The Impact of Employee Stock Ownership Plans on Capital Structure Decisions: Evidence from China

Abstract

The determination of the capital structure is a critical component of a company’s financial decision-making process. The question of how to optimize a firm’s capital structure to increase its value has been a significant topic of interest within the financial community. The employee stock ownership plan (ESOP) has developed rapidly in China’s capital market over the past decade, providing a suitable context for studying the impact of employee equity incentives on capital structure decisions. This paper employs cross-sectional ordinary least squares regression models and unbalanced panel fixed effect models to investigate the impact of employee stock ownership plans (ESOPs) on firms’ capital structure decisions. The analysis is conducted on a sample of Chinese A-share listed companies on the Shanghai and Shenzhen Stock Exchanges. The research considers both static capital structure choice and dynamic capital structure adjustment. We find that the implementation of an ESOP reduces the level of corporate debt and accelerates the dynamic adjustment of capital structure, suggesting that employee equity incentives play a role in optimizing firms’ capital structure decisions. We also find that the impact of ESOPs on the dynamic adjustment of capital structure is asymmetric. Specifically, the implementation of ESOPs markedly accelerates the downward adjustment of capital structure, yet has no impact on the upward adjustment of capital structure. Further analysis demonstrates that the impact of ESOPs on capital structure decisions is contingent upon the macroeconomic environment, industry characteristics, corporate governance, and ESOP contract designs. First, the optimization of ESOPs on capital structure decisions is more pronounced in an economic boom environment, in a poor market climate, or in competitive industries. Second, the reduction effect of ESOPs on corporate debt is more pronounced in non-state-owned companies, high-tech companies and those with lower ownership concentration. In contrast, the acceleration effect of ESOPs on capital structure adjustment is more pronounced in state-owned companies, non-high-tech companies and those with higher ownership concentration. Ultimately, ESOPs financed by loans from a firm’s major shareholders—or with a longer lock-up period, smaller shareholding size or executive subscription ratio—demonstrate a more pronounced optimization effect on capital structure decisions. This paper not only contributes to the existing literature on the relationship between equity incentives and capital structure decisions, but also provides guidance for listed companies on the reasonable design of their ESOPs and the optimization of their capital structure decisions.

1. Introduction

The decision regarding the capital structure of a company represents a significant aspect of the financial decision-making process. An investigation into the factors affecting a firm’s capital structure enables the identification of the determinants of capital structure decisions and the formulation of strategies for optimizing capital structure and enhancing the firm’s value. In accordance with the static and dynamic trade-off theories, an enterprise is presumed to possess an optimal capital structure, which it is then expected to adjust dynamically towards the target capital structure [1]. Previous studies examining the factors influencing capital structure decisions from both macro and micro perspectives have identified economic conditions, marketization degree, corporate characteristics, and governance structure as key factors affecting firms’ capital structure decisions [2,3,4,5]. In addition, as key decision-makers within enterprises, senior executives have the final decision-making power regarding capital structure. Consequently, the willingness of senior executives to adjust capital structure also directly affects the dynamic adjustment of capital structure [1].

In recent years, several scholars have paid attention to the impact of executive equity incentives on capital structure decisions. Their findings indicate that executive equity incentives and the design of their contract terms exert a considerable influence on dynamic capital structure adjustment [6,7,8]. However, these scholars fail to consider the role of employee equity incentives, limiting their analysis to the impact of executive equity incentives. As a significant constituent of the enterprise, the conduct of ordinary employees directly affects the enterprise’s development. However, they typically lack the right to participate in the decision-making processes of their companies. An employee stock ownership plan (ESOP) represents one avenue through which equity incentives may be provided, affording ordinary employees the opportunity to participate in the decision-making processes of their respective organizations. This can serve to enhance employee engagement and vigilance in monitoring executive conduct, consequently influencing the trajectory of organizational decision-making. Therefore, this paper adopts a capital structure decision lens to examine the potential governance role of employee equity incentives within the corporate decision-making process.

In accordance with the tenets of agency theory, the implementation of ESOP can serve to align the interests of employees with those of shareholders. This, in turn, can foster a culture of active employee engagement and proactive supervision of management, thereby mitigating the agency problem inherent in the separation of ownership and control [8]. Consequently, it can enhance the management’s willingness to adjust the capital structure. Furthermore, the implementation of an ESOP can convey favorable indications to the capital market, which can assist in reducing the asymmetry between internal and external information and alleviate corporate financing constraints [9], thus reducing the cost of adjusting capital structure for management. Consequently, theoretically, the implementation of an ESOP can optimize corporate capital structure by enhancing management’s willingness to adjust the capital structure, and reducing the cost of adjusting capital structure for management, thereby accelerating the dynamic adjustment of capital structure.

On 20 June 2014, the China Securities Regulatory Commission (CSRC) published the “Guiding Opinions on Pilot Employee Stock Ownership Plan for Listed Companies” (hereinafter referred to as the “Guiding Opinions”). This document prompted a surge in the adoption of ESOPs by listed companies in China. However, given the relatively brief history of the development of the employee stock ownership system in China and the absence of supporting legislation, there are numerous issues pertaining to the design and implementation of ESOPs in practice. For instance, some companies that released draft ESOPs ultimately ceased implementation due to low employee participation [10]. Conversely, other companies that successfully implemented ESOPs encountered challenges such as excessively lengthy lock-up periods and difficulties when employees wanted to withdraw their shares, which prompted the company to announce the premature termination of the ESOP implementation. The question thus arises as to whether ESOPs recently implemented in China can play a role in corporate governance, thereby optimizing the capital structure decisions of enterprises. This paper attempts to answer this question.

This study employs a sample comprising A-share listed companies in China’s Shanghai and Shenzhen Stock Exchanges, spanning the period from 2014 to 2021. First, the firm’s capital structure is measured using the gearing ratio. Subsequently, the impact of ESOPs on the static capital structure choice is estimated using cross-sectional ordinary least squares regression models. Second, the efficiency of dynamic capital structure adjustment is evaluated based on the speed of capital structure adjustment. The impact of ESOPs on dynamic capital structure adjustment is then estimated using unbalanced panel fixed effect models. The results demonstrate that the implementation of ESOPs not only reduces the company’s debt level, but also accelerates the speed of capital structure adjustment. This suggests that employee equity incentives contribute to the optimization of the company’s capital structure decisions. Further analysis reveals an asymmetry in the effect of ESOPs on the dynamic adjustment of capital structure. Specifically, the implementation of ESOPs significantly accelerates the speed of downward adjustment of capital structure, but has an insignificant effect on the speed of upward adjustment of capital structure. This indirectly validates the finding that the implementation of ESOPs reduces the level of company leverage.

Moreover, this study reveals that the impact of ESOPs on capital structure decisions is contingent upon the internal and external environment of enterprises, as well as the design of ESOP contracts. Regarding the impact of the macroeconomic environment, the optimization effect of ESOPs on capital structure is more pronounced in a favorable economic environment or in an unfavorable market climate. In terms of the impact of industry characteristics, the reduction effect of ESOPs on leverage is more pronounced in competitive industries and high-tech industries, whereas the acceleration effect of ESOPs on capital structure adjustment is more pronounced in competitive industries and non-high-tech industries. Regarding the impact of corporate governance, the reduction effect of ESOPs on leverage is more pronounced in non-state-owned companies and those with dispersed equity, whereas the acceleration effect of ESOPs on capital structure adjustment is more pronounced in state-owned companies and those with concentrated equity. In terms of the impact of ESOP contract design, it can be observed that ESOPs financed by loans from the firm’s major shareholders—or with a longer lock-up period, or a smaller shareholding size or executive subscription ratio—have a stronger optimization effect on capital structure decisions.

This paper makes three contributions to the existing literature. First, this study contributes to the existing literature on the economic consequences of ESOPs by investigating the impact of ESOPs on a firm’s capital structure decisions. The extant literature focuses on the impact of ESOPs on shareholder wealth, firm performance, or firm innovation. This paper extends the economic consequences of ESOP implementation to the field of capital structure decisions, thereby enriching the empirical evidence on ESOPs’ incentive effects. Second, the existing literature on the determinants of capital structure decisions is extended by considering the role of employee equity incentives. A substantial body of research has been conducted on the factors influencing capital structure decisions. Some scholars have also analyzed the impact of equity incentives on capital structure decisions. However, their focus has been on the relationship between executive equity incentives and capital structure decisions, with few studies analyzing the impact of employee equity incentives on capital structure decisions. By considering the role of employees in the decision-making process of firms, this paper finds that the implementation of ESOPs can help to optimize the capital structure of enterprises, thereby providing direct empirical evidence on the governance role of ESOPs. Third, the employee stock ownership system provides significant support for enterprise development, capital market construction and the transformation of the overall economic system in emerging capital markets. The findings of this study can inform the design of ESOP contracts in other emerging market economies and facilitate the implementation of ESOPs.

The remainder of the paper is structured as follows: Section 2 introduces the institutional background of ESOPs in China, and reviews the previous relevant literature; Section 3 develops the hypotheses of this paper; Section 4 describes the samples, variables, and models investigated in this paper; Section 5 reports the empirical results; Section 6 provides the conclusions and discussion.

2. Institutional Background and Literature Review

2.1. Institutional Background

The employee stock ownership system has its roots in the two-factor economic theory proposed by American economist Louis Kelso in 1958 [11]. The two-factor economic theory posits that wealth is created by the two key factors of labor and capital. Furthermore, it maintains that the contribution of capital to production remains greater than that of labor throughout the process of industrialization. This phenomenon ultimately gives rise to significant injustices in social distribution. Consequently, Kelso proposed the strategic combination of capital and labor through the encouragement of employee stock ownership, thereby establishing the preliminary form of ESOP. ESOP was officially recognized by the United States, and developed and promoted rapidly over the subsequent four decades; largely due to the government’s provision of financial incentives, tax preferences and other support systems for its advancement.

China’s employee stock ownership system, which commenced in the mid-1980s, represents a distinctive approach to equity system arrangements. It has been continuously developed in the context of the ongoing reform of state-owned enterprises’ shareholding structures. At that time, China was in the pilot phase of transforming the operational mechanisms of enterprises and implementing shareholding reform, during which some state-owned enterprises and collective enterprises made several attempts at shareholding. In the 1990s, influenced by the employee stock ownership system in Western developed countries, a bold attempt at employee stock ownership was carried out in all parts of China, resulting in the flourishing of internal employee stock ownership. However, due to the absence of owners of state-owned enterprises and the imperfect design of the employee stock ownership system, internal employee stock ownership has been utilized by some individuals, resulting in significant losses to state-owned assets. Consequently, internal employee stock ownership has been repeatedly suspended by regulatory authorities. In summary, the initial ESOPs in China have undergone a series of developments, from rapid expansion to emergency suspension and subsequent suspension of issuance.

Until 20 June 2014, the China Securities Regulatory Commission (CSRC) issued the Guiding Opinions, which set forth comprehensive regulations governing the participants, stock sources, capital sources, shareholding scale, duration, management mode, and information disclosure of listed companies implementing ESOPs. The CSRC’s regulations standardized and unified the relevant approval process for implementing ESOPs. The release of the Guiding Opinions prompted a surge in the number of listed companies implementing ESOPs, with the practice of employee stock ownership experiencing significant growth over the past decade. As evidenced by the Wind database, from 2014 to 2023, 1137 A-share listed companies announced the implementation of ESOPs, representing a total of 1952 ESOPs. Consequently, ESOPs have emerged as a prominent topic in Chinese accounting circles over the past decade.

2.2. Literature Review

The existing literature on this topic includes studies on the implementation effect of ESOPs and the determinants of capital structure decisions. This review will summarize these two aspects.

2.2.1. The Consequences of ESOP Implementation

This paper presents a review of the existing literature on the implementation effect of ESOPs, organized according to three key aspects: capital market consequences, firm performance, and firm innovation.

(1)

The capital market consequences of implementing ESOPs

The extant literature on the capital market consequences of ESOPs primarily examines these consequences from the perspectives of the shareholder wealth effect, the cost of equity capital, and stock price crash risk. In terms of the shareholder wealth effect, most western scholars concur that the announcement of an ESOP has a considerable impact on shareholder wealth [12]. The magnitude of this effect is contingent upon the rationale behind the ESOP’s implementation and its influence on corporate control [13]. Since the issuance of the Guiding Opinions, Chinese scholars have commenced empirical investigations into the announcement effect of ESOPs in the new era. Their findings consistently indicate that the announcement of an ESOP can result in a substantial short-term increase in shareholder wealth. Additionally, the wealth effect is found to be contingent upon the specific design of the ESOP contract [14,15,16]. Regarding the cost of equity capital, the extant literature on the relationship between employee stock ownership and the cost of equity capital has yielded disparate conclusions. In their study, Cheng and Ji [17] examine the impact of ESOP implementation on the cost of equity capital for Chinese listed companies. Their findings indicate that ESOP adoption leads to a reduction in the cost of equity capital for these enterprises. However, Aubert et al. [18] employ French listed companies as their research sample and conclude that there is no significant relationship between employee stock ownership and the cost of equity capital. Campa and Kern’s [19] research on listed companies in the United States indicates the existence of a U-shaped relationship between employee stock ownership and the cost of equity capital. Regarding stock price crash risk, most current studies have indicated that the implementation of ESOP can serve to mitigate the risk of stock price collapse [20,21]. However, from the perspective of institutional design, Guan et al. [22] have identified that the institutional design features of ESOP in China at the present stage may encourage management and major shareholders to be more inclined to conceal unfavorable information, which could potentially increase the risk of stock price collapse of listed companies in the future.

