Abstract
With the rapid development of the digital economy, digital finance, as a financial innovation combining Internet information technology with traditional finance, plays an essential role in the financial risk of microenterprises and macroeconomic operations. In this paper, the digital financial inclusion index at the provincial level is matched with the microdata of listed companies in Shanghai and Shenzhen stock markets. And, the panel data from 2011 to 2020 are set up from the theoretical and empirical analysis of digital finance on the impact of enterprise financial risk and its mechanism. Firstly, the development of digital finance in China has significantly reduced enterprise financial risk. In order to control the endogeneity, the Bartik instrumental variable is used to select the instrumental variable. Secondly, financing constraint is the function mechanism of digital finance to reduce enterprise financial risk. Thirdly, for enterprises with low debt levels and enterprises located in the eastern region, digital finance plays a more critical role in reducing financial risk.
1. Introduction
Security and stability are crucial factors for maintaining sustained and high-speed economic growth. In the context of unprecedented changes in a century, keeping stable productivity-induced economic growth and reducing uncertainty risks are of vital significance for high-quality economic development. Enterprises are the backbone to promote the growth of the real economy, and financial activities provide an indelible impetus to the growth of enterprises [1]. In reality, the unreasonable financing model and the mismatch between financial supply and demand have hindered the improvement of the quality and efficiency of China's financial industry in serving the real economy [2]. The traditional financial sector is confronted with some structural problems due to a lack of its own conditions and technology. Therefore, when financing for enterprises, the so-called “liquidity stratification” of financial resources emerges, which primarily consists of three mismatches, namely, domain, attribute, and stage.
In the new era, traditional finance can solve the financial risk and financing constraints faced by enterprises through innovative integration with new technologies and other new things. China's Fintech and Digital Financial Inclusion Development Report (2020) points out that the level of the Fintech industry in China has already ranked at the forefront of the world in China and the Fintech financing in China accounts for 52.7% of the global share. Since the launch of the Fintech Development Plan (2019–2021), banking-related Fintech companies have started to develop vigorously, and the Fintech Supervisory Sandbox has rapidly launched. China has benefited from an enormous market and new infrastructure, and the financial industry has accelerated the digitalization process. Meanwhile, the sector of digital finance is flourishing, which draws the attention of domestic and foreign scholars [3, 4]. At present, a large number of related literature studies have been made, but most of them focus on the relationship and mechanism between digital finance and total factor productivity, the influence on household financial market participation, and bank performance [5–7]. Some of them find that digital finance can reduce the risk of stock price collapse and enterprise risk-taking [8, 9]. The aforementioned studies provide some theoretical basis and techniques for this paper to explore the relationship between digital finance and enterprise financial risk. However, there is still a lack of research on the actual impact of digital finance on enterprise financial risk. Further research on this issue has great theoretical significance and practical value as China's economy enters a new normal.
The paper mainly aims to examine the influence of digital finance on enterprise financial risk and its mechanism. Moreover, it attempts to analyze whether digital finance corrects the distortion of traditional financial elements and whether it can effectively improve enterprise financial risk and further verify the inclusion and transmission mechanism of digital finance. On the one hand, the in-depth exploration of such issues is helpful to theoretically prove that digital finance can reduce the enterprise financial risk; on the other hand, it also provides empirical support for the policy-making of China's financial supervision.
2. Literature Review and Research Hypotheses
The extended definition of enterprise financial risk is the process in which the probability of enterprise financial distress is constantly expanding. Among them, financial distress is generally depicted as the situation when enterprises are unable to repay their debts. Financial risk derivation mainly comes from the changes of two internal factors, namely, the capital structure and the value creation ability of enterprises. The competitive environment and regulatory environment are external factors of enterprise value creation, naturally constraining or improving the changes in its financial risk. In addition, transaction constraints, information constraints, and political or administrative constraints can also cause enterprise managers to encounter obstacles when implementing their preferred policies. According to the classical Free Cash Flow Theory, debt financing certainly has the advantages of reducing free cash flow constraints and curbing agency [10]. In this regard, enterprise managers will release the signal of “a clean hand wants no washing” to the outside for the benefit of the enterprise's development. It will be more inclined to make the decision to use financing. Without the support of financial markets and other systems, the production and operation of enterprises will be in a dilemma and then the financial risk of enterprises will more likely to fall into high-level bankruptcy risks.
