The Impact of Marketization on Enterprise Performance from the Perspective of Enterprise Debt

Abstract

The problem of corporate debt is a hidden danger to China’s economic growth and financial stability. This paper studies the effect of regional marketization on the debt burden of the local enterprises based on a fixed effects model. The result demonstrates that the higher the market-oriented degree, the lower the debt burden of the enterprises in that region. Further analysis demonstrates that the improvement of marketization not only directly reduces the enterprise debt, but also reduces debt burden by increasing the enterprise productivity and promoting the enterprise property rights reform. Improving the level of regional marketization is conducive to the enterprise debt issue, and this provides some suggestions to the policy makers.

1. Introduction

On 2 April 2018, the Central Financial and Economic Commission held its first meeting, proposing the task of “structural deleveraging” for the first time, which raised two issues worth considering. First, why does leverage need to come down as soon as possible? Statistics demonstrate that after the global financial crisis, the trend of leverage in Chinese (non-financial) enterprises is very obvious. According to the Bank for International Settlements (BIS), as of the end of 2018, the leverage ratio of China’s non-financial enterprises was 153.6%, up 56.6% from the end of 2007, far exceeding 70% in the US, 101.8% in Japan and 104% in the euro zone, 70 percentage points higher than the average of developed countries and 53 percentage points higher than the average of emerging economies [1]. The growing problem of corporate leverage cannot be ignored, and has become a hidden danger for the healthy and stable development of finance. Second, why emphasize deleveraging through structural adjustment? The study found that although the over-all corporate asset–liability ratio increased, the changes in the asset–liability ratio of enterprises of different ownership types were inconsistent. The average asset–liability ratio of private enterprises has demonstrated a downward trend for most of the time after the financial crisis, while state-owned enterprises have entered an upward channel since 2008, showing obvious leverage and strong deleveraging ‘rigidity’. Based on the above phenomenon, [2] proposed that since the outbreak of the financial crisis, the debt ratio of Chinese enterprises has been rising at the overall level and demonstrating structural characteristics. That is to say, corporate debt related to ownership has become the main structural factor driving up the debt of the whole society. This kind of structural enterprise debt will affect the sustainable development of the economy from three aspects: one is the low efficiency and resource waste of state-owned enterprises, which affects the supply efficiency of public goods and services; two, the financing constraints of private enterprises inhibit the innovation capability of enterprises; three, the enterprise’s debt mismatch increases the overall economic financing costs and increases the burden of corporate debt.

A market-oriented reform is conducive to improving corporate performance. Corporate performance includes productivity (TFP) and asset–liability ratio (LEV). Good corporate performance should match the level of productivity with the level of debt ratio, which can maintain the sustainable and healthy development of enterprises. Productivity progress is sustainable enough, and enterprises can rely on self-financing accumulation to overcome the capital mismatch caused by debt barrier, thereby improving enterprise performance; at the same time, the reform of property rights will affect the governance structure and incentive mechanism of enterprises, and then affect the behavior characteristics and performance of enterprises. Promoting a market-oriented reform will strengthen the property rights reform of state-owned enterprises. The property rights reform has a significant positive effect on the total factor productivity of state-owned enterprises, which is conducive to reducing the asset–liability ratio of enterprises [3,4].

At present, there is much literature on enterprise debt, most of which study corporate deleveraging from the perspective of financial deepening [5,6,7,8,9]; there is little research on corporate deleveraging from the perspective of market transformation and property rights [10]. A noteworthy phenomenon is that before the global financial crisis, China’s corporate debt ratio is declining; after the global financial crisis, corporate debt ratio increased significantly. At the same time, another phenomenon is also worthy of attention. Unlike past market-oriented economic reforms, after the global financial crisis, the number of ‘The State Advances, the Private Sector Retreats’ cases of state-owned enterprises acquiring or annexing private enterprises has increased significantly. Research also demonstrates the existence of ‘The State Advances, the Private Sector Retreats’ phenomenon [11,12]. Therefore, is the resistance to a market-oriented transformation and property rights reform related to the rise of corporate debt? If so, what is its mechanism?

The existing literature mainly discusses the issue of the market-oriented transformation and resource allocation efficiency, and believes that the market-oriented transformation is conducive to improving resource allocation efficiency [13,14,15]. With the improvement of resource allocation efficiency, capital can be transferred from inefficient areas to efficient areas, thereby increasing TFP [16,17]. Market-oriented transformation also promotes property rights reform, and the participation of non-state-owned shareholders in the high-level governance of state-owned enterprises can alleviate government intervention and thus increase corporate productivity [10]. There is not much literature on the impact of market-oriented transformation on corporate debt, and there is little literature on the impact of productivity on corporate debt [18,19]. The existing literature is more empirical in studying the relationship between policy and corporate deleveraging, and although there are discussions on the mechanisms of deleveraging, they are mostly indirect mechanisms, with little discussion on the intrinsic direct mechanisms; they do not reflect the role of the intrinsic direct mechanisms of corporate productivity on deleveraging [3,4,5,6,7,8,9]. This paper integrates the market-oriented external environment and internal corporate governance into a unified framework, and discusses not only the indirect mechanisms of the market-oriented transformation and property rights reform on corporate deleveraging, but also the intrinsic direct mechanisms of deleveraging through improving corporate productivity. This paper establishes a research framework, taking market-oriented transformation and property rights reform as exogenous institutional shocks, to explain the way in which China improves corporate performance by means of institutional improvement. The key step is to internalize exogenous institutions (market-oriented transformation and property rights reform) into corporate production efficiency (productivity), thereby improving corporate debt ratio. Using the 1998–2007 industrial enterprise database of the National Bureau of Statistics and the provincial marketization process index calculated by [20], this paper empirically examines whether market-oriented transformation and property rights reform are conducive to corporate deleveraging.

The empirical study of this paper finds that the improvement of the marketization level is indeed conducive to reducing the debt ratio of enterprises. Secondly, the degree of marketization not only directly has an impact on corporate debt ratio, but also by affecting corporate productivity and corporate ownership structure, thereby affecting the debt ratio. That is to say, the degree of marketization to improve the corporate debt problem has multiple ways. The first is the direct effect, that is, a market-oriented institutional reform and business environment improvement to improve corporate debt problems. This includes the continuous improvement of the legal system, regulatory means, and the increasingly perfect securities market. The second is the self-financing mechanism, that is, the degree of marketization by raising corporate TFP, enhancing corporate self-financing capabilities, and thereby improving corporate debt problems. This finding provides empirical support for the financial mismatch theory proposed by Moll [21]. Previous empirical studies often ignored the role of the corporate self-financing capacity [21]. Third, there is the debt structure adjustment, that is, market-oriented enterprises to promote the adjustment of property rights, thereby improving the corporate debt structure. This is mainly to combine the financing advantages of the public ownership economy and the productivity advantages of the non-public economy to restructure the debt. For example, transforming state-owned enterprises into private enterprises through a mixed ownership reform or public ownership economy buys a share in private enterprises. In fact, this approach is also a solution to the problem of debt mismatch between private enterprises and state-owned enterprises [2].

The contribution of this paper is mainly reflected in the following aspects. First of all, it provides empirical support for marketization to improve corporate debt, and emphasizes the impact of corporate total factor productivity on debt ratio. Few studies have discussed this before, and this study makes up for this deficiency. Secondly, it clarifies the internal mechanism of market-oriented transformation and property rights reform affecting corporate debt ratio, which is conducive to deepening our understanding of the coordinated promotion of market-oriented economic transformation and property rights reform. Finally, this paper provides evidence from Chinese industrial enterprises for the theory of a financial mismatch, and provides a solution to the problem of debt mismatch between private enterprises and state-owned enterprises [2]. It has important public policy value and provides a theoretical basis for ‘structural deleveraging’.

This paper is organized as follows: the second part is the literature review and theoretical framework of deleveraging; the third part puts forward a simple theory; the fourth part is the empirical test of the theory; finally, the summary and policy recommendations are presented.

2. Literature Review

2.1. Research on Deleveraging

Ref. [22] analyzed the relationship between market-oriented reform and leverage ratio from the macro level. Ref. [23] research demonstrates that mixed ownership reform is conducive to the improvement of corporate performance and the reduction of the leverage of state-owned enterprises. There are also many studies discussing deleveraging from the micro-enterprise level; see

Table 1. Overview of deleveraging literature.