(2)

The impact of ESOPs on firm performance

Most early studies indicate that the implementation of ESOPs has the potential to significantly enhance corporate performance [23]. However, several studies have identified that ESOPs have no discernible impact on corporate performance [24], while a few studies have also reported a negative impact of ESOPs on corporate performance [25]. In recent years, scholars in China have conducted further analysis of the relationship between employee stock ownership and enterprise performance, specifically in the context of ESOPs implemented in the current era. Their findings have been inconsistent. In their study, Shi and Cui [26] conclude that the implementation of ESOP has a positive effect on the financial performance of enterprises. They find that ESOPs can improve the profitability, solvency, operational capacity, and development potential of enterprises, while simultaneously reducing the financial risk faced by these entities. However, Shen et al. [27] conclude that the implementation of ESOPs in state-owned enterprises does not result in a notable enhancement of corporate governance and business performance, due to an absence of sufficient management incentives. Ren et al. [28] find that, in comparison to non-implementing companies, the absolute performance of companies implementing ESOPs is higher both before and after implementation; yet the relative performance does not improve after implementation.

(3)

The impact of ESOPs on enterprise innovation

The research on the impact of employee equity incentives on enterprise innovation is relatively limited in the West, whereas in China numerous studies on the relationship between ESOP and enterprise innovation have emerged in recent years. These studies concentrate on the influence of ESOP on the innovation output of enterprises. They consistently demonstrate that the implementation of ESOP has a positive effect on the innovation output, with the strength of this impact varying according to the characteristics of employees, companies or ESOPs [29,30,31,32]. Furthermore, Cao et al. [33] suggest that the implementation of ESOPs can markedly enhance the dual innovation of enterprises. Li and Ding [34] present evidence indicating that ESOPs have a significant incentive effect on innovation. However, the relationship between the two exhibits a U-shaped cumulative effect. Hong [35] categorizes ESOPs according to their contractual characteristics, distinguishing between governance type ESOPs, incentive type ESOPs, and binding type ESOPs. Subsequently, he determines that, in comparison to governance type ESOPs, both incentive and binding type ESOPs can enhance the innovation output of enterprises, with the effect of binding type ESOPs being particularly pronounced. Additionally, a limited number of studies have investigated the correlation between ESOP and innovation input. To illustrate, Luo and Wang [36] examine the correlation between R&D expenditure and the executive subscription ratio in ESOPs, and ascertain that both the company’s R&D expenditure and the expensed R&D expenditure exhibit a U-shaped relationship with the executive subscription ratio in ESOPs. Cheng et al. [37] investigate the influence of ESOP contract elements on innovation input, concluding that the participation degree, capital source, stock source, lock-up period, duration, management mode, shareholding scale, and executive subscription ratio of ESOPs have a significant impact on innovation input.

2.2.2. The Determinants of Capital Structure Decisions

The decision-making process regarding capital structure typically encompasses two key aspects: the selection of an optimal capital structure and the dynamic adjustment of that structure over time. This paper presents a comprehensive review of the existing literature on the factors influencing capital structure choice and the determinants of capital structure adjustment.

(1)

The determinants of capital structure choice

At the macro-policy level, Zhang [38] identifies a reduction in the level of corporate debt resulting from industrial policy support. Conversely, Huang et al. [39] confirm that the level of corporate debt is increased by tax collection and administration. At the firm level, Titman and Wessels [4] employ factor analysis to investigate the influence of firm attributes on capital structure. Their findings indicate that, in general, larger enterprises tend to exhibit higher debt ratios, as do those with greater asset values and faster growth rates. Conversely, an elevated level of profitability or non-debt tax shields is associated with a reduction in debt. Zhang and Peng [40] identify a negative correlation between R&D investment and capital structure, indicating that enterprises tend to reduce debt financing in response to heightened financial risks when R&D investment is high. At the executive level, Gao and Yang [41] investigate the influence of CEO education on capital structure and conclude that CEO education is positively correlated with the asset-liability ratio of managed enterprises. Zhao et al. [42] identify an inverted U-shaped relationship between executive compensation incentives and capital structure. Cheng et al. [43] propose that the appointment of a CFO as board secretary is associated with a reduction in corporate debt. Chen and Ma [44] posit that founder overconfidence is associated with an increase in corporate leverage.

(2)

The determinants of capital structure dynamic adjustment

At the macro-environmental level, Hackbarth [1] posits that economic conditions exert an influence on the speed of capital structure adjustment. During periods of economic growth, enterprises tend to accelerate the speed of capital structure adjustment. Jiang and Huang [3] investigate the influence of the institutional background of the marketisation process on the dynamic adjustment of capital structure. Their findings indicate that, as the marketisation degree improves, the adjustment speed of the capital structure increases; while the deviation degree of actual capital structure from the target capital structure reduces. At the enterprise level, Jalivand and Haris [45] indicate that larger enterprises tend to exhibit superior business performance and reputation, facilitating more expedient capital structure adjustment through financing and other avenues. Sheng et al. [46] posit that the strategic differences between enterprises serve to reduce the adjustment speed of capital structure. In other words, the higher the degree of strategic differentiation, the slower the speed of capital structure adjustment. At the managerial level, Zhang [5] identifies that MBA education facilitates the dynamic adjustment of capital structure. Jia et al. [47] investigate the influence of CEO background characteristics on the speed of dynamic capital structure adjustment and conclude that the effect of dynamic capital structure adjustment is enhanced when the CEO is male, older, or has been through higher education.

In terms of the impact of equity incentives, the existing literature primarily addresses the relationship between equity incentives and capital structure decisions from the perspective of executive equity incentives. Hou et al. [6] investigate the influence of management equity incentive contracts and their stipulations on the dynamic adjustment of capital structure. Their findings indicate that the implementation of equity incentive contracts by enterprises can facilitate the dynamic adjustment of capital structure. The implementation of equity incentive contracts with higher standard exercise conditions, longer validity periods, and stronger incentive strengths has been observed to facilitate the dynamic adjustment of capital structure towards the target level. In their study of the impact of executive equity incentives on the dynamic adjustment of capital structure, Sheng et al. [1] find that executive equity incentives have a significant promoting effect on capital structure adjustment. Furthermore, this promoting effect is asymmetrical in terms of upward and downward adjustments. In other words, when the capital structure requires downward adjustment, the promoting effect of equity incentives on the dynamic adjustment speed of capital structure is greater.

In conclusion, the extant literature is deficient in two respects. First, the extant literature on the implementation effect of ESOPs is largely confined to an examination of the consequences for capital markets, firm performance and firm innovation. There is a paucity of studies that address the impact of ESOPs on capital structure decisions. Second, existing studies have conducted comprehensive analyses of the factors influencing capital structure decisions, yet have devoted less attention to the impact of equity incentives on capital structure decisions, particularly in the context of employee equity incentives. It is important to note that ESOPs and equity incentive plans differ significantly in terms of their incentive object, incentive mechanism and assessment method. Consequently, the existing research conclusions on the impact of equity incentive plans on capital structure decisions may not be applicable to ESOPs. It is therefore necessary to discuss the impact of ESOP on capital structure decisions. As a component of the company’s equity incentives, ESOP plays an important role in alleviating agency problems among shareholders, executives and employees, supervising management behavior, improving the overall performance of the company, and enhancing corporate transparency, which may have an impact on capital structure decisions. The purpose of this paper is therefore to explore the factors influencing capital structure decisions from the perspective of employee equity incentives.

3. Hypothesis Development

3.1. The Impact of ESOPs on Static Capital Structure Choice

This paper argues that ESOPs may affect firms’ capital structure choice in three ways (as described in Figure 1).

First, ESOP has the potential to reduce debt levels by motivating employees and promoting teamwork to improve corporate performance. An ESOP enables employees to assume the dual identities of laborers and owners, linking their personal interests with those of shareholders. It also enhances the motivation and capacity of employees to engage in corporate governance, thereby alleviating interest conflicts among shareholders, managers and employees. This, in turn, reduces agency costs and improves the financial performance of enterprises [8,27]. Concurrently, the ESOP fosters a positive environment for mutual learning and supervision among employees [48], facilitates teamwork, enhances employee motivation and satisfaction with their work, and thus improves the business performance of enterprises [49]. An improvement in the company’s financial and business performance can lead to an increase in operating profit and shareholder wealth, thereby providing the company with greater financial flexibility and capacity to repay debts. Additionally, it can attract external investors to purchase the company’s shares, which may in turn influence the management’s preference for equity financing and contribute to a reduction in the company’s overall debt levels.

Second, in terms of alternative financing tools, ESOP has the potential to reinforce a firm’s inclination towards equity financing, diminish reliance on debt financing and consequently reduce the overall level of debt. Some studies have indicated that, in contrast to the west, China’s ESOP is employed more as a financing instrument to effectively address the issue of excessive corporate indebtedness and to make the company’s capital structure more conservative [50]. From an operational standpoint, ESOPs that source their stock through non-public offering shares can directly replenish the company’s working capital, alleviate financing constraints, and increase equity financing [51]. ESOP enables firms to align employee interests with corporate performance without the need for cash payments. Consequently, firms that adopt ESOPs tend to favor equity financing [52], which in turn reduces reliance on debt financing and lowers debt levels. Moreover, if employee remuneration is used as a source of funding for an ESOP, it can replace the monetary remuneration paid to employees, reducing the company’s cash outflow, improving the balance sheet and lowering the level of debt [53].

Third, ESOP has the potential to mitigate the level of liabilities by reinforcing the efficacy of internal control mechanisms. Wang et al. [54] posit that ESOP can enhance internal control effectiveness by strengthening internal supervision, improving information and communication efficiency, and optimizing control activities. The implementation of enhanced internal controls can facilitate the creation of an internal environment that is conducive to the achievement of performance targets [55], thereby increasing company profits and reducing debt levels. Furthermore, enhanced internal controls can be an effective means of mitigating the adverse effects of managerial overconfidence and political affiliation on capital structure decisions [8]. By monitoring the actions of management, employees can deter self-serving and unscrupulous behavior, thereby ensuring the enterprise’s value is optimized as a business objective. This will encourage managers to select a conservative capital structure and reduce the level of indebtedness. In addition, in comparison to managers, employees demonstrate a lower capacity for diversifying risk and a greater propensity for risk aversion [55]. Consequently, employees can circumvent risk and diminish leverage by monitoring the risks associated with production and operations.

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

Hypothesis 1 (H1).

The implementation of an ESOP can reduce the company’s debt level.

3.2. The Impact of ESOPs on Dynamic Capital Structure Adjustment

This paper argues that ESOPs may affect the dynamic adjustment of capital structure in the following three ways (as described in Figure 2).

First, the signaling theory posits that an ESOP serves to reassure external stakeholders of the company’s positive development prospects, thereby alleviating concerns about the potential uncertainty of the company’s future. This, in turn, has the effect of reducing financing constraints, lowering the adjustment cost of capital structure and promoting the dynamic adjustment of capital structure. In accordance with the pertinent regulations, the draft ESOP is to be subjected to review at multiple levels—including by the labor union, the board of directors, and the general meeting of shareholders—prior to its release to the public. Moreover, the principle of voluntary employee participation is to be observed. As the primary source of internal information, employee share purchases demonstrate their recognition of the company’s prospects. Management, in contrast, possesses more pronounced insights into the company’s operational performance. Their voluntary subscription behavior reflects an enhanced corporate governance structure and can facilitate the transmission of positive signals, reduce the adjustment cost of capital structure through financing, and thus promote the dynamic adjustment of enterprise capital structure.

Second, as posited by the tenets of asymmetric theory, the implementation of ESOP can serve to mitigate the informational asymmetry that exists between employees and enterprises, as well as between external investors and internal managers. This is achieved by enhancing the transparency of corporate information and reducing the costs associated with adjusting the capital structure, thereby facilitating the dynamic adjustment of capital structure. ESOP can enhance internal control, facilitate effective supervision of management, and deter self-serving behavior among management. This, in turn, can lead to greater transparency of information and a reduction in the degree of asymmetry between internal and external information. On this basis, corporate information can be transmitted among stakeholders, thereby eliminating the need for creditors to demand higher compensation for information risk [56]. This can also enhance the confidence of capital market investors in corporate stocks and stimulate investors’ willingness to invest in enterprises [55,57]. A reduction in the costs of debt and equity financing can diminish the expense associated with the realignment of a company’s capital structure. This, in turn, can bolster the motivation to modify the structure and speed up adjustment. The financial capacity of enterprises has been augmented, thereby facilitating a more expedient realignment of their capital structure through financing activities.

Third, in accordance with the tenets of agency theory, ESOPs serve to incentivize both participating executives and employees, thereby reducing the agency costs of enterprises and facilitating the dynamic adjustment of capital structures. The separation of the two rights between shareholders and executives results in inconsistent value orientation, which in turn gives rise to moral hazard and adverse choices by executives, and ultimately, agency problems. The implementation of an ESOP can facilitate the alignment of enterprise ownership and control with employee shareholdings, thereby encouraging managers to reduce self-interested behaviors and make investment decisions that maximize the value of the enterprise [8]. This may involve selecting investment projects with high returns and high risks. Given the inherent risk associated with such projects, enterprise managers are compelled to closely monitor the debt and income situation of the projects post-investment. This enables them to enhance their capacity to adjust the capital structure, speed up the capital structure adjustment, and facilitate the dynamic adjustment of capital structure.