The financial inclusion policy aims to provide more equitable and convenient financial services for low-income and disadvantaged groups. Digital inclusive finance promotes inclusive finance through digital financial services, which can make full use of modern digital information technology, expand the scope of financial services, reduce the cost of financial services, enhance the convenience and penetration of financial services, and improve efficiency and level of financial services. From a microperspective, the development of digital inclusive finance can enable small, medium, and microenterprises that were originally located at the bottom of the economic pyramid to enjoy more financial services at lower costs. From a macroperspective, the development of digital inclusive finance can alleviate the contradiction between financial and economic structures and create favorable conditions for building a new development pattern in which domestic and international dual cycles promote each other.
Hsu et al. [11] argue that the financial element is a major element for microentities to carry out their business activities and the matching degree of financial supply can severely affect the progress of enterprise activities. However, China's existing financial system still has some structural problems. High-quality development has been seriously restricted since traditional finance has not played its proper role. However, digital finance has a huge impact on traditional finance and real life. Gomber et al. [12] consider that digital finance can process big data at low cost and low risk with the support of technical tools such as the Internet, big data, and artificial intelligence, thus lowering the threshold of access to financial services for long-tail groups. At the same time, the high-quality development of digital finance also has the function of empowerment, and its technical tools can empower the development of enterprises, enable them to make optimal decisions, accelerate their development, and help them optimize their development path towards high-quality development [13].
First of all, theoretically speaking, digital finance is a new type of financial innovation which exerts extensive influence on real life and subverts the traditional financial system to some extent. Digital finance uses technologies such as artificial intelligence to establish a data warehouse by improving algorithms and evaluation mechanisms [14], construct a transparent and information-based credit system [15, 16], introduce innovative models or tools to enhance the efficiency of capital allocation in the financial sector [13, 17], and strengthen risk warning and management capabilities. However, digital finance can reduce spatial and temporal constraint, serve the fields that are difficult to cover by traditional finance [18], alleviate the mismatch of credit resources [19], and thus provide support for enterprises to reduce financial risk. Furthermore, as a key factor affecting the investment and cash flow of enterprises, financing constraint plays an important role in enterprise financial risk. The existing literature holds that financing constraints can reduce enterprise financial risk [20, 21] and digital finance can reduce information asymmetry, thus alleviating financing constraints. Then, financing constraints may be the mechanism of digital finance affecting financial risk. Finally, digital finance itself has the function of a financial accelerator [22] and also has the characteristics of improving efficiency and increasing risk probability. Because the industries, regions, and other individual natures of enterprises are significantly different, the influence of digital finance may also be significantly different, so efforts should also be made to explore the policy performance and differentiated effect of digital finance. Hence, it is more realistic to consider the heterogeneity when studying the role of digital finance in enterprise financial risk. On this basis, the following hypotheses are proposed: H1: digital finance can reduce enterprise financial risk H2: the effect of digital finance on reducing enterprise financial risk is significantly different due to the heterogeneity of enterprises H3: digital finance reduces the enterprise’s financial risk by alleviating the financing constraint
3. Research Design
3.1. Data Sources
This paper takes A-share listed companies in the Shanghai and Shenzhen stock markets from 2011 to 2020 as the research object and matches their data with digital finance index to form a panel data set. After referring to the existing literature, this paper cleans the data. Firstly, this paper excludes the enterprises that are listed on ST and delisted. Secondly, financial enterprises are excluded because this paper studies the influence of digital finance. Thirdly, since participating in an IPO has a significant impact on the finance and development of enterprises, enterprises experiencing an IPO during the sample period are excluded. Fourthly, considering the influence of abnormal values on the regression results, this paper uses a two-sided 1% winsorization of continuous variables. Among them, the data related to listed companies all come from the CSMAR database, and the digital finance index comes from the Digital Financial Inclusion Index, which is the data publicly released by Peking University.
3.2. Specification of Variables
3.2.1. Explained Variable
Enterprise financial risk: in this paper, the revised Altman Z value is used to measure enterprise financial risk. The comparative advantage of this index is that it has less interference, eliminates the interference of the stock market index on financial risk measurement, and is more suitable for evaluating enterprise financial risk under the new financial norm. The specific calculation formula is as follows: enterprise financial risk = (3.3 earnings before interest and tax + 1.0 sales revenue + 1.4 retained earnings + 1.2 working capital)/total assets. Graham (2000) and Byoun (2008) think that the larger the revised Altman Z value, the smaller the financial risk correspondingly [23, 24].