Literature	Deleveraging Mechanism
Lu Dong, Jiang Muzi. (2021) [5]	Bank competition speeds up capital structure adjustment by raising financing costs for zombie firms.
Li Huamin, Ren Ding, Wu Fei, etc. (2020) [6]	Interest rate liberalization improves the flow efficiency of financial resources. The flow of monetary resources is guided by the level of marginal output, thereby reducing the overall corporate leverage.
Tan Xiaofen, Li Yuan, Wang Kexin. (2019) [7]	The enhancement of the marketization of financial structure will reduce the leverage ratio of enterprises by promoting the increase in the scale of equity financing.
Lin Aijie, Liang Qi, Fu Guohua. (2021) [8]	Failure to examine the impact of digital financial development on corporate leverage transmission mechanism.
Xie Zhouliang, Zhou Suhua. (2021) [9]	The mitigation of financing constraints and the improvement of financing channels have a certain degree of optimization of corporate leverage.
Ma Xinxiao, Dou Xiaochen. (2022) [10]	The participation of non-state-owned shareholders in the high-level governance of state-owned enterprises can alleviate government intervention, improve executive incentives and strengthen internal control, thus exerting a governance effect on the leverage manipulation of state-owned enterprises.
Jiang Lingduo, Lu Yi. (2018) [24]	The bank implements soft budget constraints on state-owned enterprises to protect state-owned inefficient enterprises.
Zhang Binbin, He Dexu, Zhang Xiaoyan. (2020) [25]	FinTech can reduce the leverage ratio by reducing financing constraints and financial costs.
Wu Jinghua, Qiu Quanfeng, Wang Hongjian. (2019) [26]	Interest rate control policy and income tax rate affect the financial policy of enterprises, which in turn affects the capital structure of enterprises.
Shen Guangjun, Zhang Yan, Wang Rong. (2018) [27]	The VAT transformation reduces the tax expenditure of enterprises, and requires more long-term liabilities to support.
Dai Qinwen, Xu Weihang. (2019) [28]	The change of market accessibility of enterprises has an impact on their leverage ratio through government intervention in the market, financial marketization, and distance from central cities.
Yu Bo. (2017) [29]	Technological innovation regulates corporate leverage by strengthening corporate market competitiveness and adjusting cost hedging capabilities.
Li Xiaoxi, Rao Pingui. (2022) [30]	The rule of law promotes the adjustment of enterprises, so as to achieve the optimal capital structure.
This paper	Marketization not only directly improves the allocation of resources, but also enhances the total factor productivity of enterprises and promotes the reform of property rights, making the self-financing mechanism and the debt structure adjustment mechanism work, thereby reducing leverage.

.

From Table 1, we can observe that most of the literature discusses the impact of financial deepening on corporate deleveraging. There are also studies on the impact of corporate deleveraging from the perspectives of taxation, innovation, the legal system, and market accessibility. Another part of the literature discusses the relationship between corporate leverage and corporate performance [31,32,33]. The existing literature is more about the indirect mechanism of deleveraging, which does not reflect the effect of enterprise productivity on deleveraging, does not discuss the impact of property rights reform on corporate debt under different marketization conditions, and does not integrate the external environment of marketization and the internal governance of enterprises (property rights structure) into a unified framework to discuss the impact mechanism of marketization transformation and property rights reform on corporate deleveraging.

2.2. Theoretical Framework

First, Fan et al. (2003) constructed a marketization index to measure the process of marketization transformation [34]. The marketization index quantifies the institutional variables, making it possible to quantitatively examine the economic significance of institutional factors. The existing literature mainly studies the impact of the market-oriented transformation on the economy from the macro and micro levels. At the macro level, the main research focuses on the impact on TFP and believes that the market-oriented transformation is conducive to improving the efficiency of resource allocation [13,14], thereby improving productivity [15]. That is to say, in the case of a higher degree of marketization, capital can move faster from inefficient areas to efficient areas, thereby increasing TFP. At the micro level, it mainly studies the impact on enterprise productivity and debt ratio. The study found that the improvement of marketization level promoted the improvement of enterprise productivity [16], but the impact of marketization on the corporate debt ratio was not agreed. For example, the empirical research of [2] shows that the debt ratio determination method of private enterprises is more consistent with that of western enterprises, or is more market-oriented; there are more ‘non-market’ elements in the determinants of the debt ratio of state-owned enterprises. Ref. [17] found that marketization has no significant effect on corporate debt ratio. The former does not directly define and index the marketization. Although the latter introduces the marketization index, the conclusion may be distorted due to the serious lack of data. The authors introduce a complete marketization index and further discuss the influence mechanism of marketization on the corporate debt ratio. The empirical results demonstrate that marketization has a significant negative impact on corporate debt ratio, and the results are robust.

Secondly, the empirical analysis based on the theory of property rights economics found that public property rights will have an impact on corporate financing costs [18]. Further study found that the nature of enterprises and equity will affect corporate credit financing. This is mainly manifested as: 1. The supplier of funds does not allocate funds entirely based on the principle of profit, but also depends on the nature of the enterprise. Compared to private enterprises, state-owned enterprises have an increasing advantage in obtaining loans [2]; 2. State-owned equity can bring financing advantages, and state-owned equity of private enterprises can rely on its relationship with the government to provide more economic resources for the development of private enterprises [35]. Although public property rights can bring credit financing advantages, the efficiency is poor. In terms of bank loans and government funding as a proportion of the total investment, state-owned enterprises have more than three times the proportion of private enterprises, but in terms of capital gains, state-owned enterprises are only half of private enterprises [36]. The above research explains the important reasons for the formation of corporate structural debt, but further research is needed on how to improve corporate structural debt. In theory, property rights and the information economics theory of cooperative teams believe that mixed ownership reform can take advantage of the respective advantages of different property rights capital, improve the efficiency of resource allocation [37,38], and then reduce the debt ratio of enterprises under the self-financing mechanism [21]. Empirically, few studies have researched this. This paper uses the total factor productivity (TFP) of enterprises to measure the self-financing ability, which provides a new perspective for understanding the internal mechanism of the self-financing ability affecting the corporate debt ratio.

Finally, property rights economics has always emphasized the decisive influence of the institution on contract structure. For example, the impact of the judicial system on the corporate ownership structure and financial markets has been confirmed by a large number of cross-country comparative literature [39]. In China, there are also some studies that verify that the marketization process and the nature of the enterprise ownership have an important impact on the debt maturity structure of enterprises [19]. Methodologically, this type of literature uses the marketization process index and corporate attributes (state-owned or private) as proxy variables for institutional and corporate ownership structures. Since the interaction between the two is not considered, these studies cannot well reflect how the differences in the marketization process affect the change of the corporate ownership structure, which in turn affects corporate debt decisions. The authors will expand the enterprise attributes, including not only the holding situation (state-owned or private), but also the proportion of state-owned enterprises. On the basis of the original literature, the interaction between the marketization process and the proportion of state-owned enterprises is further considered, which can better examine the impact of the marketization process on the ownership structure of enterprises, and then affect the mechanism of the corporate debt ratio.

Based on the above literature review, the research framework proposed in this paper follows the ‘market-behavior-performance’ organization theory; that is, the firm performance is mainly affected by the external market environment and internal governance structure. Generally speaking, in addition to the direct effect of the external environment, the internal stress behavior of the enterprise also plays a role. The key to the internal and external co-action is the effective allocation of resources. Figure 1 summarizes our theoretical framework, including three levels: First, the authors adopt the basic framework of the impact of market-oriented transformation (measuring the external environment of enterprises) and property rights reform (measuring the internal structure) on corporate performance. Second, the key to the influence of the internal structure and external environment is to improve the efficiency of resource allocation. From the perspective of China’s situation, it is mainly to reduce the resource mismatch caused by the financial market friction. Third, the fundamental way to improve corporate debt is the matching of the productivity and debt ratio.