In conclusion, the implementation of ESOPs can reduce the adjustment costs of capital structure for enterprises by reducing the costs of debt and equity financing [17,56], thereby accelerating the adjustment speed of capital structure. Therefore, this paper proposes the following hypothesis:

Hypothesis 2 (H2).

The implementation of an ESOP can speed up the dynamic adjustment of capital structure.

4. Methodology

4.1. Data and Sample

This paper employs a sample of A-share listed companies in China’s Shanghai and Shenzhen Stock Exchanges from 2014 to 2021. The following screening procedures are then carried out on this sample: (1) Excluding samples from the financial and insurance industries; (2) Excluding financially distressed companies (i.e., companies subject to special treatments by the regulators); (3) Excluding samples with ESOP implementation failures, such as “suspension of implementation” and “failure to pass the shareholders’ meeting”; and (4) Excluding samples with missing data of variables required for this study. Furthermore, to eliminate the potential impact of outliers on the empirical findings, this study winsorizes all continuous variables at the 1st and 99th percentiles. The final sample comprises 19,703 firm-year observations.

In this paper, we define the starting year of the ESOP as the date of approval of the ESOP proposal by the shareholders’ meeting. The ending year of the ESOP is defined as the date of announcement of completion of share sales. If the current year falls between the commencement and conclusion of the ESOP period, this indicates that the firm is implementing the ESOP in the current year. In instances where a single company has implemented multiple ESOPs over the course of the sample period, the earliest starting year and the latest ending year of ESOP are retained. The data on ESOP implementation and the contractual factors required for the study were obtained from the Wind database. The capital structure and other corporate financial data, as well as the data on governance structures, were sourced from the CSMAR database.

4.2. Variables

4.2.1. Dependent Variables

(1)

Capital structure

In accordance with the tenets of static and dynamic trade-off theories, the capital structure decision can be classified into two distinct categories: static capital structure choice and dynamic capital structure adjustment. Previous studies have identified two broad definitions of capital structure, distinguished by whether they include short-term liabilities. Given that the debt maturity structure of Chinese listed companies is typified by short-term debt financing, the proportion of short-term liabilities in total liabilities is markedly higher than that of long-term liabilities. Additionally, there is a tendency for short-term liabilities to be recycled [40]. Consequently, this paper adopts the broad concept of capital structure and employs the gearing ratio to quantify the firm’s capital structure.

(2)

Dynamic adjustment of capital structure

To ascertain the adjustment speed of the capital structure, it is first necessary to determine the optimal capital structure for the firm. As evidenced in previous literature, the firm’s target capital structure is typically shaped by a combination of firm-specific characteristics and shifts in the internal and external environment [58]. Accordingly, this paper employs the methodologies of Cheng and Wang [43], Jiang and Huang [3], and Sheng et al. [1] to construct a multiple regression model incorporating firm characteristic variables for estimating the target capital structure, as follows:

$$\begin{array}{c}{\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}^{*}=\mathsf{\beta }{\mathbf{X}}_{\mathbf{i},\mathbf{t}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{R}\mathrm{o}\mathrm{a}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_2{\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_3{\mathrm{G}\mathrm{r}\mathrm{o}\mathrm{w}\mathrm{t}\mathrm{h}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_4{\mathrm{T}\mathrm{a}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}\\ +{\mathsf{\beta }}_5{\mathrm{D}\mathrm{e}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_6{\mathrm{O}\mathrm{c}\mathrm{f}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_7{\mathrm{I}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}}+\sum {\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}}\end{array}$$

(1)

where ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}^{*}$ denotes the target capital structure of firm i in year t + 1; β denotes the vector of regression coefficients; ${\mathbf{X}}_{\mathbf{i},\mathbf{t}}$ denotes the firm characteristic variables and year fixed effects for firm i in year t. ${\mathrm{R}\mathrm{o}\mathrm{a}}_{\mathrm{i},\mathrm{t}}$ is the return on total assets; ${\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}}_{\mathrm{i},\mathrm{t}}$ is the natural logarithm of total assets; ${\mathrm{G}\mathrm{r}\mathrm{o}\mathrm{w}\mathrm{t}\mathrm{h}}_{\mathrm{i},\mathrm{t}}$ is the growth rate of operating revenue; ${\mathrm{T}\mathrm{a}\mathrm{n}\mathrm{g}}_{\mathrm{i},\mathrm{t}}$ is the net fixed assets divided by total assets; ${\mathrm{D}\mathrm{e}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ is total depreciation and amortization divided by total assets; ${\mathrm{O}\mathrm{c}\mathrm{f}}_{\mathrm{i},\mathrm{t}}$ is the net cash flow from operating activities divided by total assets; ${\mathrm{I}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}}$ is the median gearing ratio of an industry in year t. ${\mathsf{\beta }}_0$ denotes the constant term; ${\mathsf{\beta }}_1$ to ${\mathsf{\beta }}_7$ denote the regression coefficients. The control variables are defined in

Table 1. Definitions of variables.

Symbol	Variable Definition
Lev	It represents firm’s capital structure, which is equal to the total liability divided by total assets at the end of the year.
$\delta$	It represents the speed of capital structure adjustment, which is estimated by Formula (2).
Esop	It represents the implementation of ESOP, which is equal to 1 when the firm is implementing ESOP in the current year and is otherwise equal to 0.
Roa	It represents operating profitability, which is the net profit in the current year divided by the total assets at the end of the year.
Size	It represents firm size, which is the natural logarithm of total assets at the end of the year.
Growth	It represents sales growth, which is the difference between the current year’s revenue and last year’s revenue divided by last year’s revenue.
Tang	It represents capital intensity, which is equal to the fixed assets divided by total assets at the end of the year.
Dep	It represents the growth ability, which is the total depreciation and amortization divided by total assets.
Ocf	It represents the growth ability, which is the net cash flows from operating activities divided by total assets.
Ind	It represents the industrial capital structure, which is the median asset–liability ratio for a given industry by year by industry.
Soe	When the nature of the company’s equity is a state-owned enterprise, it is 1, otherwise it is 0.
Dual	It represents a dual structure, which is equal to 1 when the chairman of the board serves concurrently as general manager and is otherwise equal to 0.
Mpay	It represents the managerial pay, which is the natural logarithm of the top three executives’ compensation.
Top1	It represents the ownership concentration, which is the shareholding ratio of the largest shareholder, in %.
Year	It represents year-fixed effects, which are represented by 7-year dummies. The year 2014 is used as the baseline in setting year dummies.

.

Drawing on existing studies, the standard partial adjustment model proposed by Flannery and Rangan [58] is utilized to estimate the speed of firms’ capital structure adjustment, as follows:

$${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}-{\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}={\displaystyle \frac{{\mathrm{D}}_{\mathrm{i},\mathrm{t}+1}}{{\mathrm{A}}_{\mathrm{i},\mathrm{t}+1}}}-{\displaystyle \frac{{\mathrm{D}}_{\mathrm{i},\mathrm{t}}}{{\mathrm{A}}_{\mathrm{i},\mathrm{t}}}}=\mathsf{\delta }\left({\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}^{*}-{\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}\right)+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^1$$

(2)

where, ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$ and ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$ represent the actual capital structure of company i at the end of year t and t + 1, respectively, expressed by the asset–liability ratio of the company in that year; δ represents the average adjustment speed of capital structure of the sample companies in that year; ${\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^1$ represents the residual.

In accordance with the dynamic trade-off theory, the actual capital structure will inevitably deviate from the target structure due to the inherent costs associated with capital structure adjustment. Consequently, the enterprise will pursue a trajectory of capital structure adjustment that balances the benefits and costs of such adjustments, thereby aligning the target structure with the actual structure. This implies that the target capital structure and the adjustment speed of capital structure should be estimated concurrently. Accordingly, this paper employs a simultaneous estimation approach, whereby the target capital structure and the speed of capital structure adjustment are both calculated concurrently. This is achieved by substituting Formula (1) into Formula (2), which yields the following equation:

$${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}=\mathsf{\delta }\mathsf{\beta }{\mathbf{X}}_{\mathbf{i},\mathbf{t}}+\left(1-\mathsf{\delta }\right){\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^2$$

(3)

In addition, Flannery and Rangan [58] posit that the fixed-effect estimator utilizing panel data (FE regression model) is superior to the ordinary least squares estimator employing pooled cross-sectional data (OLS regression model) in estimating the speed of dynamic capital structure adjustment. The OLS regression model is thought to underestimate the speed of dynamic capital structure adjustment due to the omission of firm fixed effects. Furthermore, our research sample includes companies that have missing data for at least one year during the sample period. Therefore, we employ the FE regression model based on unbalanced panel data to estimate the speed of dynamic capital structure adjustment.

4.2.2. Independent Variable

In this paper, we introduce a new variable, Esop, which is used to indicate whether a company is implementing an ESOP in the current year. This variable takes the value of 1 when the company is implementing an ESOP in the current year and 0 otherwise. As a long-term incentive for the company, one ESOP generally has a lock-up period of between one and three years. Accordingly, in the case of the sample comprising companies with ESOPs, the approval of the ESOP proposal by the shareholders’ meeting is taken to mark the commencement of the ESOP, while the announcement of the completion of the sale of the underlying stock marks its conclusion. All years falling within these periods are regarded as encompassing the implementation of the ESOP.

4.2.3. Control Variables

In light of studies conducted by Cheng and Wang [43] and Wang and Deng [59], this paper selects the following firm characteristics and governance structure variables as the control variables of the empirical model: profitability (Roa); firm size (Size); growth rate of operating revenue (Growth); ability to pledge assets (Tang); non-debt tax shield (Dep); cash flows from operation (Ocf); industry capital structure (Ind); ownership nature (Soe); dual structure of board (Dual); executive compensation (Mpay); equity concentration (Top1). Furthermore, this paper controls for year-fixed effects. The precise definitions of the control variables are provided in Table 1.

4.3. Models

4.3.1. The Model Testing the Effect of ESOP on Capital Structure Choice

To test the impact of ESOP implementation on capital structure choice, this paper constructs the following multiple regression model:

$${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+\mathsf{\beta }{\mathbf{C}\mathbf{o}\mathbf{n}\mathbf{t}\mathbf{r}\mathbf{o}\mathbf{l}\mathbf{s}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^3$$

(4)

The dependent variable ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$ represents the capital structure of company i in year t + 1, and the independent variable ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ represents the dummy variable proxied for ESOP implementation of company i in year t. The Controls represents control variables, including company size (Size), profitability (Roa), sales growth (Growth), asset collateral (Tang), non-debt tax shield (Dep), cash flows from operation (Ocf), industry capital structure (Ind), equity nature (Soe), dual structure of board (Dual), management payment (Mpay), ownership concentration (Top1) and year fixed effect (Year). ${\mathsf{\alpha }}_0$ represents the constant term; ${\mathsf{\alpha }}_1$ represents the regression coefficient of the independent variable; $\mathsf{\beta }$ represents the regression coefficient matrix of the control variable; and ${\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^3$ represents the residual of Equation (4). This paper focuses on the regression coefficient ${\mathsf{\alpha }}_1$, which measures the difference in capital structure between companies with ESOPs and those without ESOPs. If ${\mathsf{\alpha }}_1$ is significantly negative, it indicates that the implementation of ESOP can significantly reduce the company’s debt level.

4.3.2. The Model Testing the Impact of ESOP on Capital Structure Adjustment

To test the impact of ESOP implementation on the dynamic adjustment of capital structure, this paper draws on the models of Jiang et al. [3] and Huang et al. [60] to construct the following fixed effect model:

$${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}={\mathsf{\gamma }}_0+{\mathsf{\gamma }}_1{\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\gamma }}_2{\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\gamma }}_3{\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}\times {\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}+\mathsf{\beta }{\mathbf{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\lambda }}_{\mathrm{i}}{+\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^4$$

(5)

where, ${\mathsf{\gamma }}_0$ represents the constant term; ${\mathsf{\gamma }}_1$ represents the regression coefficient of the capital structure in year t + 1 on the capital structure in year t; $\mathsf{\beta }$ represents the regression coefficient vector of control variables (i.e., all explanatory variables in Equation (1)); ${\mathsf{\lambda }}_{\mathrm{i}}$ represents firm fixed effect; and ${\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^4$ represents the residual of Equation (5). By comparing Equation (5) with Equation (3), we can get the expression of capital structure adjustment speed: $\mathsf{\delta }=1-{\mathsf{\gamma }}_1-{\mathsf{\gamma }}_3{\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$. When the enterprise is implementing ESOP, ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ equals to 1, and the capital structure adjustment speed is $1-{\mathsf{\gamma }}_1-{\mathsf{\gamma }}_3$. When the enterprise has not implemented ESOP in year t, ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ equals to 0, and the capital structure adjustment speed is $1-{\mathsf{\gamma }}_1$. The difference in the speed of capital structure adjustment between companies with ESOPs and those without ESOPs is $-{\mathsf{\gamma }}_3$. Therefore, the interaction coefficient ${\mathsf{\gamma }}_3$ can be used to measure the impact of ESOP implementation on the speed of capital structure adjustment. If ${\mathsf{\gamma }}_3$ is significantly negative, it indicates that the implementation of ESOP can significantly accelerate the adjustment speed of capital structure.

5. Empirical Results and Analysis

5.1. Descriptive Analysis

5.1.1. Descriptive Statistics of the Full Sample

Table 2. Descriptive statistics of the full sample.

