Controls	YES	YES	YES
firm	YES	YES	YES
quarter	YES	YES	YES
N	5300	1603	2723
Adj R-Square	0.5768	0.5929	0.5871

Note: t statistics in parentheses. * p < 0.10; ** p < 0.05. The control variables remain consistent with those in the baseline regression model described earlier.

, show that the coefficients of the interaction terms between the key explanatory variable (Affected) and the proxies for net trade credit provision to downstream customers (TC1, TC2, TC3) are statistically significant at the 5% level. This suggests that firms, after experiencing exogenous shocks, consider the high costs of terminating existing supply chain relationships and instead provide more trade credit to downstream customers to maintain the continuity of customer relationships, thereby stabilizing the supply chain.

4.5. Heterogeneity Analysis

This paper posits that when firms face intense competition in the product market, the value of existing customer relationships appears to be higher. If a firm encounters operational difficulties, customers are more likely to terminate their relationships, potentially forcing the affected firm to increase the supply of trade credit to a greater extent to maintain these relationships. Drawing on the research of Ersahin et al. (2024), the Herfindahl-Hirschman Index (HHI) is employed to characterize the competitive environment of the industry in which the firm operates [68]. Firms are classified into the higher HHI group if their product market concentration HHI in the current quarter exceeds the median HHI across all industries; otherwise, they are classified into the lower HHI group. Column (1) and column (2) in

Table 9. Results of heterogeneity analysis.

	Market Concentration		Input Heterogeneity		Ownership Type
	(1) Higher HHI	(2) Lower HHI	(3) Higher Unique	(4) Lower Unique	(5) SOEs	(6) non-SOEs
Affected	0.0067	0.0637 *	−0.0672	0.1448 *	0.1161 ***	−0.0233
	(0.0309)	(0.0327)	(0.0782)	(0.0756)	(0.0322)	(0.0298)
Controls	YES	YES	YES	YES	YES	YES
firm	YES	YES	YES	YES	YES	YES
quarter	YES	YES	YES	YES	YES	YES
city	YES	YES	YES	YES	YES	YES
N	3195	2882	540	563	2781	3393
Adj R-Square	0.5854	0.5983	0.3706	0.3883	0.5289	0.6286

Note: t statistics in parentheses. * p < 0.10; *** p < 0.001. The control variables remain consistent with those in the baseline regression model described earlier.

present the results of the grouped regressions, indicating that the higher the level of competition a firm faces, the stronger its motivation to maintain the continuity of existing supply chain relationships.

According to Barrot and Sauvagnat (2016), when the heterogeneity of a firm’s input products is higher, exogenous shocks are more likely to trigger the transmission of supply chain risks, thereby affecting the stability of the entire supply chain [5]. Following Barrot and Sauvagnat (2016), this paper uses the ratio of R&D expenditure to operating revenue as a proxy for the heterogeneity of input products [5]. Firms are divided into the higher Unique and the lower Unique groups based on whether their ratio exceeds the industry median in the current quarter. Column (3) and column (4) in Table 9 present the results of the grouped regressions, showing that when the heterogeneity of a firm’s input products is lower, providing trade credit to downstream customers after an exogenous shock helps maintain the sustainability of supply chain relationships.

Currently, China’s economy is undergoing a transitional phase, characterized by market uncertainties. State-owned enterprises (SOEs) often possess competitive advantages in resources, making it necessary to distinguish between SOEs and non-SOEs when examining the impact of exogenous shocks on supply chains. This paper categorizes the sample into SOEs and non-SOEs and conducts regressions accordingly. Column (5) and Column (6) in Table 9 present the regression results, revealing that the coefficient of Affected is significantly positive for the SOE sample. This suggests that SOEs are more likely to maintain the sustainability of supply chain relationships after experiencing exogenous shocks, which is attributed to their resource advantages and stronger proactive awareness of mitigating supply chain risks and maintaining supply chain stability.

5. Discussion

This study examines how the persistence of supply chain relationships changes after suppliers experience exogenous shocks, using the COVID-19 pandemic as a natural experiment. The results indicate that firms experiencing exogenous shocks exhibit enhanced persistence in their supply chain relationships. This study investigates the role of trade credit as a channel through which exogenous shocks affect the persistence of supply chain relationships, revealing the underlying mechanisms and shedding light on the strategies firms adopt in response to such shocks.

This study contributes to the literature by providing new insights into the dynamics of supply chain relationships under exogenous shocks and highlighting the strategic use of trade credit as a stabilizing mechanism. The findings have important implications for firms and policymakers in managing supply chain risks and enhancing resilience. Through a detailed analysis of the data, the findings not only validate the theoretical foundations of existing literature but also extend the scope of current research by providing novel insights and perspectives for supply chain management. The results contribute to a deeper understanding of how firms can navigate supply chain risks and enhance relationship continuity in the face of external disruptions.