2.3. Theoretical Analysis of Market-Oriented Reform Deleveraging

China’s marketization index consists of five aspects: the relationship between the government and market; the development of the non-state-owned economy; the development of the product market; the development of the factor market; the development of market intermediary organizations and the legal environment of market. By combing the literature, it can be concluded that the enhancement of the marketization degree of the market-oriented financial structure will promote the increase in the scale of corporate equity financing, which will eventually lead to the decrease in corporate leverage. The interest rate marketization improves the flow efficiency of financial resources, making the flow of monetary resources oriented to the level of marginal output, thus reducing the leverage ratio of enterprises. Legalization promotes the adjustment of the capital structure of enterprises, so as to achieve the optimal capital structure that can maximize the value of enterprises, and then achieve the goal of deleveraging. The change of the market accessibility of enterprises affects their leverage ratio through government intervention in the market and the degree of financial marketization, thus reducing leverage. The bank competition reduces leverage by raising the financing costs of zombie firms. The participation of non-state-owned shareholders in the high-level governance of state-owned enterprises can alleviate government intervention, improve executive incentives, and strengthen internal control, thus playing a governance role in the leverage manipulation of state-owned enterprises and reducing corporate leverage. Digital finance can alleviate the financing constraints of private enterprises and improve the financing channels, thus optimizing the leverage ratio of enterprises to some extent. At the same time, it is found that market-oriented transformation is conducive to improving the efficiency of resource allocation, thereby improving corporate productivity (TFP) and reducing the corporate debt ratio under the self-financing mechanism. However, there is no literature to study this mechanism. Based on this, this paper proposes the following research hypotheses:

H1. 

Market-oriented reform reduces corporate leverage through resource allocation improvement.

H2. 

Market-oriented reform improves corporate productivity by optimizing resource allocation efficiency, and then corporate leverage.

3. Theoretical Model of Productivity and Debt

According to the above framework, this part establishes a simple theoretical model, and obtains the measurement equation based on this theoretical model. Firstly, according to the explanation of [40], the authors assume that the production function is a Hicks-neutral Cobb–Douglas production function:

$$y_{it}=f\left(A_{it},k_{it},l_{it}\right)=A_{it}{k}_{it}{}^a{l}_{it}{}^{\beta }$$

(1)

where the subscript $i$ represents the enterprise, $t$ represents the year, $k_{it}$ is the capital investment, $l_{it}$ is the labor investment, and $A_{it}$ is the total factor productivity (TFP). It can be observed from the production function that an important power source for the sustainable development of enterprises is the improvement of TFP in the case of limited resources. The market environment outside the enterprise and the property right and governance inside the enterprise are the main factors affecting TFP; thus, the TFP form is set as follows:

$$A_{it}=A{e}^{\delta Z_{it}+\phi X_{it}}$$

(2)

Among them, $Z_{it}$ represents the influence mechanism of the external market environment and internal governance on TFP. The external environment includes the marketization process and the degree of product competition. The internal governance is mainly the ownership structure. The ownership structure includes the proportion of state-owned enterprises, the shareholding situation, and the mixed index. In addition, in (2), $X_{it}$ is other factors that affect TFP, including the enterprise survival age, enterprise size, new product ratio, export scale, subsidy income, total debt, total asset turnover, and return on assets. The econometric model of TFP can be established by (3):

$$\log A_{it}=\log A+\delta Z_{it}+\phi X_{it}+{\mu }_i+{\epsilon }_{it}.$$

(3)

${\mu }_i$ is the inherent characteristics of the enterprise $i$, and ${\epsilon }_{it}$ is the random disturbance term.

Secondly, when there is friction in the financial market, the Moll (2014) [21] model assumes that enterprises obtain favorable investment opportunities, but the debt capacity is limited by $k_{it}\le {\lambda }_{it}{a}_{it} \mathrm{and} {\lambda }_{it}\ge 1$; thus, enterprises have to rely on their own capital accumulation slowly. According to this logic, if enterprises miss the best investment opportunities, productivity improvement will be hindered, resulting in the accumulation of debt. The authors define ${\eta }_{it}$ as the firm’s financing capacity, and the firm’s debt accumulation is constrained by:

$$k_{it}\le {\lambda }_{it}{a}_{it}\Rightarrow \frac{k_{it}-a_{it}}{k_{it}}\le 1-\frac{1}{{\lambda }_{it}}\stackrel{\Delta }{=}{\eta }_{it}$$

(4)

Under the condition of market competition, the constraint conditions in Formula (4) are tight; that is to say, in order to invest and develop, enterprises hope to obtain external financing, and the total asset–liability ratio reflects the external financing ability of enterprises [41]. In the Moll (2014) [21] model, TFP is exogenously given, and the corporate debt capacity can be any function of TFP. The external market environment, internal property rights and governance also affect the total asset–liability ratio. Therefore, the total asset–liability ratio function is as follows:

$${\eta }_{it}=f(A_{it},Z_{it},{X^{\prime }}_{it})$$

(5)

Among them, $Z_{it}$ represents the influence mechanism of the external market environment and internal governance on financing ability. The external environment includes the marketization process and the degree of product competition. The internal governance is mainly the ownership structure. The ownership structure includes the proportion of state-owned enterprises, the shareholding situation, and the mixed index. In addition, ${X^{\prime }}_{it}$ is other influencing factors of enterprise debt capacity, including enterprise survival age, enterprise size, subsidy income, total liabilities, total asset turnover, return on assets, and liquidation ratio. Based on the above discussion and the heterogeneity of enterprises, the design quantity model is as follows:

$${\eta }_{it}=\eta +\varphi \log A_{it}+\theta Z_{it}+\gamma {X^{\prime }}_{it}+{{\mu }^{\prime }}_i+{{\epsilon }^{\prime }}_{it}$$

(6)

where ${{\mu }^{\prime }}_i$ is the inherent characteristic of the $i$ enterprise, and ${{\epsilon }^{\prime }}_{it}$ is the random disturbance term.

In fact, due to the correlation between internal and external factors, the model measures both the external environment of the enterprise (including market-oriented transformation and competition degree) and the internal governance of the enterprise (including the state-owned proportion, holding situation, and mixed index), which can effectively avoid the estimation bias caused by endogenous problems. However, the impact of TFP on the asset–liability ratio in Equation (6) is often questioned, such as whether there is an endogenous problem due to the characteristics of corporate financial capabilities. The fixed effect model can control the factors that do not change over time, but if there are factors that change over time, then this concern is justified. Therefore, the authors consider the instrumental variable method. Equation (3) can be understood as the estimation of the first stage of the endogenous variable TFP, and Equation (6) is the estimation of the second stage. The instrumental variables the authors consider include the new product ratio and export size. In China, the government often encourages enterprises to innovate and export, so these two variables are related to external policies and have little to do with the financial ability of enterprises. However, many studies have found that these two variables can often affect the TFP of enterprises.

4. Empirical Testing

This paper combines the micro data of industrial enterprises of the National Bureau of Statistics (1998–2007) and the provincial marketization index of China calculated by Fan Gang et al. The following is a detailed description.

4.1. Database Description

Chinese Industrial Enterprise Database conducts statistical surveys on all state-owned enterprises and non-state-owned enterprises with annual sales revenues of more than 5 million yuan. According to previous studies, the values of some variables in this database may be inconsistent with the facts due to misstatement or underreporting by enterprises. According to the following principles, this paper deletes some enterprises with doubtful data: 1. The key financial indicators cannot be missing or negative. These financial indicators include total assets, owner’s equity, paid-in capital, total liabilities, net fixed asset balance, sales revenue, and total industrial output value; 2. In accordance with the recommendations of [15], the number of employees of enterprises cannot be less than eight people; 3. Enterprises whose sales revenue is less than 5 million yuan and corporate capital are not classified. In addition, like [42], this paper deletes enterprises that do not meet the following principles according to general accepted accounting principles: total assets must be greater than the net balance of fixed assets; total assets shall not be less than current assets or total fixed assets; the paid-in capital should be the sum of the capital of all kinds of investors. On the basis of deleting some data, the total factor productivity, asset–liability ratio, Herfindahl–Hirschman index, new product ratio, export product ratio, subsidy income, enterprise scale, enterprise survival age, logarithm of total liabilities, total asset turnover rate, asset return rate, and liquidation ratio were reduced by 0.5% at both ends. The non-public economy and public economy are defined from the perspective of paid-in capital. Individual capital, capital from Hong Kong, Macao and Taiwan, foreign capital, and part of corporate capital are regarded as a non-public economy, while a state-owned capital, collective capital, and part of a corporate capital are regarded as a public economy.