Variable	N	Mean	S.D.	Min	P25	Median	P75	Max
Lev	19,703	0.430	0.208	0.059	0.265	0.419	0.580	0.947
Esop	19,703	0.143	0.351	0	0	0	0	1
Roa	19,703	0.031	0.077	−0.395	0.012	0.034	0.064	0.201
Size	19,703	22.254	1.306	19.649	21.330	22.093	22.992	26.190
Growth	19,703	0.171	0.493	−0.652	−0.041	0.090	0.253	3.273
Tang	19,703	0.208	0.160	0.002	0.083	0.173	0.299	0.691
Dep	19,703	0.022	0.015	0.001	0.011	0.019	0.030	0.070
Ocf	19,703	0.047	0.070	−0.176	0.008	0.047	0.088	0.248
Ind	19,703	0.415	0.106	0.145	0.347	0.395	0.457	0.720
Soe	19,703	0.346	0.476	0	0	0	1	1
Dual	19,703	0.285	0.451	0	0	0	1	1
Mpay	19,703	14.564	0.680	13.010	14.108	14.523	14.964	16.545
Top1	19,703	0.337	0.146	0.084	0.222	0.314	0.433	0.743

presents descriptive statistics for the full sample. As illustrated in the table, in terms of explained variables, the mean and median values of Lev are 0.430 and 0.419, respectively. Meanwhile, the minimum and maximum values of Lev are 0.059 and 0.947. This indicates that the asset–liability ratios of more than half of sample companies are lower than 50% and that there is a considerable range in the asset–liability ratio among companies. Regarding the explanatory variable, the mean value of Esop is 0.143, indicating that within the selected sample periods, only 14.3% of sample companies have implemented ESOPs. This suggests that the scope of ESOP implementation is limited in Chinese listed companies and that ESOP promotion is still in its pilot phase. It also indicates that most enterprises are cautious about implementing ESOPs.

5.1.2. Descriptive Statistics of the Subsamples

Table 3. Descriptive statistics of subsamples implementing ESOPs and not implementing ESOPs.

Variable	(1) Esop = 0			(2) Esop = 1			Difference = (2) − (1)
	N	Mean	Median	N	Mean	Median	Mean Diff.	Median Diff.
Levt+1	16,877	0.436	0.426	2826	0.443	0.436	0.007	0.010 **
Levt	16,877	0.425	0.414	2826	0.427	0.425	0.002	0.011 **
Roa	16,877	0.030	0.034	2826	0.032	0.039	0.002	0.006 ***
Size	16,877	22.209	22.039	2826	22.519	22.371	0.309 ***	0.333 ***
Growth	16,877	0.167	0.084	2826	0.194	0.129	0.027 ***	0.045 ***
Tang	16,877	0.213	0.177	2826	0.177	0.155	−0.036 ***	−0.022 ***
Dep	16,877	0.022	0.020	2826	0.020	0.018	−0.002 ***	−0.002 ***
Ocf	16,877	0.047	0.047	2826	0.047	0.047	0.00004	0.001
Ind	16,877	0.418	0.398	2826	0.397	0.369	−0.021 ***	−0.029 ***
Soe	16,877	0.382	0	2826	0.129	0	−0.253 ***	-
Dual	16,877	0.274	0	2826	0.346	0	0.072 ***	-
Mpay	16,877	14.536	14.499	2826	14.729	14.664	0.193 ***	0.165 ***
Top1	16,877	0.341	0.319	2826	0.309	0.291	−0.032 ***	−0.029 ***

Note. The symbol *** and ** indicate significance at the 1% and 5% level.

presents a comparison between the ESOP subsample and the non-ESOP subsample in terms of capital structure, firm characteristics, and governance structure. Regarding the capital structure variable, the median difference tests of Lev_t+1 and Lev_t are both found to be significant at the 5% level. This indicates that the asset–liability ratio of enterprises implementing ESOPs is higher than that of enterprises not implementing ESOPs, which contradicts Hypothesis 1. However, as univariate analysis does not control for the effects of other variables, this finding may be due to confounding factors. For instance, firm size is positively correlated with both debt level and ESOP implementation, resulting in higher debt levels for ESOP firms. Consequently, we need to further use the multiple regression analysis to exclude the influence of confounding factors for formal hypothesis testing.

In terms of corporate characteristics, ESOP companies are observed to exhibit a larger size, stronger profitability, and faster sales growth in comparison to non-ESOP companies. However, they display weaker asset mortgage ability and a greater number of non-debt tax shields. Regarding the corporate governance variables, it is observed that 12.9% of enterprises which have implemented ESOPs are state-owned enterprises, whereas 38.2% of enterprises which have not implemented ESOPs are state-owned enterprises. This indicates that non-state-owned enterprises are more likely to implement ESOPs than state-owned enterprises. The ratio of dual-role structures in ESOP companies is 7.2% higher than that in non-ESOP companies, indicating that enterprises with the dual roles of chairman and CEO are more likely to implement ESOPs. The mean shareholding ratio of the largest shareholder in enterprises with ESOPs is 0.291, compared to 0.341 in enterprises without ESOPs. This suggests that enterprises with lower ownership concentration are more likely to implement ESOPs.

5.2. Correlation Analysis

To preliminarily test the correlation between variables, the Pearson correlation coefficients between the major variables are calculated, and the results are presented in

Table 4. Pearson correlations between the main variables.

Variable	Levt+1	Levt	Esop	Roa	Size	Growth	Tang	Dep	Ocf	Ind
Levt	0.898 ***	1
Esop	0.005	−0.002	1
Roa	−0.325 ***	−0.351 ***	0.008	1
Size	0.433 ***	0.461 ***	0.083 ***	0.051 ***	1
Growth	0.025 ***	0.019 ***	0.019 ***	0.201 ***	0.035 ***	1
Tang	0.029 ***	0.060 ***	−0.078 ***	−0.050 ***	0.094 ***	−0.085 ***	1
Dep	−0.027 ***	0.0002	−0.048 ***	−0.111 ***	0.008	−0.112 ***	0.764 ***	1
Ocf	−0.203 ***	−0.178 ***	0.0002	0.348 ***	0.065 ***	−0.006	0.227 ***	0.258 ***	1
Ind	0.396 ***	0.418 ***	−0.069 ***	−0.105 ***	0.343 ***	−0.007	−0.022 ***	−0.113 ***	−0.112 ***	1
Soe	0.218 ***	0.254 ***	−0.186 ***	−0.048 ***	0.366 ***	−0.075 ***	0.193 ***	0.115 ***	−0.013 *	0.240 ***
Dual	−0.101 ***	−0.119 ***	0.056 ***	0.029 ***	−0.180 ***	0.026 ***	−0.086 ***	−0.063 ***	0.004	−0.131 ***
Mpay	0.094 ***	0.091 ***	0.099 ***	0.168 ***	0.434 ***	0.019 ***	−0.119 ***	−0.073 ***	0.166 ***	0.070 ***
Top1	0.028 ***	0.040 ***	−0.077 ***	0.147 ***	0.217 ***	−0.014 *	0.097 ***	0.049 ***	0.116 ***	0.126 ***

Note. The symbol *** and * indicate significance at the 1% and 10% levels, respectively.

. As can be seen from the table, the correlation between Lev_t+1 and Esop is not significant, indicating that Hypothesis 1 is not supported. Consequently, further tests should be conducted using multiple regression analysis. Significant correlations exist between capital structure and most control variables (except for Dep), thereby validating the selection of control variables to a certain extent. Furthermore, there were significant correlations between Esop and most control variables (except for Roa and Ocf), indicating that there are notable differences in corporate characteristics and governance structure between ESOP companies and non-ESOP companies. This serves to reinforce the importance of controlling for these factors when examining the impact of ESOP implementation on capital structure decisions. Furthermore, the absolute values of the correlation coefficients among the control variables (except for Tang and Dep) and between the independent variables and the control variables are all less than 0.500, indicating that the empirical models constructed in this paper do not exhibit serious multicollinearity issues.

5.3. Regression Analysis

5.3.1. The Impact of ESOP on Capital Structure Choices

To test the impact of ESOP implementation on capital structure, Equation (4) is estimated with ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$ as the dependent variable and ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ as the explanatory variable. The results are shown in

Table 5. Regression results: the impact of ESOP on capital structure choices.

Variable	Baseline Model:  ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$		Model (4):  ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
	Coefficient	t-Value	Coefficient	t-Value
Esop	−0.011 ***	−3.22	−0.009 ***	−2.73
Roa			−0.822 ***	−27.97
Size			0.066 ***	56.23
Growth			0.039 ***	11.67
Tang			0.020	1.63
Dep			0.127	0.91
Ocf			−0.301 ***	−12.61
Ind			0.443 ***	31.10
Soe			0.008 **	2.56
Dual			0.003	1.27
Mpay			−0.011 ***	−5.02
Top1			−0.020 **	−2.40
C	0.438 ***	97.14	−1.027 ***	−36.79
Year FE	yes		yes
Adj-R2	0.001		0.402
N	19,703		19,703
F-value	3.84 **		843.45 ***

Note. The symbols *** and ** indicate significance at the 1% and 5% levels, respectively; C represents the constant term of the regression models.

. As can be observed from the table, the regression coefficients of the explained variable ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$ on the explained variable ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ before and after adding control variables are −0.011 and −0.009, respectively. Both are significant at the 1% level, indicating that the implementation of ESOP is significantly negatively correlated with the asset–liability ratio of enterprises. This evidence supports the hypothesis that the implementation of ESOP can significantly reduce the debt level of enterprises. From an economic standpoint, the implementation of ESOP results in a reduction of the company’s debt level by 0.9%. In other words, companies with ESOPs are more likely to adopt a conservative capital structure than non-ESOP companies. From an employee perspective, the implementation of ESOP encourages greater efforts, thereby improving corporate performance, increasing profits and shareholder wealth. From the perspective of the company, it replaces monetary compensation, reduces cash expenditure, supplements working capital, and improves assets and liabilities. From the perspective risk management, the implementation of ESOP is helpful in improving the construction of internal control of enterprises and strengthening the supervision of managers, encouraging enterprises to avoid risks and choose a more stable capital structure.

5.3.2. The Impact of ESOP on the Dynamic Adjustment of Capital Structure

To test the impact of the implementation of ESOP on the speed of capital structure adjustment, this paper uses unbalanced panel data fixed-effect models to estimate Equation (5), with the results being displayed in column (1) of

Table 6. Regression results: the impact of ESOP on the dynamic adjustment of capital structure.

Variable	(1) Full Sample	(2) Downward-Adjusted Subsample	(3) Upward-Adjusted Subsample
	${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev	0.440 *** (56.30)	0.294 *** (28.82)	0.207 *** (24.43)
Esop	0.034 *** (4.94)	0.033 ** (2.55)	0.008 (1.38)
Lev × Esop	−0.067 *** (−4.72)	−0.046 ** (−2.11)	0.008 (0.45)
Roa	−0.138 *** (−10.69)	−0.271 *** (−16.50)	−0.159 *** (−12.58)
Size	0.022 *** (10.42)	0.019 *** (7.53)	0.041 *** (17.49)
Growth	0.002 (1.09)	0.001 (0.50)	0.007 *** (4.39)
Tang	0.031 *** (2.59)	0.045 *** (3.05)	0.014 (1.16)
Dep	−0.496 *** (−4.48)	−0.530 *** (−3.91)	−0.002 (−0.02)
Ocf	−0.093 *** (−7.58)	−0.100 *** (−7.17)	−0.065 *** (−4.93)
Ind	−0.015 (−0.63)	0.074 *** (2.72)	0.108 *** (3.84)
Soe	0.008 (1.38)	0.006 (0.83)	0.006 (0.97)
Dual	0.003 (1.22)	−0.004 (−1.49)	0.004 (1.51)
Mpay	0.003 (1.33)	0.003 (1.30)	0.006 *** (2.80)
Top1	−0.032 ** (−2.39)	−0.068 *** (−4.12)	−0.006 (−0.47)
C	−0.257 *** (−5.18)	−0.052 (−0.86)	−0.808 *** (−14.62)
Year FE	yes	yes	yes
Firm FE	yes	yes	yes
Adj-R2	0.304	0.322	0.275
N	19,703	9104	10,599
F-value	316.07 ***	155.69 ***	124.36 ***

Note. The symbols *** and ** indicate significance at the 1% and 5% levels, respectively; the numbers inside the brackets are t values, and the numbers outside the brackets are coefficients; C represents the constant term of the regression models.

. As can be seen from the table, the regression coefficient of the explained variable ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$ on the interact term ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$ × ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ is −0.067 and significant at the 1% level. This indicates that the capital structure adjustment speed of enterprises implementing ESOPs is, on average, 6.7% faster than that of enterprises not implementing ESOPs. This suggests that the implementation of ESOP can markedly accelerate the dynamic adjustment of capital structure, thereby supporting Hypothesis 2. This may be attributed to the fact that the implementation of ESOP can assist in reducing the costs associated with debt and equity financing for enterprises [16,55], and alleviate the constraints faced by enterprises in terms of financing, which in turn reduces the cost of capital structure adjustment for enterprises and consequently accelerates the dynamic adjustment of capital structure.

Previous studies have demonstrated that, in comparison to debt levels below the target capital structure, corporate decision-makers exhibit heightened sensitivity to debt levels exceeding the target capital structure. This observation indicates an asymmetry in the dynamic adjustment of capital structure, whereby the speed of upward adjustment is markedly faster than the speed of downward adjustment [60]. Considering the above evidence, this paper further investigates whether an asymmetry exists in the influence of ESOP on the speed of capital structure adjustment. The sample is initially divided into two groups; namely, the upward-adjusted group (${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$ < ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}^{*}$) and the downward-adjusted group (${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$ > ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}^{*}$), based on the relationship between the actual capital structure at the beginning of the year t + 1 (${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$) and the target capital structure (${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}^{*}$). Subsequently, grouping regression estimations are performed in Equation (5), and the results are presented in columns (2) and (3) of Table 6.