3.2.2. Core Explanatory Variable
Digital finance: the measurement of digital finance refers to the practice of Guo et al. [25]. This paper adopts the Digital Financial Inclusion Index of Chinese Provinces compiled and measured by Institute of Digital Finance, Peking University. In the core empirical part of this paper, the digital financial inclusion index at the provincial level is adopted and the logarithmic processing is carried out.
3.2.3. Mechanism Variable
Financing constraint: by referring to Hadlock and Pierce [26], this paper employs the SA index to measure the financing constraints of enterprises. The specific calculation formula is as follows: cooperate financing constraint = (−0.737 size + 0.043 size 2–0.04 age). The smaller the SA index, the greater the financing constraint on enterprises.
3.2.4. Control Variables
To avoid the influence of missing variables on the regression results as much as possible, this paper introduces a series of microvariables from the perspective of enterprises. The level is the proportion of total liabilities to total assets. Size is the natural logarithm of enterprise assets. Rate is the profit rate of main business. Cash is the proportion of cash to total assets. Capital is the proportion of total assets to main business. Equity is the concentration of the largest shareholder. From the results of VIF expansion factor in Table 1, it can be seen that VIF is all less than 10, so there is no serious multicollinearity in this empirical study. The descriptive statistics of the variables are demonstrated in Table 2.
3.3. Model Setting and Empirical Strategies
The following basic model is constructed:
In Model (1), the explained variable is enterprise financial risk, which is measured by the revised Altman Z value; the core explanatory variable is the provincial digital finance index; control variables include the set of control variables described earlier; and is the random interference term of the model.
In the empirical study, this paper also carries out the following actions. First, considering that enterprises in the same province have some common characteristics that cannot be observed, the robust standard error of clustering to provinces is adopted in the regression test to eliminate the interference of this factor. Second, this paper adopts the triple fixed-effect model for empirical study so as to absorb the relevant fixed-effect and gain robust results.
In terms of the robustness test and endogenous treatment, this paper takes the following measures: (1) The year-industry-region joint fixed-effect is tested. As year, industry, and region are separate fixed-effects, joint effect is more common in the actual situation. Therefore, year, industry, and region are jointly tested for fixed effect. (2) Municipalities directly under the Central Government are excluded. Due to the vast territory of China and the extremely uneven development of digital technology in different regions, the growth effects of digital finance development may be quite different. Therefore, the data of Beijing, Tianjin, Shanghai, and Chongqing are removed for further reexamination. (3) In the endogeneity test, this research adopts the instrumental variable method to test, refers to the practice of Bartik, uses the existing data to construct an instrumental variable (specifically, use the product of multiplying the first difference Δ of the digital finance index of the last period, namely, t-1 period and that of the digital finance index in time), and then takes it as an instrumental variable.
4. Empirical Results and Economic Explanation
4.1. Baseline Regression
First of all, the results of the baseline regression are presented. The marginal effect of the digital finance index is 0.256, and it is significant at the level of 1%, which indicates that the development of digital finance obviously increases the Z value measuring enterprise financial risk. Meanwhile, the larger the Z value, the lower the enterprise financial risk, which shows that the development of digital finance promotes the reduction of enterprise financial risk. The coefficient of level is 0.116, but it is not significant. The coefficient of size is −0.378, which is significant at the level of 1%, indicating that the enterprise financial risk gradually expands with the increase of enterprise scale. The coefficient of cash is significant at the level of 1%, which is 0.573, suggesting that the expansion of cash flow lowers the enterprise’s financial risk, which is consistent with the research of Chen et al. and Weifeng He et al. [27, 28]. The coefficient of capital is significant at the level of 1%, and the marginal effect is −0.402, demonstrating that the more concentrated the capital distribution of enterprises, the greater the enterprise financial risk, which is consistent with the research of Zhong and Yong [29].