4.2. Marketization, Property Rights, and Core Variables

Market transformation and property rights reform are combined. In the industrial organization theory, the Herfindahl–Hirschman index is often used to describe the degree of competition in the industry in which the enterprise is located. It reflects the living environment of the industry and will have an important impact on the production decision of the enterprise. However, the marketization index has a broader meaning, reflecting the environment of the enterprise survival. The marketization index measures the relative process of marketization transformation and quantifies the institutional variables, which is one of the focuses of this paper. Figure 2 shows that from the time dimension, the overall degree of marketization is rising. From the regional perspective, the developed coastal areas not only have a high degree of marketization, but also the process is fast, central inland areas followed, and the western region is slower. The interaction term between the proportion of state-owned enterprises and the marketization index reflects the impact of marketization through the change of the proportion of state-owned enterprises, and reflects the indirect impact of marketization on enterprises. The enterprises are divided into state-owned enterprises (the proportion of state-owned enterprises is greater than 50%) and private enterprises (the proportion of state-owned enterprises is less than 50%) according to the holding situation, which reflects the actual control and dominance of enterprise assets. The proportion of state-owned enterprises considers the change of ownership structure from a quantitative perspective, reflecting the impact of marketization on enterprises. Finally, the mixed index is the number of different types of capital participation, reflecting the breadth of participation.

Core variables include TFP and the asset–liability ratio. TFP is an indicator of productivity. It is the ‘residual value’ after removing the input of labor, capital, and other factors, which can truly reflect the efficiency essence of individual enterprises. Ref. [43] defined TFP as marketization and other factors that affect innovation efficiency, including both the impact of marketization on TFP through other channels and other control variables (mainly enterprise size and economic performance). The asset–liability ratio is a comprehensive indicator to measure the level of corporate debt, reflecting the proportion of corporate borrowings in total assets, reflecting the borrowing capacity of different enterprises, corresponding to the leverage ratio, and essentially reflecting the ratio of corporate equity capital to debt. Ref. [44] believe that factors such as technology, innovation, or institutional progress are the driving force for corporate debt ratios. From the micro perspective of the MM theorem, it is concluded that information and transaction costs, the tax burden, the soft budget constraint, and the incentive mechanism are closely related to the debt ratio of enterprises. Specifically, the enterprise scale, tax shield effect (debt), soft budget constraint [45], and rigid payment (all attributes of enterprises) will also affect the debt ratio of enterprises.

There are differences in TFP among different types of enterprises, such as the statistical description of TFP of various types of enterprises (see

Table 2. Descriptive statistics of TFP of various type of enterprises.

Category	Sample	Mean	Variance	Median	Minimum	Maximum
State-owned	84,577	6.96	1.29	6.89	1.81	11.71
Mixed	35,845	7.25	1.24	7.16	1.83	11.68
Private	382,690	7.12	1.10	7.05	1.80	11.71

) and the probability density of TFP distribution (see Appendix A Figure A1). From the data statistics and graphics, it can be found that the mean and median of TFP of private enterprises are greater than that of state-owned enterprises, with the highest TFP for mixed enterprises. From the overall economy, the mixed ownership property rights reform is conducive to improving the total factor productivity of state-owned enterprises.

Different types of enterprises not only have differences in production efficiency, but also face different financing constraints. From

Table 3. Asset–liability ratio descriptive statistics of various type enterprises.

Category	Sample	Mean	Variance	Median	Minimum	Maximum
State-owned	84,577	0.610	0.230	0.640	0	1
Mixed	35,845	0.580	0.220	0.600	0	1
Private	382,690	0.560	0.240	0.590	0	1

, we can observe that whether it is the mean or median, the asset–liability ratio of state-owned enterprises is higher than that of private enterprises, which demonstrates that state-owned enterprises have advantages in borrowing. In general, private enterprises are more difficult to finance than state-owned enterprises, but the reform of mixed ownership property rights helps to alleviate the financing difficulties of private enterprises (see Appendix A Figure A2).

From the perspective of dynamic changes, the average asset–liability ratio of private enterprises has demonstrated a downward trend for most of the time after the financial crisis, while state-owned enterprises have entered an upward trend since 2008 (Figure 3).

4.3. Other Variables and Statistical Description

The selection of other variables in the econometric model is based on the previous literature. The definition of variables and statistical indicators are reported in

Table 4. Definition and description of variables.

Classification	Variable Name	Code	Explanation
Kernel variable	Total factor productivity	tfp	The part of output exceeding the growth rate of factor input
	Asset–liability ratio	lev	Total liabilities/total assets
Market circumstances	Marketization index	market	Fan Gang’s China Marketization Index, reflecting the relative process of marketization in each region
	Herfindahl–Hirschman index (control variable)	hhi	Sum of squares of percentage of total assets, reflecting the degree of market competition (calculated by industry four-digit code)
Ownership structure	Proportion of state-owned	ratio	Public economy/paid-in capital
	Holdings (sample distinguishing variables)	contr	State-owned enterprises (ratio > 0.5)
	Mixed index (control variable)	mix	Or, private enterprise (ratio < 0.5)
Key indicators	Proportion of state-owned × Market index	rm	The interaction between state-owned proportion and marketization index
Other control variable indicators	New Product Ratio	new	New product output/total output
	Export product ratio	ex	Export delivery value/sales revenue/10
	Subsidy income	sub	Log (1 + subsidy income)
	Enterprise size	size	Log (number of employees)
	Enterprise survival age	age	(Registration year − registration year)/100
	The logarithm of total liabilities	liability	Log (total liabilities)
	Total asset turnover rate	manage	Net operating income/average total assets
	Return on assets	ROA	Total pre-tax profit/total assets
	Liquidation ratio	Tang	Total fixed assets/total assets

and

Table 5. Variable Description Statistics and Correlations.

	Tfp	Lev	Market	Hhi	Ratio	Mix	New	Ex	Sub	Size	Age	Liability	Manage	ROA	Tng
1.tfp	1
2.lev	−0.0744 *	1
3.market	0.0641 *	0.0064 *	1
4.hhi	−0.0035	0.0144 *	−0.1416 *	1
5.ratio	−0.0453 *	0.0727 *	−0.4395 *	0.0988 *	1
6.mix	0.0879 *	−0.000900	−0.1232 *	0.0324 *	0.1717 *	1
7.new	0.1015 *	0.0072 *	0.0356 *	0.0419 *	0.0160 *	0.0475 *	1
8.ex	0.0237 *	−0.0182 *	0.1933 *	−0.0491 *	−0.1682 *	0.00530	0.0288 *	1
9.sub	0.1435 *	0.0208 *	0.0282 *	0.0222 *	0.1049 *	0.0801 *	0.0925 *	−0.00270	1
10.size	0.4620 *	0.0700 *	−0.1321 *	0.0378 *	0.1394 *	0.1352 *	0.1000 *	0.2152 *	0.1753 *	1
11.age	0.0487 *	0.0949 *	−0.2216 *	0.0614 *	0.4097 *	0.0764 *	0.0555 *	−0.0793 *	0.0877 *	0.2602 *	1
12.liability	0.4768 *	0.4306 *	0.0110 *	0.0674 *	0.1035 *	0.1390 *	0.1336 *	0.0343 *	0.2462 *	0.6293 *	0.2026 *	1
13.manage	0.1853 *	−0.1169 *	0.0471 *	−0.0331 *	−0.1058 *	−0.0587 *	−0.0460 *	−0.0197 *	−0.0943 *	−0.1364 *	−0.1332 *	−0.3148 *	1
14.ROA	0.2254 *	−0.2398 *	0.00520	−0.0145 *	−0.0573 *	−0.0141 *	−0.0148 *	−0.0628 *	−0.0309 *	−0.0809 *	−0.0848 *	−0.2150 *	0.6026 *	1
15.Tang	−0.0210 *	−0.2341 *	−0.1871 *	−0.0216 *	0.0388 *	−0.0061 *	−0.0343 *	−0.0770 *	−0.0439 *	0.0714 *	0.0082 *	−0.0797 *	0.0285 *	0.0718 *	1
Mean	7.1	0.57	8.14	0.03	0.21	1.19	0.04	0.2	0.86	4.93	0.1	9.26	2.15	0.08	0.36
SD	1.15	0.24	2.2	0.05	0.39	0.45	0.16	0.36	2.12	1.09	0.1	1.55	2.69	0.19	0.20
Minimum	1.8	0	0	0	0	1	0	0	0	2.4	0	0	0	−0.23	0.00
Maximum	11.71	1	11.71	1	1	6	18.49	1	14.22	12.02	8.07	18.02	361.4	36.5	1
Number	503,112	503,112	503,112	503,112	503,112	503,112	458,448	503,058	502,835	503,112	503,112	501,492	503,112	503,112	503,112

* represents the significance levels of 10%.