When the actual capital structure is lower than the target capital structure and requires upward adjustment, the ESOP has no significant effect on the speed of capital structure adjustment (the coefficient on ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$ × ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ is 0.008 and not statistically significant). When the actual capital structure is higher than the target capital structure and requires downward adjustment, the ESOP significantly accelerates the dynamic adjustment of capital structure (the coefficient on ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$ × ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ is −0.046 and statistically significant at the 5% level). This indicates that the influence of the ESOP on the speed of dynamic capital structure adjustment is asymmetrical. Specifically, the impact of the ESOP on capital structure adjustment is evident when the capital structure requires downward adjustment, but insignificant when the capital structure requires upward adjustment. This indirectly validates the hypothesis that ESOP implementation can assist in reducing a company’s debt level. From the perspective management, a proclivity towards risk aversion may provide a stronger incentive to adjust the capital structure downward, that results in reduction in leverage. Furthermore, the ESOP may engender an increased risk aversion among executives, given that their personal wealth is more closely aligned with the success of the business. Therefore, the implementation of the ESOP will facilitate a downward adjustment of the capital structure, thereby stabilizing the financial structure and reducing risk exposure. Conversely, there is a reduced incentive for management to recapitalize upwards (i.e., increase leverage). The concern is that excessive leverage will increase the risk of bankruptcy, thereby damaging their human capital and the value of their equity. The ESOP does not significantly alter management’s appetite for risk. Consequently, ESOPs have a limited impact on the upward adjustment of the capital structure.

5.3.3. Robustness Tests

(1)

Propensity scores matching method

To enhance the robustness of the research results, this paper further employs the propensity score matching method to address the endogeneity problem that is caused by sample self-selection bias. First, the Esop variable is taken as the treatment variable, with the result variable being ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$. The control variables of Equation (4) are employed as covariances, and a Logit model is used to estimate the probability of implementing an ESOP for each sample. The probability is taken as the propensity scores, and then the nearest neighbor matching method is used to match the covariate one-to-one. Second, the matching results are evaluated for balance. The results demonstrate that, in comparison to the pre-matching data, the absolute value of the standard deviation of all matching variables is less than 10% following matching, with a notable reduction in deviation. This suggests that the selected matching variables and method are appropriate. Moreover, the t-statistics after matching are not significant, indicating that there are no significant differences in matching variables between the treatment and the control groups after matching, and that the matching result is satisfactory.

Finally, the regression estimations of Models (4) and (5) are performed again using the matched sample, and the results are shown in

Table 7. Regression results based on the PSM sample.

Variable	$\mathbf{Model} \left(4\right): {\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$		$\mathbf{Model} \left(5\right): {\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
	Coefficient	t-Value	Coefficient	t-Value
Lev			0.403 ***	16.71
Esop	−0.009 **	−1.97	0.051 ***	3.92
Lev × Esop			−0.117 ***	−4.66
Roa	−0.701 ***	−12.50	−0.125 ***	−4.15
Size	0.071 ***	29.55	0.031 ***	5.54
Growth	0.031 ***	5.11	−0.000	−0.01
Tang	0.126 ***	4.81	−0.027	−0.74
Dep	−0.622 **	−2.35	−0.205	−0.72
Ocf	−0.319 ***	−6.59	−0.036	−1.17
Ind	0.535 ***	18.80	−0.026	−0.40
Soe	−0.003	−0.48	0.028 **	2.34
Dual	0.011 **	2.30	0.005	0.92
Mpay	−0.011 ***	−2.58	0.004	0.81
Top1	0.026	1.52	0.012	0.36
C	−1.200 ***	−22.67	−0.493 ***	−3.69
Year FE	yes		yes
Firm FE	no		yes
Adj-R2	0.424		0.248
N	4234		4234
F-value	233.12 ***		38.53 ***

Note. The symbols *** and ** indicate significance at the 1% and 5% levels, respectively; C represents the constant term of the regression models.

. The estimated results of Model (4) indicate that the regression coefficient of the explained variable (${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$) on the explanatory variable (Esop) is −0.009 and is significant at the 5% level. This suggests that the implementation of ESOP is negatively correlated with the asset–liability ratio of enterprises. This indicates that the implementation of ESOP reduces the level of corporate debt, thereby supporting Hypothesis 1. According to the estimation results of Model (5), the regression coefficient of the explained variable (${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}$) on the interact term ${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}$ × ${\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}$ is −0.117 and is significant at the 1% level, indicating that the implementation of ESOP speeds up the dynamic adjustment of the capital structure. This provides further support for Hypothesis 2.

(2)

Instrumental variable approach

Table 5 illustrates a notable negative correlation between the implementation of ESOP and the level of corporate debt. Nevertheless, the direction of causality between ESOP and corporate debt remains inconclusive. It is plausible that a reverse causality exists, whereby companies with low debt levels may be more inclined to implement ESOPs. Consequently, we employ the instrumental variable (IV) approach to address the endogeneity resulting from potential reverse causality bias. The implementation of ESOPs among enterprises in the same industry and region is subject to certain imitations, and the implementation of ESOPs by other enterprises in the same industry and region does not directly affect the debt level of the enterprise. Considering the findings presented by Cao et al. [61], this study employs the proportion of other enterprises in the same industry and province that have implemented ESOPs in the previous year as the instrumental variable (Esop_Ind_t−1).

Table 8. Regression results of the two-stage least squares method.

Variable	$\mathbf{The} \mathbf{First} \mathbf{Stage}: {\mathbf{E}\mathbf{s}\mathbf{o}\mathbf{p}}_{\mathbf{i},\mathbf{t}}$		$\mathbf{The} \mathbf{Second} \mathbf{Stage}: {\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
	Coefficient	t-Value	Coefficient	z-Value
Esop_Indt−1	1.147 ***	51.45
Esop			−0.017 *	−1.75
Roa	0.005	0.13	−0.775 ***	−29.53
Size	0.038 ***	16.18	0.064 ***	48.92
Growth	0.005	1.03	0.024	7.62
Tang	−0.234 ***	−9.70	0.003	0.22
Dep	0.848 ***	3.17	−0.768 ***	−4.80
Ocf	−0.169 ***	−7.87	−0.225 ***	−16.94
Ind	−0.234 **	−2.35	0.156 ***	2.85
Soe	−0.418 ***	−3.48	0.135 ***	10.66
Dual	0.025 ***	4.10	0.002	0.62
Mpay	−0.003	−0.08	−0.006 **	−2.54
Top1	−0.139 ***	−7.88	−0.017 *	−1.95
C	−0.636 ***	−8.39	−0.885 ***	−22.21
Year FE	yes		yes
Adj-R2	0.197		0.407
N	17,370		17,370
F-value/Chi2	51.92 ***		14,106.99 ***

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; C represents the constant term of the regression models.

presents the regression results obtained through the two-stage least squares method. This indicates that, after addressing the potential reverse causality bias, the implementation of ESOP is still negatively correlated with the level of corporate debt at the 10% level, thereby providing further support for Hypothesis 1.

(3)

Replacing the calculation method of capital structure

In line with numerous studies on capital structure in China [1,2,3], this research employs the book value of equity to ascertain the firm’s capital structure. However, in the context of Western developed capital markets, the prevailing approach among researchers is to utilize the market value of equity in the calculation of a firm’s capital structure [58]. Accordingly, the market debt ratio (MDR) is employed as a further measure of the firm’s capital structure, with the main regression analysis then being re-performed. The MDR represents the ratio of total debt to the total market value of assets. The total market value of assets is calculated by adding the book value of debt to the market value of equity.

Table 9. Regression results based on market debt ratio.

Variable	$\mathbf{Model} \left(4\right): {\mathbf{M}\mathbf{D}\mathbf{R}}_{\mathbf{i},\mathbf{t}+1}$		$\mathbf{Model} \left(5\right): {\mathbf{M}\mathbf{D}\mathbf{R}}_{\mathbf{i},\mathbf{t}+1}$
	Coefficient	t-Value	Coefficient	t-Value
MDR			0.436 ***	36.43
Esop	−0.012 ***	−4.28	−0.010 **	−2.51
MDR × Esop			−0.068 ***	−4.62
Roa	−0.498 ***	−25.20	−0.045 ***	−2.97
Size	0.099 ***	93.95	0.023 ***	6.72
Growth	0.003	1.16	−0.001	−0.77
Tang	0.098 ***	9.25	0.002	0.12
Dep	−1.003 ***	−8.74	−0.467 ***	−3.06
Ocf	−0.277 ***	−16.32	−0.086 ***	−6.69
IndMDR	0.413 ***	43.78	0.050 ***	3.10
Soe	0.020 ***	8.21	0.010 *	1.79
Dual	−0.003	−1.44	−0.002	−0.68
Mpay	−0.027 ***	−14.96	0.004	1.37
Top1	−0.064 ***	−9.22	−0.013	−0.75
C	−1.654 ***	−67.99	−0.464 ***	−5.67
Year FE	yes		yes
Firm FE	no		yes
Adj-R2	0.632		0.841
N	18,821		18,821
F-value	1741.52 ***		441.78 ***

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; C represents the constant term of the regression models.

presents the regression results for Models (4) and (5) based on MDR. The results demonstrate that the coefficient on ESOP in Model (4) and the coefficient on MDR × ESOP in Model (5) are both negative and statistically significant at the 1% level. This suggests that the implementation of ESOP can reduce a firm’s debt level and facilitate the dynamic adjustment of its capital structure, thereby providing further support for our hypotheses.

(4)

Replacing the proxy for the industry capital structure

In the above analysis, the median debt level within the same industry (Ind) for the same year was employed as a means of controlling for industry capital structure. Nevertheless, the correlation between Ind and Lev is considerable (0.418), which may give rise to a multicollinearity issue in the capital structure adjustment model. Therefore, the industry median debt level was replaced with the industry fixed effects to control for industry capital structure, and the main regression analysis was then re-performed (see

Table 10. Regression results of replacing the proxy for the industry capital structure.

Variable	$\mathbf{Model} \left(4\right): {\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$		$\mathbf{Model} \left(5\right): {\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
	Coefficient	t-Value	Coefficient	t-Value
Lev			0.452 ***	35.79
Esop	−0.007 **	−2.20	0.029 ***	3.57
Lev × Esop			−0.055 ***	−3.15
Roa	−0.821 ***	−30.33	−0.117 ***	−5.71
Size	0.064 ***	51.84	0.016 ***	4.49
Growth	0.031 ***	9.92	0.001	0.55
Tang	0.061 ***	4.20	0.028	1.60
Dep	−0.430 ***	−2.70	−0.423 **	−2.24
Ocf	−0.259 ***	−10.96	−0.089 ***	−5.52
Soe	0.017 ***	5.63	0.013 *	1.72
Dual	0.004	1.33	0.001	0.47
Mpay	−0.009 ***	−4.11	0.005	1.60
Top1	−0.041 ***	−4.92	−0.031	−1.56
C	−0.802 ***	−23.97	−0.176 **	−2.03
Industry FE	yes		yes
Year FE	yes		yes
Firm FE	no		yes
Adj-R2	0.386		0.717
N	19,703		19,703
F-value	237.88 ***		441.78 ***

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; C represents the constant term of the regression models.

). The results demonstrate that the exclusion of the independent variable Ind from the empirical models does not alter our findings.

5.4. Further Analysis

The incentive and governance effects of ESOP have been demonstrated to vary in accordance with the internal and external environments of enterprises and the contractual design of ESOPs [26,61]. Accordingly, this paper further investigates the heterogeneity of the impact of ESOP on capital structure decisions from four perspectives: macro environment, industry characteristics, corporate governance, and ESOP contract design. First, the economic environment and market conditions are employed as proxies for the macro environment. Second, two significant industry characteristics are considered, namely industry concentration and industry technology content (i.e., whether the industry in question can be classified as high-tech). Third, two governance variables are selected to proxy for the internal governance environment of corporations. These are the nature of enterprise property rights and the degree of ownership concentration. Ultimately, in accordance with the Guiding Opinions, our investigation is concentrated on the impact of specific elements within the ESOP contract design, including the source of funding, the source of stock, the management mode, the lock-up period, the scale of shareholding, and the executive subscription ratio.

5.4.1. The Moderating Effect of the Macro Environment

(1)

The moderating role of the economic environment

To ascertain whether the influence of the ESOP on capital structure decisions is contingent upon the macroeconomic environment, we initially divide the comprehensive sample into two distinct groups based on China’s annual per capita GDP growth rate. The initial cohort comprises observations from the years 2019 and 2020, and is characterized by a rate of annual per capita GDP growth that is less than 6%. The second group comprises observations from 2014 to 2018, representing rapid economic growth, with annual per capita GDP growth exceeding 6%. Subsequently, group regression estimations were performed on Models (4) and (5), respectively. The results are presented in

Table 11. The impact of ESOP on capital structure decisions under different economic environment.