Overall, due to the limited availability of data on inter-firm supply chain relationships in China, research in this area remains relatively scarce. This study contributes to the literature by expanding the understanding of firm-level supply chain dynamics. To further advance this line of research, there are several avenues for improvement in future studies. First, due to the limited availability of supply chain data in China, where listed firms are only required to disclose information on their top five customers and suppliers, this study is constrained to analyzing relationships involving these top five entities. Future research could expand the sample to include a broader range of supply chain partners. Most similar studies have only considered listed companies as samples. Therefore, if data on non-listed companies are accessible and of reliable quality, attempting to include samples from non-listed companies could enhance the generalizability of the research findings. Second, in the channel analysis, this study examines the total volume of trade credit received by firms from all suppliers. However, due to data limitations, it is not possible to precisely measure the one-to-one flow of trade credit within specific “supplier–quarter–customer” relationship pairs. Refining this aspect represents a promising direction for future research. Third, the empirical analysis uses firms’ headquarters addresses to identify whether they are affected by exogenous shocks. Given that firms’ production facilities may be geographically dispersed, matching headquarters locations with the locations of shocks may introduce some imprecision. Future studies could improve accuracy by considering the actual locations of firms’ operations. These potential improvements highlight opportunities for more nuanced and precise analyses in subsequent research.

6. Conclusions and Policy Suggestions

6.1. Conclusions

This paper uses all A-share listed companies in China from the second quarter of 2018 to the fourth quarter of 2021 as the sample, constructing a “supplier–quarter–customer” relational dataset. It takes the COVID-19 pandemic as an exogenous shock experiment to study the impact of exogenous shocks on the sustainability of supply chain relationships, with a focus on how the supplier’s relationships within the supply chain are affected. For simplification, the sustainability of corporate customer relationships is used as a measure of supply chain relationship sustainability. First, a multi-period difference-in-differences (DID) baseline regression model is employed to examine the impact of exogenous shocks on the sustainability of supply chain relationships. Second, drawing on the mechanism testing method of Dai et al. (2021) [67], an empirical test is conducted to investigate whether the trade credit extended by firms to downstream customers serves as a channel through which exogenous shocks affect the sustainability of supply chain relationships. Furthermore, this study explores the heterogeneity in the effects of exogenous shocks on supply chain relationship sustainability under different product market concentration, product input heterogeneity, and ownership structures. The following conclusions are drawn from the study: (1) After an exogenous shock, the sustainability of supply chain relationships (measured by the sustainability of downstream customer relationships) actually increases. This is attributed to the proactive awareness of firms in maintaining supply chain stability. (2) In response to exogenous shocks, firms maintain the sustainability of their supply chain relationships by providing more trade credit to downstream customers, thereby preserving existing customer relationships. (3) A further analysis reveals that the sustainability of supply chain relationships tends to strengthen when companies face a more competitive market environment, has lower product heterogeneity, or is a state-owned enterprise, following an exogenous shock.

6.2. Policy Suggestions

Based on the results of this paper, we put forward suggestions for supply chain risks management.

6.2.1. For Companies

(1)

With the refinement of market division of labor, the coordination requirements from production to sales at each stage necessitate increasingly close relationships between major suppliers and clients, leading to a scenario where prosperity and adversity are shared. Consequently, enterprises must prioritize supply chain relationship management, not only to meet current production demands but also to effectively mitigate the risks of external shocks transmitted through the supply chain.

(2)

Research indicates that trade credit, as an alternative financing channel, can provide short-term liquidity for firms and serve as a tool to maintain supply chain relationships sustainability. Therefore, to prevent the risk of supply chain disruptions, firms should leverage trade credit as a buffer against supply chain risks, thereby ensuring the stability of the supply chain.

(3)

Firms should optimize their existing supply chain structures by, for example, reducing the concentration of clients and suppliers, acquiring more heterogeneous resources, avoiding excessive dependency, and fostering a cooperative and mutually beneficial environment to prevent the deep propagation of supply chain risks.

(4)

Since small- and medium-sized enterprises (SMEs) or non-state-owned private enterprises often occupy a weaker position in the supply chain, they should actively establish long-term cooperative relationships with upstream and downstream enterprises to secure credit support from upstream partners. At the same time, they should establish mutual assistance mechanisms within the supply chain to jointly cope with external shocks alongside industry peers.

(5)

The results of the heterogeneity analysis in this study indicate that only enterprises with low input product heterogeneity can maintain the sustainability of existing supply chain relationships through trade credit when facing exogenous shocks. Therefore, this study also suggests that while enterprises pursue product uniqueness to enhance competitiveness, they should avoid becoming overly innovative. When the heterogeneity of input products is too high, enterprises may struggle to find suitable substitutes during production and operational disruptions, making it difficult to mitigate losses caused by risks.

6.2.2. For Governments

(1)

Governments and industry associations should introduce relevant policies to encourage firms to stabilize supply chains through trade credit adjustments. For example, tax incentives could be provided to firms that offer trade credit support, such as tax reductions or exemptions.

(2)

Regulatory agencies should actively guide and encourage firms to disclose information about their suppliers and customers while protecting business confidentiality. At the same time, they should continuously monitor the concentration of suppliers and customers to establish a supply chain risk early warning system and response strategies. This aims to prevent the rapid spread of risks caused by excessively high supply chain concentration and avoid potential negative effects.

(3)

Governments should also formulate and implement relevant laws and regulations to create a healthy competitive environment for supply chains. Differentiated management strategies should be adopted based on varying market competition conditions to prevent market monopolies. Additionally, governments should encourage enterprises to pursue win–win cooperation, thereby avoiding unfair competition.