, respectively, as follows: (1) Enterprise innovation. The empirical study of industrial organization finds that innovation activities and R & D activities can significantly improve the production efficiency of enterprises [16,46]. The ratio of new products reflects the innovation ability of enterprises. The larger the ratio, the greater the innovation intensity. (2) Product export. Our theoretical framework emphasizes the role of marketization. In fact, this idea coincides with the new trade theory—due to the expansion of the market, the productivity of export enterprises in the same industry is higher than that of domestic enterprises [16]. However, some studies have found that the average productivity of China’s export enterprises is generally lower than that of domestic enterprises, and there is a ‘productivity paradox’ [47]. It can be observed that whether the domestic market or the international market is more important to enterprise productivity, the researchers did not agree. (3) Government subsidies. Some studies have pointed out that government subsidies can offset the negative impact of financing restrictions on enterprises and improve enterprise productivity [48]. However, ref. [49] found that the increase or inhibition in enterprise productivity depends on the intensity of government subsidies. This demonstrates that it is controversial whether government subsidies promote or hinder the development of enterprise productivity. (4) Enterprise size and age. In theory, the higher the production efficiency, the larger the scale. The size and age of enterprises are one of the main sources of productivity heterogeneity. However, ref. [50] points out that for developing country firms, there is not sufficient evidence that firms with larger sizes are more productive. Similarly, the relationship between firm age and firm productivity is uncertain. Ref. [51] found a positive correlation between firm size and debt ratio, and also found that the relationship between firm size and debt ratio is uncertain [2,18]. (5) Financial environment. Total liabilities as a proxy variable for external financing; ref. [52] points out that external financing promotes productivity growth. Ref. [16] believes that total liabilities hindered the improvement of TFP. Because debt acts as a tax shield, it has a positive relationship with the debt ratio [53]. (6) Enterprise management. The total asset turnover rate is an important index to comprehensively evaluate the utilization efficiency and management quality of all assets in an enterprise. The faster the total asset turnover rate, the stronger the sales and management ability, and the higher the enterprise productivity [16]. Generally speaking, the higher the efficiency of the asset operation, the stronger the profitability of enterprises, but [54] found that the relationship between the asset operation efficiency and debt ratio is uncertain. (7) Profitability. ROA reflects the profitability of enterprises. It is generally believed that it has a positive relationship with TFP and a negative relationship with the asset–liability ratio [13,18]. Another study found that its relationship with debt ratio is uncertain [54]. (8) Asset structure. Liquidation ratio as a proxy variable of asset structure; the higher the liquidation ratio of the company to recover loans, the more secure. However, there are differences in the relationship between the liquidation ratio and the debt ratio. Ref. [19] found that it is a positive relationship, while [18] believe that it is a negative relationship.

Generally speaking, when evaluating the influencing factors of the debt ratio, different studies often have great differences. For example, the actual results of the relationship between the Chinese firm size, liquidation ratio, profitability and debt ratio are often contrary to predictions based on the trade-off theory or pecking order theory [54]. This paper tries to control the influence of these factors.

4.4. TFP Regression Results Analysis

Firstly, this paper estimates the total factor productivity of micro enterprises by referring to the methods of [55,56]. There are two problems in estimating TFP by the ordinary least squares (OLS) method. One is the simultaneity bias, and the other is the sample selection bias. It is usually considered necessary to use the semi-parametric method of Olley and Pakes (OP, 1996) to estimate the production function at the two-digit industrial level, and then obtain a consistent estimator of the total factor productivity at the enterprise level [57]. However, the problem with the OP method is that the investment needs to be greater than zero. Ref. [58] solved this problem by replacing investment with intermediate investment. The modified OP method draws on the LP idea and uses the logarithmic form of actual intermediate inputs as a proxy variable for productivity shocks that are known to entrepreneurs but not observable to modelers, thereby solving the problem of simultaneity bias, which can also reduce data deletion problems due to missing investment data. The modified OP method calculates the exit probability of the enterprise to solve the problem of selective deviation.

Then, we use Equation (3) to analyze the influence mechanism of the external environment and internal governance on the total factor productivity of enterprises. Since the Hausman test results demonstrate that it is not the consistency of FE and RE methods, the author believes that the fixed effect model results should be used instead of the random effect model results. Therefore, this paper reports the results of the fixed effect model. Ref. [59] found that the governance structure and corporate performance of Chinese-listed companies are related to the ownership nature of the largest shareholder. Non-state holding companies are more flexible in operation and more efficient in corporate governance. Similarly, this paper classifies and regresses enterprises from the perspective of holding. In

Table 6. Total factor productivity (TFP) regression results.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	State-Owned Enterprises	Private Enterprises	Overall	State-Owned Enterprises	Private Enterprises	Overall	State-Owned Enterprises	Private Enterprises	Overall
market	0.086 ***	0.124 ***	0.117 ***				0.107 ***	0.125 ***	0.118 ***
	(33.73)	(88.67)	(103.46)				(7.53)	(87.01)	(92.11)
ratio				−0.147 ***	−0.502 ***	−0.179 ***	0.045	0.283 ***	0.039 **
				(−3.00)	(−13.11)	(−28.39)	(0.42)	(2.86)	(2.49)
rm							−0.023	−0.050 ***	−0.012 ***
							(−1.54)	(−3.70)	(−5.60)
hhi	−0.368 ***	−0.372 ***	−0.352 ***	−0.592 ***	−0.914 ***	−0.831 ***	−0.369 ***	−0.370 ***	−0.350 ***
	(−4.93)	(−8.23)	(−9.45)	(−7.85)	(−19.96)	(−21.98)	(−4.94)	(−8.19)	(−9.39)
mix	0.017 **	0.015 ***	0.013 ***	−0.003	0.012 **	−0.010 **	0.008	0.018 ***	0.016 ***
	(2.25)	(2.70)	(3.35)	(−0.37)	(2.03)	(−2.35)	(0.87)	(3.08)	(3.86)
sub	0.648 ***	0.809 ***	0.761 ***	0.900 ***	1.245 ***	1.174 ***	0.650 ***	0.808 ***	0.764 ***
	(4.34)	(9.32)	(10.48)	(5.94)	(14.05)	(15.84)	(4.34)	(9.32)	(10.51)
new	0.116 ***	0.063 ***	0.074 ***	0.130 ***	0.102 ***	0.109 ***	0.115 ***	0.063 ***	0.073 ***
	(5.17)	(6.13)	(8.09)	(5.70)	(9.68)	(11.65)	(5.15)	(6.10)	(8.00)
ex	−0.014	−0.012	−0.007	−0.016	−0.016 *	−0.010	−0.014	−0.012	−0.007
	(−0.49)	(−1.44)	(−0.86)	(−0.54)	(−1.90)	(−1.26)	(−0.47)	(−1.43)	(−0.86)
age	0.110 **	0.314 ***	0.211 ***	0.573 ***	1.449 ***	1.067 ***	0.112 **	0.312 ***	0.216 ***
	(1.97)	(8.47)	(7.44)	(10.50)	(40.81)	(38.57)	(2.02)	(8.42)	(7.62)
size	0.106 ***	0.190 ***	0.183 ***	0.0760 ***	0.239 ***	0.209 ***	0.106 ***	0.189 ***	0.182 ***
	(12.34)	(47.49)	(52.89)	(8.77)	(59.02)	(59.10)	(12.28)	(47.39)	(52.19)
liability	0.112 ***	0.096 ***	0.096 ***	0.123 ***	0.099 ***	0.102 ***	0.112 ***	0.096 ***	0.096 ***
	(36.37)	(83.49)	(92.61)	(39.52)	(84.77)	(96.19)	(36.35)	(83.51)	(92.64)
manage	0.241 ***	0.184 ***	0.198 ***	0.310 ***	0.256 ***	0.273 ***	0.241 ***	0.184 ***	0.198 ***
	(37.92)	(75.35)	(89.38)	(50.67)	(108.63)	(128.06)	(37.92)	(75.20)	(89.06)
ROA	0.932 ***	0.726 ***	0.763 ***	0.933 ***	0.761 ***	0.799 ***	0.932 ***	0.726 ***	0.762 ***
	(26.59)	(55.09)	(63.40)	(26.25)	(56.50)	(64.96)	(26.59)	(55.07)	(63.31)
cons	3.327 ***	3.138 ***	3.121 ***	3.435 ***	3.241 ***	3.226 ***	3.300 ***	3.133 ***	3.125 ***
	(49.62)	(127.75)	(141.24)	(39.84)	(129.10)	(141.96)	(26.65)	(127.39)	(139.74)
r2_w	0.173	0.247	0.241	0.150	0.213	0.207	0.173	0.247	0.241
r2_b	0.423	0.304	0.345	0.394	0.381	0.377	0.423	0.303	0.343
r2_o	0.444	0.320	0.360	0.422	0.396	0.398	0.444	0.320	0.358
F	778.3	5026.2	6370.6	657.5	4139.3	5240.2	659.2	4254.5	5398.0
P	0	0	0	0	0	0	0	0	0
96,897	358,884	456,692	96,897	358,884	456,692	96,897	358,884	456,692	96,897