Variable	Slow Economic Growth		Rapid Economic Growth
	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		−0.093 **		0.328 ***
		(−2.25)		(22.06)
Esop	−0.006	0.018	−0.009 **	0.027 ***
	(−1.06)	(0.68)	(−2.23)	(3.16)
Lev × Esop		−0.029		−0.045 **
		(−0.54)		(−2.38)
Roa	0.059 ***	0.023 **	0.067 ***	0.018 ***
	(26.980)	(2.173)	(44.208)	(4.156)
Size	−0.776 ***	−0.024	−0.842 ***	−0.181 ***
	(−19.015)	(0.711)	(−22.647)	(−6.411)
Growth	0.042 ***	−0.005	0.026 ***	0.002
	(6.554)	(−0.921)	(7.408)	(0.858)
Tang	0.038	−0.018	0.050 ***	0.028
	(1.533)	(−0.501)	(2.831)	(1.297)
Dep	−0.004	0.191	−0.495 **	−0.757 ***
	(−0.015)	(0.705)	(−2.506)	(−3.086)
Ocf	−0.244 ***	−0.108 ***	−0.273 ***	−0.081 ***
	(−5.776)	(−3.487)	(−9.535)	(−4.004)
Ind	0.141	0.181 ***	0.224 ***	0.089 **
	(0.731)	(3.739)	(3.692)	(2.462)
Dual	0.011 **	0.008	0.018 ***	0.002
	(2.074)	(0.684)	(5.003)	(0.151)
Mpay	0.001	0.009	0.006 *	0.001
	(0.250)	(1.318)	(1.689)	(0.237)
Top1	−0.013 ***	−0.001	−0.005 *	0.006
	(−3.103)	(−0.185)	(−1.794)	(1.478)
C	−0.694 ***	−0.077	−1.052 ***	−0.209 **
	(−7.167)	(−0.304)	(−22.654)	(−2.055)
Year FE	yes	yes	yes	yes
Firm FE	no	yes	no	yes
F-Value	56.10 ***	8.75 ***	98.89 ***	59.98 ***
Adj-R2	0.389	0.011	0.405	0.752
N	6364	6364	13,339	13,339

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; the numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients; C represents the constant term of the regression models.

.

The results in Table 11 demonstrate that in the capital structure choice model (i.e., Model (4)), the implementation of ESOP significantly reduces the level of corporate debt in the rapid economic growth group (coefficient = −0.009, t-value = −2.23), yet exerts insignificant influence on corporate debt in the slow economic growth group (coefficient = −0.006, t-value = −1.06). In the capital structure adjustment model (i.e., Model (5)), the implementation of the ESOP significantly speeds up the dynamic adjustment of capital structure in the rapid economic growth group (coefficient = −0.045, t-value = −2.38), but has no significant impact on the dynamic adjustment of capital structure in the slow economic growth group (coefficient = −0.029, t-value = −0.54). The above results suggest that the optimization effect of ESOP on capital structure decisions is more pronounced in an economic boom environment. This may be attributed to the fact that a favorable economic climate enhances employees’ confidence in the company’s prospects, thereby motivating them to participate more actively in ESOPs. This, in turn, helps to reduce agency costs and financing costs for businesses, thereby optimizing the capital structure decisions of enterprises.

(2)

The moderating role of the market climate

To ascertain whether the influence of ESOP on capital structure decisions differs contingent on the market climate, the full sample was initially divided into two groups based on the annual CSI 300 (an index of large companies listed in Shanghai and Shenzhen). The initial group comprises observations from years in which the market climate was favorable (2014, 2015, 2017, 2019 and 2020), characterized by an annual closing index that exceeded the opening index. The second group comprises observations from the years 2016 and 2018, which were characterized by a poor market climate. In these years, the annual closing index was lower than the opening index. Subsequently, group regression estimations were performed on Models (4) and (5), respectively. The results are presented in

Table 12. The impact of ESOP on capital structure decisions under different market climate.

Variable	Poor Market Climate		Good Market Climate
	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		0.330 ***		0.526 ***
		(14.67)		(26.02)
Esop	−0.008 *	0.066 ***	−0.006	−0.003
	(−1.76)	(4.76)	(−1.29)	(−0.27)
Lev × Esop		−0.103 ***		−0.003
		(−3.62)		(−0.12)
Roa	0.067 ***	0.029 ***	0.062 ***	0.015 ***
	(34.95)	(4.74)	(38.33)	(3.21)
Size	−0.797 ***	−0.253 ***	−0.830 ***	−0.109 ***
	(−17.83)	(−7.48)	(−24.41)	(−3.60)
Growth	0.027 ***	0.003	0.034 ***	0.002
	(6.11)	(1.01)	(7.76)	(0.54)
Tang	0.056 **	0.069 **	0.062 ***	0.012
	(2.43)	(2.24)	(3.35)	(0.48)
Dep	−0.500 *	−0.894 ***	−0.370 *	−0.490 **
	(−1.94)	(−2.81)	(−1.85)	(−1.99)
Ocf	−0.247 ***	−0.067 **	−0.273 ***	−0.114 ***
	(−6.53)	(−2.36)	(−9.02)	(−5.11)
Ind	0.132	0.011	0.414 ***	0.123 ***
	(1.54)	(0.24)	(6.16)	(3.25)
Dual	0.020 ***	−0.006	0.014 ***	0.027 ***
	(4.38)	(−0.33)	(3.58)	(2.82)
Mpay	0.005	0.005	0.002	0.005
	(1.27)	(0.87)	(0.60)	(1.26)
Top1	−0.008 **	0.005	−0.010 ***	0.005
	(−2.36)	(1.05)	(−3.35)	(1.33)
C	−0.961 ***	−0.416 ***	−0.911 ***	−0.249 **
	(−16.14)	(−2.95)	(−17.86)	(−2.16)
Year FE	yes	yes	yes	yes
Firm FE	no	yes	no	yes
F-Value	88.73 ***	39.60 ***	143.45 ***	96.30 ***
Adj-R2	0.393	0.716	0.386	0.780
N	8161	8161	11,542	11,542

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; The numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients; C represents the constant term of the regression models.

.

The results in Table 12 show that in the capital structure choice model, the implementation of an ESOP significantly reduces the level of corporate debt by 0.8% in the poor market climate group (coefficient = −0.008, t-value = −1.76), yet exerts insignificant influence on corporate debt in the good market climate group (coefficient = −0.006, t-value = −1.29). In the capital structure adjustment model, the implementation of an ESOP significantly speeds up the dynamic adjustment of capital structure by 10.3% in the poor market climate group (coefficient = −0.103, t-value = −3.62), but has no significant impact on the dynamic adjustment of capital structure in the good market climate group (coefficient = −0.003, t-value = −0.12). The above results indicate that the optimization effect of ESOP on capital structure decisions is more pronounced in a poor market climate. When the market climate is poor, the financing of enterprises is more challenging, the capital cost of equity financing is higher, and the debt financing is constrained. In such circumstances, ESOP is more beneficial for enterprises in reduce financing pressure and optimizing capital structure.

5.4.2. The Moderating Effect of Industry Characteristics

(1)

The moderating role of industry concentration

To ascertain whether the influence of an ESOP on capital structure decisions is contingent upon industry concentration, the full sample was initially divided into two groups based on the Herfindahl index, calculated by operation revenue. One group is designated as the low industry concentration group (below the median value of the Herfindahl index), and the other is designated as the high industry concentration group (above the median value of the Herfindahl index). Subsequently, group regression estimations were performed on Models (4) and (5), respectively. The results are presented in

Table 13. The impact of ESOP on capital structure decisions under different industry concentration.

Variable	Industries with High Concentration		Industries with Low Concentration
	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		0.409 ***		0.412 ***
		(18.22)		(19.28)
Esop	−0.008	0.019	−0.010 **	0.036 ***
	(−1.56)	(1.45)	(−1.96)	(3.23)
Lev × Esop		−0.034		−0.077 ***
		(−1.22)		(−3.09)
Roa	0.068 ***	0.020 ***	0.066 ***	0.009
	(35.16)	(3.53)	(34.05)	(1.60)
Size	−0.793 ***	−0.093 ***	−0.835 ***	−0.145 ***
	(−15.64)	(−2.93)	(−20.27)	(−4.93)
Growth	0.031 ***	0.002	0.032 ***	0.001
	(6.62)	(0.51)	(6.36)	(0.34)
Tang	0.034	0.017	0.085 ***	0.043
	(1.49)	(0.60)	(3.93)	(1.47)
Dep	−0.986 ***	−0.521 *	−0.201	−0.726 **
	(−4.06)	(−1.72)	(−0.86)	(−2.34)
Ocf	−0.258 ***	−0.109 ***	−0.265 ***	−0.044
	(−6.99)	(−4.12)	(−7.39)	(−1.63)
Ind	0.093	0.048	0.433 ***	0.041
	(1.19)	(1.07)	(3.28)	(0.52)
Dual	0.003	0.010	0.031 ***	0.030 **
	(0.59)	(0.77)	(6.66)	(2.19)
Mpay	0.005	−0.002	0.002	0.006
	(1.08)	(−0.50)	(0.51)	(1.28)
Top1	0.002	−0.001	−0.015 ***	0.011 **
	(0.53)	(−0.13)	(−4.34)	(2.35)
C	−1.072 ***	−0.170	−0.960 ***	−0.127
	(−18.44)	(−1.38)	(−13.27)	(−0.87)
Year FE	yes	yes	yes	yes
Firm FE	no	yes	no	yes
F-Value	99.42 ***	34.33 ***	113.50 ***	43.73 ***
Adj-R2	0.437	0.778	0.432	0.752
N	7621	7621	7846	7846

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; the numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients; C represents the constant term of the regression models.

.

The results in Table 13 demonstrate that in the capital structure choice model, the implementation of ESOP significantly reduces the level of corporate debt in the low industry concentration group (coefficient = −0.010, t-value = −1.96), but has insignificant impact on corporate debt in the high industry concentration group (coefficient = −0.008, t-value = −1.56). In the capital structure adjustment model, the implementation of ESOP significantly speeds up the dynamic adjustment of capital structure in the low industry concentration group (coefficient = −0.077, t-value = −3.09), but has no significant impact on the dynamic adjustment of capital structure in the high industry concentration group (coefficient = −0.034, t-value = −1.22). The above results indicate that the optimization effect of ESOP on capital structure decisions is more pronounced in competitive industries. This may be because highly competitive industries tend to have substantial capital requirements and limited financing channels. ESOP can serve as a supplementary channel for enterprise financing, which can alleviate the financial pressure on enterprises and optimize their capital structure decisions.

(2)

The moderating role of industry technology content

To ascertain whether the influence of ESOP on capital structure decisions is contingent upon the technological content of the industry in question, the full sample was initially divided into two distinct groups: high-tech industries (comprising 19 industries, as defined by the CSRC industry classification, and identified by industry codes C25-C26, C27-C29, C31-C41, I63-I65, and M73) and non-high-tech industries. Subsequently, group regression estimations were performed on Models (4) and (5), respectively. The results are presented in

Table 14. The impact of ESOP on capital structure decisions under different industry technology content.

Variable	High-Tech Industries		Non-High-Tech Industries
	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		0.440 ***		0.436 ***
		(26.82)		(19.96)
Esop	−0.014 ***	0.022 **	0.002	0.037 ***
	(−3.52)	(2.39)	(0.34)	(2.59)
Lev × Esop		−0.038 *		−0.074 ***
		(−1.77)		(−2.60)
Roa	0.068 ***	0.016 ***	0.061 ***	0.022 ***
	(39.75)	(3.47)	(33.24)	(3.81)
Size	−0.755 ***	−0.098 ***	−0.863 ***	−0.126 ***
	(−22.60)	(−4.01)	(−18.67)	(−3.37)
Growth	0.031 ***	−0.003	0.029 ***	0.003
	(7.25)	(−0.98)	(6.44)	(0.91)
Tang	0.090 ***	0.010	0.012	0.032
	(4.25)	(0.40)	(0.61)	(1.15)
Dep	−0.322	−0.115	−0.496 **	−0.792 **
	(−1.57)	(−0.49)	(−1.97)	(−2.35)
Ocf	−0.310 ***	−0.073 ***	−0.193 ***	−0.090 ***
	(−9.78)	(−3.53)	(−5.53)	(−3.60)
Ind	0.351 ***	0.165 ***	0.208 ***	0.074 *
	(5.20)	(3.97)	(2.86)	(1.86)
Dual	0.034 ***	0.023 **	−0.001	0.001
	(7.95)	(2.50)	(−0.13)	(0.08)
Mpay	−0.005	−0.002	0.019 ***	0.004
	(−1.52)	(−0.56)	(3.91)	(0.78)
Top1	−0.018 ***	0.006	0.005	0.005
	(−5.95)	(1.49)	(1.51)	(1.06)
C	−1.001 ***	−0.264 **	−1.008 ***	−0.318 **
	(−18.46)	(−2.39)	(−18.25)	(−2.34)
Year FE	yes	yes	yes	yes
Firm FE	no	yes	no	yes
F-Value	104.86 ***	78.74 ***	98.13 ***	42.85 ***
Adj-R2	0.358	0.755	0.413	0.679
N	11,452	11,452	8251	8251

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; the numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients; C represents the constant term of the regression models.

.

The results in Table 14 show that in the capital structure choice model, the implementation of ESOPs significantly reduces the level of corporate debt in high-tech industries (coefficient = −0.014, t-value = −3.52), but has no significant impact on corporate debt in non-high-tech industries (coefficient = 0.002, t-value = 0.34). In the capital structure adjustment model, the implementation of ESOPs speeds up the dynamic adjustment of capital structure by 7.4% in non-high-tech industries, but only 3.8% in high-tech industries. The above results indicate that the reduction effect of ESOP on corporate debt is more pronounced in high-tech industries, whereas the acceleration effect of ESOP on dynamic capital structure adjustment is stronger in non-high-tech industries. On the one hand, external financing for high-tech companies is more challenging, necessitating a reliance on internal accumulation. An ESOP has the potential to stimulate employee enthusiasm, enhance company profitability and strengthen the capacity for self-accumulation. ESOPs implemented by high-tech companies are more likely to result in a reduction of debt levels. On the other hand, the inherent uncertainty and the necessity for continuous large input needs inherent to the high-tech industry make it challenging for the company to adjust its capital structure in a timely manner in accordance with the actual situation. The incentive effect of an ESOP may be subject to a time lag, which could potentially impede the company’s capacity to make timely adjustments to its capital structure.