4.2. Subindex Regression
To more comprehensively identify the role of digital finance in enterprise financial risk, this paper decomposes the digital finance index into the coverage breadth of digital finance and the usage depth of digital finance. On this basis, it further analyzes which aspect of digital finance development is more conducive to reducing enterprise financial risk, namely, the role of digital finance in reducing enterprise risks demonstrated in Table 3. Is it because digital finance covers a large audience or the service of digital finance is more sophisticated? Table 4 displays the corresponding estimation results.
It is found that the breadth index (DIF-B) has a positive driving effect on reducing enterprise financial risk (the coefficient is positive and passes the 1% statistical significance test). At the same time, the depth index (DIF-D) also plays a prominent role in reducing enterprise financial risk (the coefficient is positive and passes the 1% statistical significance test). Comparing the coefficients of DIF-B and DIF-D, we can find that the depth index exerts a more significant effect in reducing enterprise financial risk, which is consistent with the research of Fei Wu et al. (2020). The empirical test in Table 4 illustrates a lot of information. The development of digital finance needs both breadth and depth to provide better services for SMEs and achieve high-quality economic development.
4.3. Mechanism Analysis
It has been proved that digital finance can reduce enterprise financial risk, so what is the mechanism by which digital finance affects enterprise financial risk? To reveal this mechanism, the financing constraint is selected as the mechanism variable to carry out the test according to the previous theoretical analysis.
Table 5 shows the corresponding results. Among them, the results in Columns (1), (3), and (5) show that both digital finance itself and its breadth and width can significantly increase the SA index, which is significant at the level of 1%. The coefficients of the SA index are all negative, and the increase of the SA index means the decrease of financing constraints. In other words, digital finance can reduce the financing constraints of enterprises. Columns (2), (4), and (6) present the results of adding the SA index to Model (1). It can be seen that the coefficients of SA are all positive and significant at the level of 5%, showing that digital finance reduces enterprise financial risk by alleviating the financing constraint of enterprises. In this sense, the financing constraint is the mechanism through which digital finance affects enterprise financial risk. In addition, in Columns (2), (4), and (6), the coefficients of digital finance still show positive significance, indicating that there are other mechanisms.
4.4. Heterogeneity Analysis
In the baseline regression, it has been found that the level is not significant because the nature of listed companies varies and their debt levels may be significantly different, which will have different effects on enterprise financial risk, thus affecting the regression results. In view of this, this paper considers the heterogeneity of enterprise debt levels and analyzes their heterogeneity. In addition, due to the vast territory of China, the economic development level, policy conditions, and other environments in different places are also different, so the heterogeneity of the provinces where listed companies are located is also considered. The endowment of resource factors and the degree of economic development in different regions of our country are different, resulting in a large gap in the level of financial development among regions. In areas with low levels of financial development, the efficiency of capital allocation is low and the development of traditional financial institutions is insufficient, thus inhibiting the development of small- and medium-sized enterprises. Digital inclusive finance can make up for the differences in the level of financial development between regions, help to promote the liberalization of interest rates, and increase the risk-taking preference on the asset side of banks, which is conducive to the formation of a good financial ecological environment.
Table 6 shows the corresponding results. Among them, the coefficient of digital finance (DIF) in Column (1) is 0.297, which is significant at the level of 1%. For enterprises with a high debt level in Column (2), the coefficient of digital finance on the financial risk is 0.188, but it is only significant at a level of 10%, which suggests that digital finance has a more significant effect on reducing enterprise financial risk when the debt level of enterprises is low. As for the heterogeneity of location, in the fourth column, the coefficient of digital finance (DIF) in the eastern region is significantly positive at the level of 1%, while that in the noneastern region is only significant at the level of 5% and the coefficient is smaller, indicating that digital finance has the strongest effect on reducing enterprise financial risk when the enterprise is located in the eastern region. This may also be associated with the development of digital finance itself, so digital finance has the highest level in the eastern developed areas. To make the results of this research more robust, the interactive model is introduced, in which size is a dummy variable from 0 to 1. When it is 1, it means that the enterprise is at a high debt level. Meanwhile, east is a dummy variable from 0 to 1. When it is 1, it means that the enterprise is in the eastern region. In the results in Columns (3) and (6) of Table 6, the coefficient of DIFsize is negative and the coefficient of DIFeast is positive, which verifies that the previous conclusion is robust.