Note: t statistical value in parentheses; *, **, *** represent the significance levels of 10%, 5%, and 1%, respectively.

, (1), (4), and (7) are state-owned enterprises; (2), (5), and (8) are private enterprises; and (3), (6), and (9) are overall enterprises. The regression results are as follows.

From Table 6, the coefficient of the marketization index is significantly positive; that is, the improvement of the marketization level is conducive to improving the TFP of enterprises, which is consistent with the research of [16,20]. The market-oriented transformation improves the efficiency of resource allocation within enterprises through incentive mechanism improvement, thus improving the micro-productivity of enterprises. The market-oriented transformation makes the input factors flow from low-productivity enterprises to high-productivity enterprises, thus improving the allocation efficiency of resources among enterprises, reducing resource misallocation, and improving the overall enterprise productivity. In Table 6, (4), (5), and (6) demonstrate that the coefficient of the state-owned proportion is significantly negative, indicating that the increase in the proportion of state-owned enterprises hinders the improvement of TFP. In Table 6 (8), the coefficient of the proportion of state-owned enterprises and the interaction term between the proportion of state-owned enterprises and the marketization index shows that marketization affects the TFP of enterprises through the change of the proportion of state-owned enterprises, and its mechanism is related to the degree of marketization; the higher the degree of marketization, the more obvious the effect of marketization. When the level of marketization is low, increasing the proportion of state-owned private enterprises is conducive to improving the TFP of enterprises. When the level of marketization is high, reducing the proportion of state-owned private enterprises is conducive to improving corporate TFP. Table 6 (9) shows that this mechanism is also validated across the enterprise. The above mechanism can be summarized, as marketization through changes in the proportion of state-owned plays a role. Table 6 (7) shows that this mechanism does not exist in state-owned enterprises. This also validates the conclusions drawn by [59] from the study of listed companies. The coefficient of Herfindahl–Hirschman Index is significantly negative, indicating that the product competition environment is conducive to enhancing the TFP of enterprises, and a monopoly will hinder the improvement of TFP. This is why it is necessary to encourage competition and promote market-oriented reforms. From Table 6 (7), (8), and (9), it can be observed that the overall mixing index is significantly positive, indicating that the breadth of participation is conducive to improving the TFP of enterprises. The reason may come from the mutual learning in the mixing process, or from the mutual checks and balances of different types of equity to improve the management efficiency.

The proportion coefficient of new products is significantly positive, indicating that the development of new products is conducive to promoting TFP; that is, innovation activities enhance the TFP of enterprises. The export share coefficient is not significant, which is different from the results of [16]. They believe that an important mechanism for increasing TFP is exporting, but they have not found strong evidence of the ‘productivity paradox’. The government subsidy coefficient is significantly positive, which is similar to the conclusion of [48], indicating that industrial policies and local support for enterprises generally promote the development of productivity. The coefficient of enterprise scale is significantly positive, indicating that it has scale effect. The survival age coefficient is significantly positive. The reason may be that Chinese enterprises have not been established for a long time and have not reached the high point of the inverted U-shaped relationship. The coefficient of the total debt is significantly positive, which indicates that the external financing promotes productivity growth. This phenomenon also appears in the results of [52]. The coefficient of the asset turnover is significantly positive, indicating that the higher the level of management, the higher the productivity of enterprises.

4.5. Analysis of Asset–Liability Ratio Regression Results

This paper uses Equation (6) to analyze the influencing factors of the asset–liability ratio. The Hausman test results based on Equation (6) demonstrate that it is not the consistency of FE and RE methods; thus, the author believes that the fixed effect model results should be used instead of the random effect model results. Therefore, this paper reports the results of the fixed effect model. Similarly, based on the research of [59], we classify and regress the samples according to the holding situation. In

Table 7. Asset–liability ratio (LEV) is the regression result of explained variable.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	State-Owned Enterprises	Private Enterprises	Overall	State-Owned Enterprises	Private Enterprises	Overall	State-Owned Enterprises	Private Enterprises	Overall
tfp	−0.028 ***	−0.034 ***	−0.035 ***	−0.036 ***	−0.046 ***	−0.046 ***	−0.028 ***	−0.034 ***	−0.035 ***
	(−36.12)	(−77.22)	(−90.56)	(−44.58)	(−100.82)	(−117.06)	(−36.15)	(−77.14)	(−90.32)
market	−0.025 ***	−0.033 ***	−0.031 ***				−0.022 ***	−0.033 ***	−0.032 ***
	(−60.22)	(−126.21)	(−146.97)				(−9.92)	(−124.42)	(−134.70)
ratio				−0.007	0.143 ***	0.043 ***	−0.002	−0.129 ***	−0.029 ***
				(−0.81)	(20.08)	(36.64)	(−0.13)	(−7.21)	(−10.28)
rm							−0.002	0.0236 ***	0.005 ***
							(−1.01)	(9.64)	(14.48)
hhi	0.023 *	0.040 ***	0.038 ***	0.083 ***	0.173 ***	0.154 ***	0.023 *	0.040 ***	0.037 ***
	(1.93)	(4.92)	(5.72)	(6.69)	(20.34)	(22.22)	(1.93)	(4.83)	(5.53)
mix	−0.008 ***	−0.012 ***	−0.010 ***	−0.007 ***	−0.012 ***	−0.004 ***	−0.010 ***	−0.014 ***	−0.010 ***
	(−6.68)	(−12.18)	(−13.73)	(−4.54)	(−10.91)	(−5.11)	(−6.90)	(−13.02)	(−14.47)
sub	−0.206 ***	−0.204 ***	−0.199 ***	−0.271 ***	−0.305 ***	−0.294 ***	−0.206 ***	−0.204 ***	−0.200 ***
	(−8.67)	(−12.96)	(−15.27)	(−10.94)	(−18.52)	(−21.60)	(−8.66)	(−12.95)	(−15.34)
new	−0.004	−0.011 ***	−0.010 ***	−0.007 **	−0.020 ***	−0.018 ***	−0.004	−0.011 ***	−0.010 ***
	(−1.21)	(−5.78)	(−6.12)	(−2.00)	(−10.18)	(−10.48)	(−1.23)	(−5.70)	(−5.91)
ex	0.011 **	−0.002	−0.001	0.011 **	−0.001	−0.001	0.011 **	−0.002	−0.001
	(2.33)	(−1.56)	(−1.02)	(2.34)	(−0.91)	(−0.46)	(2.34)	(−1.58)	(−1.05)
age	−0.009	−0.028 ***	−0.025 ***	−0.138 ***	−0.311 ***	−0.237 ***	−0.009	−0.027 ***	−0.025 ***
	(−1.06)	(−4.15)	(−4.99)	(−15.48)	(−46.97)	(−46.54)	(−1.03)	(−4.04)	(−4.93)
size	−0.046 ***	−0.044 ***	−0.048 ***	−0.036 ***	−0.054 ***	−0.053 ***	−0.046 ***	−0.044 ***	−0.047 ***
	(−33.30)	(−60.08)	(−76.78)	(−25.66)	(−71.39)	(−80.32)	(−33.31)	(−59.89)	(−75.18)
liability	0.182 ***	0.170 ***	0.170 ***	0.165 ***	0.154 ***	0.154 ***	0.182 ***	0.170 ***	0.171 ***
	(175.08)	(371.76)	(415.25)	(158.05)	(335.84)	(374.41)	(175.08)	(371.95)	(415.75)
manage	0.015 ***	0.013 ***	0.014 ***	0.012 ***	0.014 ***	0.013 ***	0.015 ***	0.013 ***	0.014 ***
	(29.10)	(63.16)	(72.46)	(23.84)	(61.36)	(67.52)	(29.10)	(63.12)	(72.21)
ROA	−0.083 ***	−0.072 ***	−0.074 ***	−0.076 ***	−0.072 ***	−0.075 ***	−0.083 ***	−0.072 ***	−0.074 ***
	(−14.74)	(−29.61)	(−34.02)	(−12.96)	(−28.55)	(−32.79)	(−14.75)	(−29.59)	(−33.90)
Tang	−0.033 ***	−0.038 ***	−0.041 ***	−0.022 ***	−0.032 ***	−0.034 ***	−0.032 ***	−0.038 ***	−0.041 ***
	(−6.94)	(−16.62)	(−20.57)	(−4.43)	(−13.51)	(−16.01)	(−6.93)	(−16.50)	(−20.51)
cons	−0.530 ***	−0.242 ***	−0.263 ***	−0.490 ***	−0.236 ***	−0.259 ***	−0.525 ***	−0.240 ***	−0.261 ***
	(−46.37)	(−49.33)	(−60.53)	(−33.23)	(−45.89)	(−56.61)	(−26.10)	(−48.90)	(−59.47)
r2_w	0.446	0.468	0.456	0.397	0.419	0.406	0.446	0.468	0.457
r2_b	0.256	0.290	0.298	0.245	0.354	0.344	0.256	0.289	0.295
r2_o	0.238	0.266	0.273	0.230	0.327	0.318	0.238	0.266	0.270
F	2530.9	11,394.4	14,252.7	2068.5	9344.6	11,642.8	2194.1	9889.1	12,384.7
P	0	0	0	0	0	0	0	0	0
N	96,897	358,884	456,692	96,897	358,884	456,692	96,897	358,884	456,692