5.4.3. The Moderating Effect of Corporate Governance

(1)

The moderating role of ownership nature

The classification of enterprises is contingent upon the nature of property rights. This leads to the distinction between state-owned enterprises and non-state-owned enterprises. In the context of China’s capital market, state-owned enterprises are observed to possess intrinsic informational advantages, relying on the government and demonstrating enhanced financial capabilities [62]. Conversely, state-owned enterprises are subject to more onerous conditions and constraints pertaining to the implementation of ESOP. Moreover, the subscription ratio in their ESOPs is comparatively lower, which may result in a diminished incentive effect of ESOPs in state-owned enterprises relative to that observed in private enterprises [27]. It is therefore necessary to examine whether the impact of ESOP on capital structure decisions depends on the nature of enterprise property rights. The initial step is to divide the entire sample into two distinct groups: state-owned enterprises and non-state-owned enterprises. Subsequently, group regression estimations were performed on Models (4) and (5), respectively. The results are presented in

Table 15. The impact of ESOP on capital structure decisions under different ownership nature.

Variable	State-Owned Enterprises		Non-State-Owned Enterprises
	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		0.466 *** (35.96)		0.411 *** (41.35)
Esop	−0.001 (−0.12)	0.043 ** (1.98)	−0.012 *** (−3.45)	0.023 *** (2.96)
Lev × Esop		−0.084 ** (−2.54)		−0.043 ** (−2.14)
Roa	−1.320 *** (−21.38)	−0.140 *** (−5.87)	−0.810 *** (−25.38)	−0.127 *** (−8.05)
Size	0.066 *** (38.50)	0.031 *** (9.34)	0.074 *** (47.90)	0.014 *** (5.17)
Growth	0.045 *** (7.85)	−0.002 (−0.77)	0.040 *** (10.04)	0.004 ** (1.97)
Tang	0.014 (0.84)	0.027 * (1.67)	0.080 *** (4.79)	0.032 * (1.92)
Dep	−0.952 *** (−4.56)	−0.372 ** (−2.38)	0.457 *** (2.65)	−0.683 *** (−4.60)
Ocf	−0.102 ** (−2.39)	−0.101 *** (−5.68)	−0.191 *** (−6.82)	−0.090 *** (−5.56)
Ind	0.367 *** (18.85)	−0.019 (−0.60)	0.520 *** (27.02)	0.006 (0.17)
Dual	0.004 (0.62)	−0.004 (−1.03)	0.002 (0.63)	0.005 * (1.74)
Mpay	−0.023 *** (−6.20)	0.002 (0.52)	−0.011 *** (−4.33)	0.002 (0.75)
Top1	−0.108 *** (−7.72)	−0.037 ** (−2.03)	0.042 *** (4.39)	−0.013 (−0.72)
C	−0.719 *** (−14.47)	−0.422 *** (−5.32)	−1.285 *** (−39.05)	−0.104 (−1.61)
Year FE	yes	yes	yes	yes
Firm FE	no	yes	no	yes
F-Value	301.19 ***	125.45 ***	568.33 ***	207.73 ***
Adj-R2	0.415	0.332	0.406	0.293
N	6816	6816	12,887	12,887

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; The numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients; C represents the constant term of the regression models.

.

As evidenced in Table 15, in the capital structure choice model, the implementation of ESOP in non-state-owned enterprises can significantly reduce the level of corporate debt (coefficient = −0.012, t-value = −3.45). Conversely, the implementation of ESOP in state-owned enterprises has no significant impact on corporate debt (coefficient = −0.001, t-value = −0.12). In the capital structure adjustment model, the implementation of ESOP in state-owned enterprises accelerates the dynamic adjustment of capital structure by 8.4%, while the implementation of ESOP in non-state-owned enterprises accelerates the dynamic adjustment of capital structure by only 4.3%. The above results suggest that there are notable discrepancies in the impact of ESOP on capital structure decisions between state-owned enterprises and non-state-owned enterprises. Specifically, ESOPs implemented in non-state-owned enterprises have a stronger effect on reducing financial leverage, but have a weaker effect on accelerating the dynamic adjustment of capital structure, in comparison to their effect on state-owned enterprises.

(2)

The moderating role of ownership concentration

The distribution of equity ownership among listed companies in the United States is relatively dispersed, whereas in China, it is relatively concentrated. In the context of concentrated equity, it is common for large shareholders to advance their interests at the expense of small and medium shareholders in Chinese listed companies. The extant literature demonstrates that major shareholders engage in self-serving actions to reduce their holdings [63]. The institutional framework of ESOP can fulfill the financial objectives of major shareholders in divesting their shares [64]. Consequently, for enterprises with relatively concentrated equity, the incentive of ESOP is inherently limited, which is not conducive to enhancing enterprise performance and challenging the reduction of debt levels. Conversely, some studies have identified that, to mitigate the risk of external acquisition, enterprises with relatively dispersed equity are more inclined to promote the implementation of ESOPs and enhance their external financing capabilities [65]. It is therefore necessary to test whether the impact of ESOP on capital structure decisions depends on the concentration of ownership in enterprises. The median equity concentration of the sample companies was used to divide the entire sample into two groups: one comprising companies with relatively concentrated equity and the other with relatively dispersed equity. Group regression estimations were then performed on Models (4) and (5), respectively. The results are presented in

Table 16. The impact of ESOP on capital structure decisions under different ownership concentration.

Variable	Companies with Relatively Concentrated Ownership		Companies with Relatively Dispersed Ownership
	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(4\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(5\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		0.420 *** (33.92)		0.408 *** (34.81)
Esop	0.003 (0.65)	0.057 *** (5.97)	−0.021 *** (−4.82)	0.017 * (1.65)
Lev × Esop		−0.104 *** (−5.34)		−0.048 ** (−2.12)
Roa	−1.127 *** (−25.69)	−0.164 *** (−7.45)	−0.674 *** (−18.09)	−0.103 *** (−5.96)
Size	0.064 *** (43.25)	0.023 *** (7.04)	0.071 *** (37.70)	0.025 *** (7.17)
Growth	0.050 *** (10.76)	0.000 (0.13)	0.031 *** (6.38)	0.002 (0.85)
Tang	−0.013 (−0.86)	0.022 (1.27)	0.052 *** (2.58)	0.054 *** (2.89)
Dep	−0.013 (−0.86)	−0.518 *** (−3.31)	0.460 ** (2.07)	−0.443 *** (−2.65)
Ocf	−0.196 *** (−6.43)	−0.060 *** (−3.67)	−0.346 *** (−9.58)	−0.113 *** (−5.96)
Ind	0.422 *** (23.50)	−0.039 (−1.12)	0.451 *** (20.48)	−0.047 (−1.26)
Soe	−0.007 * (−1.85)	−0.003 (−0.14)	0.024 *** (5.29)	0.003 (0.40)
Dual	−0.001 (−0.40)	0.001 (0.39)	0.008 ** (2.06)	0.007 * (1.90)
Mpay	0.001 (0.32)	0.003 (1.01)	−0.022 *** (−6.76)	0.008 ** (2.34)
Top1	−0.058 *** (−3.81)	−0.081 *** (−3.67)	0.065 ** (2.26)	−0.033 (−0.81)
C	−1.076 *** (−29.93)	−0.230 *** (−2.92)	−1.021 *** (−23.04)	−0.388 *** (−4.51)
Year FE	yes	yes	yes	yes
Firm FE	no	yes	no	yes
F-Value	605.63 ***	129.12 ***	311.11 ***	131.90 ***
Adj-R2	0.4686	0.2715	0.3533	0.2771
N	9847	9847	9856	9856

Note. The symbols ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively; the numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients; C represents the constant term of the regression models.

.

As evidenced in Table 16, in the capital structure choice model, the implementation of an ESOP has insignificant impact on corporate debt in the relatively concentrated equity group (coefficient = 0.003, t-value = −0.65), but has significant negative impact on corporate debt in the relatively dispersed group (coefficient = −0.021, t-value = −4.82). In the capital structure adjustment model, the implementation of ESOP accelerates the dynamic adjustment of capital structure by 10.4% in the relatively concentrated equity group, but only by 4.8% in the relatively dispersed group. The above results indicate that ownership concentration has the potential to significantly affect the impact of ESOP on capital structure decisions. Specifically, the reduction effect of ESOP on financial leverage is more pronounced when ownership is relatively dispersed, whereas the acceleration effect of ESOP on dynamic capital structure adjustment is more evident when ownership is relatively concentrated.

5.4.4. The Moderating Effect of ESOP Contract Design

To test whether the impact of ESOP on capital structure decisions differs due to the specific designs of ESOP contracts, this paper employs a set of proxy variables derived from six contract elements; namely, funding source, stock source, management mode, lock-up period, shareholding size, and executive subscription ratio. These variables are used as the core explanatory variables in the expansion of Models (4) and (5), as follows:

$${\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}={\mathsf{\alpha }}_0+\sum {{\mathsf{\alpha }}_{\mathrm{j}}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{t}\_\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}}_{\mathrm{i},\mathrm{j},\mathrm{t}}+\mathsf{\beta }{\mathbf{C}\mathbf{o}\mathbf{n}\mathbf{t}\mathbf{r}\mathbf{o}\mathbf{l}\mathbf{s}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^5$$

(6)

$$\begin{array}{c}{\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}+1}={\mathsf{\gamma }}_0+{\mathsf{\gamma }}_1{\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\gamma }}_2{\mathrm{E}\mathrm{s}\mathrm{o}\mathrm{p}}_{\mathrm{i},\mathrm{t}}+\sum {{\mathsf{\theta }}_{\mathrm{j}}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{t}\_\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}}_{\mathrm{i},\mathrm{j},\mathrm{t}}+\\ \sum {{\mathsf{\theta }}_{\mathrm{j}}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{t}\_\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}}_{\mathrm{i},\mathrm{j},\mathrm{t}}\times {\mathrm{L}\mathrm{e}\mathrm{v}}_{\mathrm{i},\mathrm{t}}+\mathsf{\beta }{\mathbf{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\lambda }}_{\mathrm{i}}{+\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}+1}^6\end{array}$$

(7)

where Contract_factor represents the contract elements of the ESOP, which are replaced by the funding source, stock source, management mode, lock-up period, shareholding size, and executive subscription ratio. According to the Guiding Opinions and the established practice of ESOP, the typical funding sources of ESOP encompass self-raised funds by employees, bonus funds set aside by the company from its profits, leveraged financing and major shareholder loans. The common stock sources of ESOP include purchasing the firm’s shares from the secondary market, subscribing to non-public offering shares and repurchasing the firm’s shares. The management modes of ESOP include entrusted management and self-management. Accordingly, this paper proposes four indicators (fund1, fund2, fund3, and fund4) for the four types of funding sources, three indicators (stock1, stock2, and stock3) for the three types of stock sources, and two indicators (mode1 and mode2) for the two types of management modes. Furthermore, the lock-up period is classified into three categories: one-year, one-to-two-year, and two-to-three-year periods, with three indicators (lockup1, lockup2, and lockup3) created for each. The shareholding scale is divided into two categories: 1–3% and 3–10%, with two indicators (shares1 and shares2) designed for each. The executive subscription ratio is divided into two categories: 0–30% and 30–100%. Two dummy variables (purchase1 and purchase2) are designed for these categories, respectively. The classification of contract element variables and the definitions of proxy variables in detail are shown in

Table 17. Definitions of ESOP contract element variables.

Contract Element	Symbol	Definition
fund sources	fund1	It is equal to 1 when the entire ESOP subscription fund is raised through employee compensation and self-financing and is otherwise equal to 0.
	fund2	It is equal to 1 when part or all the ESOP subscription funds are raised through the bonus fund and is otherwise equal to 0.
	fund3	It is equal to 1 when part or all the ESOP subscription funds are raised through leverage financing and is otherwise equal to 0.
	fund4	It is equal to 1 when part or all the ESOP subscription funds are raised by major shareholder loans and is otherwise equal to 0.
stock sources	stock1	It is equal to 1 when the underlying ESOP stock is resolved by purchasing the company’s stock in the secondary market and is otherwise equal to 0.
	stock2	It is equal to 1 when the underlying ESOP stock is resolved by participating in the subscription of non-public offering shares and is otherwise equal to 0.
	stock3	It is equal to 1 when the underlying ESOP shares are resolved by first repurchasing the Company’s shares in the secondary market and then transferring them to the ESOP account and is otherwise equal to 0.
management mode	mode1	It is equal to 1 when the company entrusts a third-party asset management institution or trust institution to manage the ESOP and is otherwise equal to 0.
	mode2	It is equal to 1 when the company establishes its own management organization to manage the ESOP and is otherwise equal to 0.
lockup period	lockup1	It is equal to 1 when the lockup period of the ESOP is equal to one year and is otherwise equal to 0.
	lockup2	It is equal to 1 when the lockup period of the ESOP is greater than 1 year and less than or equal to 2 years and is otherwise equal to 0.
	lockup3	It is equal to 1 when the lockup period of the ESOP is greater than 2 years and less than or equal to 3 years and is otherwise equal to 0.
shareholding scale	shares1	It is equal to 1 when the proportion of ESOP shares in the total share capital of the company is greater than 0 and less than or equal to 3% and is otherwise equal to 0.
	shares2	It is equal to 1 when the proportion of shares held by ESOP is greater than 3% and less than or equal to 10% of the total share capital of the parent company and is otherwise equal to 0.
executive subscription ratio	purchase1	It is equal to 1 when the share of the ESOP subscribed by the senior management is less than or equal to 30% and is otherwise equal to 0.
	purchase2	It is equal to 1 when the share of ESOP subscribed by senior executives is greater than 30% and is otherwise equal to 0.