4.5. Extended Analysis and Robustness Test
4.5.1. U-Shaped Relationship
Existing research holds that the role of digital finance may have nonlinear characteristics, so this paper also attempts to explore whether digital finance has nonlinear characteristics when it affects enterprise financial risk. Column (1) of Table 7 shows the regression results of the model. The coefficient of the DIF L (1) is negative, which is significant at the level of 5%, and the coefficient of DIF L (2) is positive, which is significant at the level of 1%. It shows that the effect of digital finance on the reduction of enterprise financial risk decreases first and then increases, and they maintain a “U” shaped relationship. This paper also shows the quadratic curve fit diagram of the Z value of digital finance and that of financial risk. From Figure 1, it can be seen that digital finance has a nonlinear influence on enterprise financial risk.
4.5.2. Robustness Test
First of all, the municipalities of Beijing, Tianjin, Shanghai, and Chongqing are excluded. The test estimation results are shown in Column (2) of Table 7. It can be seen that the coefficient of DIF is 0.263, which has no significant change compared with 0.256 in the baseline regression, and its estimation value is still significant at the level of 1%. Therefore, it can be considered that the parameter estimation and significance have not changed significantly, verifying that the results of this paper are robust. Secondly, the joint fixed test is conducted, and the estimation results are shown in Column (3) of Table 5. It can be seen that digital finance still plays a positive role in reducing enterprise financial risk, which also indicates that the results of this paper are robust. Finally, the instrumental variable method is employed to test the endogeneity. The value of the Hausman test shows that it is significant at a level of 1%, which suggests that this variable is suitable for being an instrumental variable. The weak instrumental variable value of Kleibergen-Paap is 478.790, much higher than the threshold value of the 10% confidence level of 16.38, which shows that there is no problem in the weak instrumental variable. From the results of instrumental variables, it is not difficult to find that the development of digital finance still has a positive impact on reducing enterprise financial risk shown in Figure 1.
5. Conclusions
With the rapid development of the digital economy, the changes brought about by digital finance have had a huge impact on the development of traditional finance. Meanwhile, enterprises are the lifeblood of China's economy, and financial support for the development of enterprises is exceptionally vital for maintaining the smooth operation of the economy. Based on the data of A-share listed companies in Shanghai and Shenzhen stock markets from 2011 to 2020, this paper empirically explores the impact of digital finance on enterprise financial risk. The main conclusions are as follows. First, the development of digital finance can significantly reduce enterprise financial risk. It has significant characteristics of “hierarchy,” in which the depth of digital finance is more significantly conducive to reducing enterprise financial risk. Second, digital finance reduces enterprise financial risk by alleviating the financing constraint, which is a crucial mechanism for digital finance to reduce enterprise financial risk. Third, for enterprises with low debt levels and enterprises in the eastern region, digital finance plays a more significant and stronger role in reducing their financial risk. Fourth, considering the economic environment and other advantageous factors of municipalities, Beijing, Tianjin, Shanghai, and Chongqing are excluded, and the conclusion is still robust. Given that the triple fixed-effect model is “flexible” and is not strong enough in endogeneity control, the high-level joint fixed-effect model of “year-industry-region” is adopted and the conclusion is still robust. Furthermore, Bartik's concept of instrumental variables is adopted and set for regression, and the conclusion is still robust. The above-mentioned aspects all show that the research conclusion of this research is robust.
The policy implications of the conclusions are as follows: On the one hand, the rapid development of digital finance provides stable financial support for enterprises and lowers the market access threshold, which is of great significance for the high-quality development of SMEs. Hence, the development of digital finance should be accelerated to make it better serve the real economy. On the other hand, the transparent big technology credit mechanism of digital finance can precisely identify the source and direction of loans, thereby facilitating the supervision and establishment of risk warning mechanisms. Therefore, it is necessary to strengthen financial supervision and improve the risk identification system to maintain financial stability. Furthermore, it improves the digital economy development policy and strengthens the construction of digital financial infrastructure. It promotes the development of digital inclusive finance in various regions, especially in the central and western regions, northeastern regions, and areas with poor digital financial inclusion. It activates the economic development potential of small- and medium-sized enterprises and promotes the high-quality development of the real economy by improving the financial sustainability of small- and medium-sized enterprises.
Data Availability
The data used to support the findings of this study are available from the author upon request.
Conflicts of Interest
The author declares no conflicts of interest.
Acknowledgments
This research was financially supported by the National Social Science Fund of China, under grant no. 18BJY249.