*, **, *** represent the significance levels of 10%, 5%, and 1%, respectively.

, (1), (4), and (7) are state-owned enterprises; (2), (5), and (8) are private enterprises; (3), (6), and (9) are overall enterprises. The regression results are as follows.

The marketization index in Table 7 is significantly negative, indicating that the improvement of marketization will reduce the asset–liability ratio of enterprises, which is the direct impact mechanism of marketization on the asset–liability ratio of enterprises. With the improvement of China’s marketization, the participation of non-state-owned shareholders in the high-level governance of state-owned enterprises can alleviate government intervention, improve executive incentives and strengthen internal control, thus playing a governance role in the leverage manipulation of state-owned enterprises and reducing corporate leverage. The marketization of the interest rate improves the flow efficiency and allocation efficiency of financial resources, thus reducing the leverage ratio of enterprises. To achieve the goal of deleveraging, marketization promotes bank competition and reduces leverage by increasing the financing costs of zombie firms. The legalization of corporate governance promotes the optimization of capital structure, thereby reducing corporate leverage. Different from the conclusion of [18], they examined the impact of regional marketization factors on the asset–liability ratio of enterprises, and did not find that marketization had a significant impact on the total asset–liability ratio of enterprises. One may be affected by the lack of marketization index data. They examined the period from 1998 to 2004, but the marketization data of 2000, 2003, and 2004 were missing.

The TFP regression coefficient in Table 7 is significantly negative, indicating that the increasing TFP helps to reduce the asset–liability ratio of enterprises, that is, the self-financing mechanism, which is of great significance to the current economic restructuring and deleveraging. If enterprises want to deleverage, a very important way is to increase the total factor productivity of enterprises. The improvement of total factor productivity brings economic growth. On the one hand, it helps to reduce the macro leverage ratio. On the other hand, it will also increase the return on investment and reduce the micro asset–liability ratio by improving the equity of owners. The coefficient of marketization in Table 6 is significantly positive and the TFP regression coefficient in Table 7 is significantly negative, which indicates that marketization can affect the internal governance of enterprises to improve productivity and reduce the debt ratio of enterprises. In Table 7 (5), the coefficient of the proportion of state-owned shares is significantly positive, indicating that as the proportion of state-owned shares increases, the asset–liability ratio also increases. This also confirms the research of [35] that state-owned equity can provide more economic resources for private enterprises. Different types of corporate lending face different financing constraints [2,15,17,36]. The mixed reform of state-owned enterprises and private enterprises is conducive to improving capital mismatch and reducing the overall corporate asset–liability ratio, which the authors call the debt structure adjustment mechanism. However, from Table 7 (8), in the proportion of state-owned enterprises and the coefficient of the interaction term between the proportion of state-owned enterprises and the marketization index, it can be observed that the marketization affects the asset–liability ratio of enterprises through the change of the proportion of state-owned enterprises, and the effect is related to the degree of marketization. When the level of marketization is low, the development of private enterprises due to financing constraints restricts the development of TFP, the self-financing mechanism is blocked. At this time, private enterprises to increase the proportion of state-owned, financing constraints can be eased, thereby enhancing the corporate TFP. Then, the self-financing mechanism can play a greater role, and is conducive to reducing the corporate asset–liability ratio. When the level of marketization is high, private enterprises can finance in the market. The increase in the state-owned proportion means the increase in government intervention, which will reduce the TFP of enterprises. The self-financing mechanism is frustrated, which is not conducive to reducing the asset–liability ratio of enterprises. Table 7 (9) shows that this mechanism is also verified in the overall enterprise. At the same time, it can be observed from Table 7 (8) that the role of the self-financing mechanism has been strengthened, thus accelerating the optimization of the corporate debt structure. The policy significance is that speeding up the marketization process is conducive to the effect of equity reform. The Herfindahl–Hirschman Index is significantly positive, indicating that the product competition is conducive to reducing the asset–liability ratio of enterprises, and the market monopoly will increase the asset–liability ratio of enterprises, which may be one of the reasons why state-owned enterprises obtain more capital. The mixed index is significantly negative, indicating that the breadth of participation is conducive to reducing the asset–liability ratio of the enterprise, thereby reducing the debt ratio of the enterprise, which verifies the conclusions of [60]. The possible reason for that is that the mixing of different types of equity improves the productivity of the enterprise. The results in Table 6 (8) and (9) demonstrate that the mixed index is significantly positive and may also be the result of checks and balances between different types of equity. In Table 7 (4) and (5), the role of the state-owned proportion is different. In state-owned enterprises, the increase in the state-owned proportion does not bring about the increase in the asset–liability ratio. In private enterprises, the increase in the state-owned proportion brings about the increase in the asset–liability ratio. The impact of the export scale on the asset–liability ratio is also different in state-owned enterprises and private enterprises, which verifies the conclusions of [59]. The coefficient of the return on assets is significantly negative, indicating that the enhancement of the corporate profitability is conducive to reducing the asset–liability ratio of enterprises. The coefficient of the liquidation ratio is significantly negative, which is different from the results of [19], but consistent with the conclusions of [18]. This may be an issue that future research needs to explain.

From the above empirical results, it can be observed that marketization and property rights reform are of great significance to improve corporate performance. Marketization not only affects enterprises directly, but also affects the enterprise performance through the change of the proportion of state-owned enterprises. The premise that property rights reform plays a good role is that the marketization process should reach a certain height. Otherwise, the property rights reform will not enable enterprises to reduce debt ratio and increase TFP, but will hinder TFP growth and further increase the corporate debt ratio. This means that the marketization and property rights reform should be coordinated and synchronized, otherwise, the goal of economic restructuring cannot be achieved. This has guiding significance for our current structural adjustment. The government should improve the degree of marketization while carrying out the property rights reform. Through mixed ownership and the use of state-owned capital to invest and operate companies, it should focus on managing capital rather than assets, and participate in stocks rather than necessarily holding shares. This is in order to improve the efficiency of the state-owned capital operation by changing the direction of state-owned capital investment and adjusting the layout.