.

It is important to highlight that Model (6) is used to test the impact of ESOP contract design on capital structure choice, while Model (7) is used to test the impact of ESOP contract design on capital structure adjustment. Like Models (4) and (5), both Models (6) and (7) use companies that do not implement ESOPs as a reference group. Consequently, the constant term ${\alpha }_0$ in Model (6) represents the average level of capital structure observed in companies that do not implement ESOPs. The coefficient ${\mathsf{\gamma }}_1$ of Lev in Model (7) reflects the average speed of dynamic capital structure adjustment of non-ESOP firms (1 − ${\mathsf{\gamma }}_1$). Furthermore, this paper focuses on the most prevalent categories when classifying capital sources, stock sources and lock-up periods. In the event of a rare case in practice (including ESOPs with multiple combinations of capital sources, gifts of stock sources by major shareholders, or lock-up periods exceeding 3 years), the relevant data will be excluded from the estimation of Models (6) and (7).

Table 18. The impact of fund source, stock source, and management mode of ESOP on capital structure decisions.

Variable	(1) Funding Sources		(2) Stock Sources		(3) Management Mode
	$\mathbf{Model} \left(6\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(7\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(6\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(7\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(6\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(7\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		0.451 ***		0.450 ***		0.451 ***
		(61.10)		(60.92)		(61.14)
fund1	−0.001	0.008
	(−0.36)	(1.16)
fund2	−0.020	0.024
	(−1.55)	(0.91)
fund3	−0.009	−0.002
	(−0.50)	(−0.05)
fund4	−0.019 ***	0.030 **
	(−3.34)	(2.52)
Lev × fund1		−0.021
		(−1.48)
Lev × fund2		−0.020
		(−0.35)
Lev × fund3		0.050
		(0.65)
Lev × fund4		−0.066 **
		(−2.35)
stock1			−0.001	0.011
			(−0.19)	(1.59)
stock2			−0.017 ***	0.026 *
			(−2.63)	(1.85)
stock3			−0.041 ***	−0.009
			(−2.84)	(−0.49)
Lev × stock1				−0.025 *
				(−1.82)
Lev × stock2				−0.051 *
				(−1.76)
Lev × stock3				0.017
				(0.36)
mode1					0.000	0.014 **
					(0.11)	(2.11)
mode2					−0.026 ***	0.011
					(−4.58)	(0.96)
Lev × mode1						−0.029 **
						(−2.13)
Lev × mode2						−0.019
						(−0.75)
Year FE	yes	yes	yes	yes	yes	yes
Firm FE	no	yes	no	yes	no	yes
N	20,136	20,136	20,064	20,064	20,162	20,162
F-value	769.12 ***	289.60 ***	810.13 ***	313.36 ***	851.79 ***	342.66 ***
Adj-R2	0.397	0.313	0.397	0.314	0.397	0.313

Note. Due to space limitations, the regression results of constant terms and control variables are ignored in the table. The symbols ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients.

presents the regression results pertaining to the impact of funding sources, stock sources and management modes of ESOP on capital structure decisions. Regarding the impact of capital source, ESOPs which solve the capital source by borrowing from major shareholders can significantly reduce the level of corporate debt (the coefficient on fund4 in the capital structure choice model is −0.019 and significant at the 1% level) and speed up the dynamic adjustment of capital structure (the coefficient on Lev × fund4 in the capital structure adjustment model is −0.066 and significant at 5% level). However, ESOPs implemented in the remaining three capital sources do not exert statistically significant influence on capital structure choice or capital structure adjustment.

In terms of the impact of stock sources, ESOPs which solve stock sources by subscription to non-public offering shares can not only significantly reduce financial leverage (the coefficient on stock2 in the capital structure choice model is −0.017 and significant at the 1% level), but also significantly accelerate the speed of capital structure adjustment (the coefficient on Lev × stock2 in the capital structure adjustment model is −0.051 and significant at the 10% level). Moreover, ESOPs that solve stock sources through repurchase of shares can only reduce financial leverage (the coefficient on stock3 in the capital structure choice model is significantly negative, but the coefficient on Lev × stock3 in the capital structure adjustment model is not significant), whereas ESOPs that solve stock sources through the purchase of shares on the secondary market can only accelerate the dynamic adjustment of capital structure (the coefficient on Lev × stock1 in the capital structure adjustment model is significantly negative, but the coefficient on stock1 in the capital structure choice model is not significant).

In terms of the influence of management mode, in the capital structure choice model, the coefficient on mode1 is not significant, while the coefficient on mode2 (−0.026) is significantly negative. In the capital structure adjustment model, the coefficient on Lev × mode1 (−0.029) is significantly negative, but the coefficient on Lev × mode2 is not significant. This indicates that ESOPs with self-management mode helps to reduce financial leverage, while ESOPs with entrusted management mode helps to accelerate the speed of capital structure adjustment.

Table 19. The impact of the lock-up period, shareholding size and executive subscription ratio of ESOP on capital structure decisions.

Variable	(1) Lock-Up Period		(2) Shareholding Size		(3) Executive Subscription Ratio
	$\mathbf{Model} \left(6\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(7\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(6\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(7\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(6\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$	$\mathbf{Model} \left(7\right)$ ${\mathbf{L}\mathbf{e}\mathbf{v}}_{\mathbf{i},\mathbf{t}+1}$
Lev		0.451 ***		0.452 ***		0.451 ***
		(61.14)		(62.20)		(61.83)
lockup1	−0.002	0.010
	(−0.70)	(1.61)
lockup 2	−0.007	0.001
	(−0.45)	(0.03)
lockup 3	−0.017 ***	0.029 **
	(−2.90)	(2.23)
Lev × lockup1		−0.020
		(−1.48)
Lev × lockup2		−0.007
		(−0.11)
Lev × lockup3		−0.060 **
		(−2.26)
shares1			−0.013 ***	0.017 **
			(−3.60)	(2.49)
shares2			0.026 ***	0.016
			(3.11)	(0.87)
Lev × shares1				−0.038 ***
				(−2.58)
Lev × shares2				−0.021
				(−0.63)
purchase1					−0.012 ***	0.029 ***
					(−2.83)	(3.37)
purchase2					−0.004	0.000
					(−0.71)	(0.03)
Lev × purchase1						−0.060 ***
						(−3.28)
Lev × purchase2						0.010
						(0.45)
Year FE	yes	yes	yes	yes	yes	yes
Firm FE	no	yes	no	yes	no	yes
N	20,159	20,159	20,289	20,289	20,133	20,133
F-value	809.53 ***	314.06 ***	860.10 ***	342.10 ***	842.43 ***	339.40 ***
Adj-R2	0.397	0.313	0.397	0.312	0.396	0.313

Note. Due to space limitations, the regression results of constant terms and control variables are ignored in the table. The symbols *** and ** indicate significance at the 1% and 5% levels, respectively. The numbers inside the brackets are t-values, and the numbers outside the brackets are coefficients.

presents the regression results pertaining to the influence of lock-up period, shareholding size and executive subscription ratio of ESOP on capital structure decisions. Regarding the impact of lock-up period, ESOPs with a longer lock-up period (i.e., exceeding two years and up to three years) can not only reduce financial leverage (the coefficient on lockup3 in the capital structure choice model is −0.017 and significant at the 1% level), but also accelerate the speed of capital structure adjustment (the coefficient on Lev × lockup3 in the capital structure adjustment model is −0.060 and significant at the 5% level).

In terms of the impact of shareholding size, ESOPs with smaller shareholding size (less than 3%) can not only reduce financial leverage (the coefficient on shares1 in the capital structure choice model is −0.013 and significant at the 1% level), but also speed up the dynamic adjustment of capital structure (the coefficient on Lev × shares1 in the capital structure adjustment model is −0.038 and significant at the 1% level).

In terms of the impact of executive subscription ratio, ESOPs with a lower executive subscription ratio (less than 30%) can not only reduce financial leverage (the coefficient on purchase1 in the capital structure choice model is −0.012 and significant at the 1% level), but also accelerate the dynamic adjustment of capital structure (the coefficient on Lev × purchase1 in the capital structure adjustment model is −0.060 and significant at the 1% level).

In sum, the results presented in Table 18 and Table 19 demonstrate that the impact of ESOP on capital structure decisions is contingent upon the specific design of ESOP contract elements, including the source of capital, the source of stock, the management mode, the lock-up period, the size of shareholding, and the executive subscription ratio.

6. Conclusions and Discussion

This paper employs a sample of Chinese A-share listed companies on the Shanghai and Shenzhen Stock Exchanges from 2014 to 2021 to investigate the influence of ESOPs on capital structure decisions. We consider both static capital structure choice and dynamic capital structure adjustment. The results demonstrate that the implementation of ESOPs reduces the level of corporate debt and facilitate the dynamic adjustment of capital structure. This suggests that the implementation of ESOP can assist in optimizing capital structure decisions made by enterprises. Further analysis indicates that the influence of ESOP on the dynamic adjustment of capital structure is asymmetrical. In particular, the implementation of ESOP can significantly accelerate the downward adjustment of capital structure, but has no significant effect on the upward adjustment of capital structure. This could be attributed to risk aversion on the part of management and employees.

Moreover, we find that the impact of ESOPs on capital structure decisions depends on the macroeconomic environment, industry characteristics, corporate governance, and ESOP contract design. First, the optimization effect of ESOP on capital structure decisions is more pronounced in a favorable macroeconomic environment, in poor market conditions or in competitive industries. Second, the reduction effect of ESOP on corporate debt is more pronounced in high-tech enterprises, non-state-owned enterprises, and enterprises with dispersed equity; while the promotion effect of ESOP on dynamic capital structure adjustment is more pronounced in non-high-tech enterprises, state-owned enterprises, and enterprises with concentrated equity. Third, in terms of capital structure choice, ESOPs financed by loans from the firm’s major shareholders—solving their stock sources through share repurchase, or with self-management mode, a longer lock-up period, smaller shareholding scale, and executive subscription ratio—have a stronger reduction effect on financial leverage. In terms of capital structure adjustment, ESOPs financed by loans from the firm’s major shareholders—solving their stock sources by subscribing to non-publicly offered shares, or with entrusted management mode, a longer lock-up period, smaller shareholding size and executive subscription ratio—have a stronger promoting effect on dynamic capital structure adjustment.

The findings of this study have important implications for companies, employees, investors, and regulators. From the perspective of companies, it is recommended that enterprises can optimize their capital structure decisions by actively promoting the implementation of ESOP and reasonably designing the contractual elements of ESOP. For example, according to the empirical findings on the moderating role of ESOP contract design, companies should select lock-up periods of more than two years and up to three years, design the shareholding size to be less than 3%, and select the executive subscription ratio of less than 30% to realize the optimization effect of ESOP on capital structure decisions.

From the perspective of employees, it is suggested that they should be aware that participation in an ESOP may expose them to the risk of fluctuations in the company’s share price. When the company’s share price falls during the ESOP lock-up period, employees should focus on the long-term development of the company rather than short-term share price fluctuations, and work hard to improve the company’s performance to increase the market value of the company. The company may also appropriately extend the duration of the ESOP after the expiration of the lock-up period to provide more options for the sale of the underlying shares and to reduce employees’ losses.

From the perspective of investors, the implementation of ESOPs by listed company and the design of ESOP contracts can be considered as important reference information when making investment decisions, as they can improve the accuracy of such decisions. From the perspective of regulators, it would be beneficial to encourage private enterprises, high-tech enterprises and enterprises in competitive industries to proactively implement ESOPs, with the aim of reducing their financial leverage. Furthermore, the implementation of ESOP in state-owned enterprises can be encouraged to provide a theoretical basis for the formulation of relevant policies of ESOP, enhance the flexibility and enthusiasm of the implementation of ESOP in state-owned enterprises, and facilitate the supporting role of ESOP for the mixed ownership reform of state-owned enterprises. In addition, this paper can serve as a reference for the practice of employee stock ownership in other emerging markets.

It should be noted that this paper is not without limitations. First, in examining the moderating effect of the internal governance environment on the relationship between ESOP and capital structure decisions, this paper considers only two internal governance variables: the nature of corporate property rights and the concentration of equity. These variables represent the characteristics of China’s institutional background. Further research could examine the moderating effects of additional internal and external governance mechanisms, such as board supervision, management incentives and analyst following, on the relationship between ESOP and capital structure decisions. Second, in examining the impact of ESOP contract design on capital structure decisions, this paper focuses on the basic contract elements of ESOP, such as capital source, stock source, and shareholding size, due to limitations of the available data. In the future, further examination of the impact of innovative contract elements of ESOP, such as performance assessment, phased unlocking and reserved shares setting, on capital structure decisions can be conducted based on the manual collection of relevant data. Third, there are discrepancies between the ESOPs observed in China and those in other countries, which can be attributed to varying local policies and regulations. However, due to the challenges associated with data acquisition, a comparison of the impact of these discrepancies on ESOP has not yet been conducted. This represents a promising avenue for future research.