4.6. Further Discussion

Empirical studies often encounter endogenous problems caused by missing variables. In addition to using the fixed effect model to control the factors that do not change with time, the authors also use the instrumental variable estimation method, using innovation [61,62] (new product ratio) and export (export product ratio) as the instrumental variables of TFP, because innovation and export are often encouraged by the government, generally implemented in the form of tax incentives, and independent of the financial capabilities within the enterprise; however, according to the theoretical and empirical research discussed previously, these two variables will affect the TFP of the enterprise. Therefore, innovation (new product ratio) and export (export product ratio) can be used as instrumental variables of TFP. There is no fundamental change in the instrumental variable regression results (see

Table 8. The debt-to-assets ratio (lev) is estimated by the instrumental variable method for the explained variable, taking into account the use of innovation and exports as a tool for TFP, and using the fixed effect model to estimate and control the constant characteristics of the firm (regression results classified by proportion of state-owned).

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	State-Owned Enterprises	Private Enterprises	Overall	State-Owned Enterprises	Private Enterprises	Overall	State-Owned Enterprises	Private Enterprises	Overall
tfp (IV)	−0.0700 **	−0.189 ***	−0.167 ***	−0.098 ***	−0.232 ***	−0.209 ***	−0.071 **	−0.187 ***	−0.164 ***
	(−2.19)	(−4.96)	(−6.10)	(−3.17)	(−8.66)	(−9.94)	(−2.20)	(−4.91)	(−5.96)
market	−0.021 ***	−0.014 ***	−0.015 ***				−0.018 ***	−0.014 ***	−0.016 ***
	(−7.73)	(−2.93)	(−4.72)				(−4.42)	(−2.99)	(−5.05)
ratio				−0.015	0.049 ***	0.013 ***	−0.001	−0.084 ***	−0.024 ***
				(−1.58)	(2.91)	(3.21)	(−0.04)	(−3.21)	(−6.63)
rm							−0.003	0.016 ***	0.004 ***
							(−1.29)	(4.11)	(6.92)
hhi	0.008	−0.018	−0.009	0.047 **	0.002	0.019	0.008	−0.017	−0.008
	(0.47)	(−0.98)	(−0.67)	(2.12)	(0.09)	(0.97)	(0.46)	(−0.98)	(−0.67)
mix	−0.007 ***	−0.010 ***	−0.008 ***	−0.007 ***	−0.010 ***	−0.005 ***	−0.009 ***	−0.011 ***	−0.008 ***
	(−5.36)	(−6.98)	(−8.46)	(−4.34)	(−6.14)	(−5.32)	(−6.32)	(−7.18)	(−8.62)
sub	−0.179 ***	−0.081 **	−0.099 ***	−0.216 ***	−0.075 *	−0.104 ***	−0.179 ***	−0.082 **	−0.102 ***
	(−5.60)	(−2.17)	(−3.74)	(−5.68)	(−1.83)	(−3.38)	(−5.56)	(−2.20)	(−3.85)
age	−0.005	0.020	0.003	−0.103 ***	−0.043	−0.064 ***	−0.004	0.020	0.003
	(−0.47)	(1.38)	(0.32)	(−5.17)	(−1.07)	(−2.72)	(−0.43)	(1.38)	(0.36)
size	−0.041 ***	−0.014 *	−0.023 ***	−0.031 ***	−0.009	−0.018 ***	−0.041 ***	−0.015 **	−0.023 ***
	(−10.85)	(−1.95)	(−4.54)	(−10.81)	(−1.42)	(−3.95)	(−10.86)	(−1.98)	(−4.56)
liability	0.192 ***	0.197 ***	0.196 ***	0.183 ***	0.200 ***	0.197 ***	0.192 ***	0.197 ***	0.195 ***
	(25.58)	(29.17)	(37.08)	(19.78)	(29.96)	(35.21)	(25.50)	(29.16)	(37.04)
manage	0.019 ***	0.028 ***	0.026 ***	0.020 ***	0.032 ***	0.030 ***	0.019 ***	0.028 ***	0.026 ***
	(5.31)	(7.74)	(10.03)	(5.21)	(12.01)	(13.94)	(5.30)	(7.69)	(9.87)
ROA	−0.044	0.041	0.027	−0.019	0.070 ***	0.056 ***	−0.044	0.040	0.025
	(−1.45)	(1.46)	(1.29)	(−0.63)	(3.38)	(3.27)	(−1.44)	(1.42)	(1.18)
Tang	−0.045 ***	−0.074 ***	−0.074 ***	−0.043 ***	−0.082 ***	−0.080 ***	−0.045 ***	−0.074 ***	−0.073 ***
	(−4.20)	(−7.85)	(−10.29)	(−3.61)	(−10.33)	(−12.09)	(−4.19)	(−7.77)	(−10.14)
cons	−0.383 ***	0.265 **	0.172 *	−0.265 **	0.401 ***	0.298 ***	−0.378 ***	0.261 **	0.164 *
	(−3.40)	(2.12)	(1.91)	(−2.33)	(4.38)	(4.15)	(−3.31)	(2.09)	(1.81)
r2_b	0.290	0.367	0.366	0.294	0.347	0.353	0.290	0.367	0.365
r2_o	0.271	0.338	0.338	0.278	0.319	0.324	0.272	0.339	0.338
F	0	0	0	0	0	0	0	0	0
N	96,897	358,884	456,692	96,897	358,884	456,692	96,897	358,884	456,692

*, **, *** represent the significance levels of 10%, 5%, and 1%, respectively.

), indicating that our results are robust.

5. Conclusions

5.1. Research Conclusions and Policy Implications

The empirical studies have found that the improvement of marketization is conducive to improving corporate productivity and improving corporate debt problems. In addition to the direct impact on the corporate debt ratio and productivity, the degree of marketization also improves corporate debt by affecting corporate productivity and the internal ownership structure. However, the premise of the mechanism is that the marketization process should reach a certain height, which means that the local marketization and enterprise property rights reform should be coordinated and synchronized; otherwise, it cannot achieve the goal of structural adjustment. The key to the current reform is to improve the degree of local marketization to reduce the misallocation of resources caused by financial market frictions, so that productivity and resources can be more effectively matched. A consensus has been reached, acknowledging that mixed ownership reform can not only improve the financing constraints of private enterprises in the financial market, but also solve the problem of inactive systems and mechanisms of state-owned enterprises, thereby increasing the productivity of enterprises and improving corporate debt. The mixed ownership reform can also improve the layout and structural adjustment of the state-owned economy, and achieve the strategic goal of the state-owned economy.

The research in this paper shows that there are three ways to improve the debt problem of enterprises by improving the degree of marketization. The first is the direct effect; that is, the market-oriented institutional reform and business environment improvement improves the debt problem of enterprises. The second is the self-financing mechanism; that is, the degree of marketization improves the corporate debt problem by enhancing the TFP of enterprises and enhancing the self-financing ability of enterprises. Third, there is the debt structure adjustment; that is, marketization promotes the reform and adjustment of the corporate property rights structure, thereby reducing the capital mismatch and improving the overall corporate debt structure. Combined with the mechanism discussed in this paper, the authors believe that the following policies and reform measures can be taken to improve the corporate debt problem. First of all, from the perspective of the direct mechanism, we should improve the financial market system and legal system construction, develop investment banks and stock markets, and use market-oriented equity financing to solve the investment gap. At the same time, we should accelerate the introduction of debt disposal regulations and strengthen corporate responsibility, in accordance with market rules to eliminate zombie enterprises. Secondly, from the perspective of the self-financing mechanism of marketization to enhance the TFP of enterprises, the process of marketization should be accelerated, and the competition mechanism should be introduced. At the same time, the tax side can also make arrangements conducive to the enterprise innovation and improve the productivity level of enterprises. Finally, in order to promote the reform of the enterprise property structure and adjust the debt structure adjustment from the market, we should promote the reform of the mixed ownership of state-owned enterprises, and through equity restructuring to promote debt restructuring, let the debt structure adjustment mechanism play a role.

5.2. Limitations and Future Perspectives

This study uses the data of industrial enterprises from 1998 to 2007, which can only be applied to the period before the 2008 financial crisis, and cannot reflect the situation after the financial crisis. Future research can use the data of listed companies to expand the scope of the data. It can not only study before the financial crisis, but also after the financial crisis, to observe whether there are structural changes to verify whether the mechanism still exists. At the same time, the data can also cover the epidemic to verify whether the deleveraging mechanism found in this paper is still effective.