Digital financial inclusion, Financial Mismatch and Small and medium-sized enterprises Financing Constraints

: Digital financial inclusion through the use of big data, artificial intelligence and other emerging technologies is becoming increasingly sophisticated in China. Using a sample of 843 SMEs in Shenzhen Stock Exchange from 2012-2021, this paper investigates the impact of digital financial inclusion on SMEs' financing constraints and its mechanism of action using a fixed effects model. It is found that digital financial inclusion can significantly alleviate the financing constraints of SMEs. Further analysis of the mechanism of action reveals that digital financial inclusion can indeed reduce the level of financing constraints by alleviating the financial mismatch of enterprises. The heterogeneity analysis finds that this mitigating effect is greater among non-state versus eastern SMEs. This study sheds new light on alleviating the financing constraints of SMEs and improving financial mismatch.


Introduction
SMEs play a very important role in the country's economic development and they are an essential part of the market economy, but the financing needs of SMEs are also increasing with the growth of the economy, but the problems of expensive and difficult financing they face has still not been fundamentally solved. Especially since the coronavirus swept the world in 2020, SMEs have faced shutdowns, disruptions in logistics and transportation, and unprecedented challenges to business operations, which have made the financing problems they face even more severe. Premier Li Keqiang has repeatedly stressed that SMEs are an important part of the real economy, and every effort should be made to ensure their survival. [China government website, Li Keqiang: We should do everything possible to let small, medium and micro enterprises and individual households survive, March 20, 2020, http://www.gov.cn/premier/202003/20/content_5493614.htm] In the face of the current financing difficulties of SMEs, whether a "new engine" can be intervened to make finance better serve the real economy and thus enhance the accessibility of the real economy has become one of the key concerns of the government and the academia at present. In addition, China's SMEs also face an unbalanced allocation of financial resources, which severely inhibits their development and access to finance. Unlike other countries, China has its own unique institutional nature, which makes it easier for resources in the market to be tilted towards state-owned enterprises and large-scale enterprises. The emergence of digital financial inclusion is to make more enterprises receive financial resources. Can digital financial inclusion, with its low cost, low threshold, and high popularity, alleviate SMEs' financing difficulties, and what is the underlying mechanism for this? This is a question worth exploring. Compared with existing studies, the possible contributions of this paper are: first, this paper explores the impact of macro-level digital financial inclusion development on corporate financing constraints using a sample of 843 SMEs in the Shenzhen Stock Exchange from 2012-2021, expanding the research on alleviating corporate financing constraints. Second, it explores the mechanism of the role of the two from the perspective of financial mismatch, and provides a basis for theoretical analysis to better utilize the function of digital financial inclusion to serve SMEs and optimize resource allocation.

Digital financial inclusion and SME
Financing Constraints With the rise of digital financial inclusion, more and more scholars have started to pay attention to the relationship between digital financial inclusion and corporate financing constraints. For example, Yi and Zhou (2018) [1] argue that digital financial inclusion can extend the service scope of traditional finance, cover more financing groups and reduce financing costs. Digital financial inclusion can alleviate the financing difficulties of SMEs through the following ways. First, broadening the financing channels of SMEs. Due to the constraints of small scale, insufficient credit and lack of guarantee for loans, SMEs are often rejected by the traditional financial system. Digital financial inclusion, on the other hand, through the use of emerging technologies such as big data and artificial intelligence, and innovation of financial products, can change the way traditional financial services are provided, broaden the audience group of services, provide credit support for the long-tail market to a certain extent, and broaden the financing channels of groups with financial service needs. [2]. [3] As is showed in figure 1, the niche market with long tail is a market with development potential but unmet demand, but the emergence of the digital financial inclusion has brought more sales circulation and storage channels to these markets. Second, reducing information asymmetry between bank and SMEs. With the development of digital finance, financial institutions are able to explore and analyze a large amount of data, obtain more accurate "soft" information, optimize the credit assessment system, and reduce the information asymmetry between the supply and demand sides [4] to alleviate the financing constraints of SMEs. Third, reducing the cost of financing, the development of digital financial inclusion improves the supply limitations of traditional finance relying on physical outlets, reduces the consumption of human and material resources, and can reduce the cost of financial services to a greater extent [5]. In conclusion, digital finance relies on emerging digital technologies to handle large amounts of data with lower risks and costs [6], which improves the allocation efficiency of the traditional financial system, significantly reduces the financing and transaction costs of enterprises, and their financing constraints can be alleviated. Based on the above analysis, research hypothesis H1: The development of digital financial inclusion, will be benefit to alleviate the financing constraints of SMEs

Digital financial inclusion, Financial Mismatch and SME Financing Constraints
The concept of financial mismatch originates from the theory of resource allocation efficiency, which refers to the fact that financial resources are in an inefficient allocation state and are not allocated according to the principle of efficiency size. Domestic and foreign scholars generally agree that financial mismatches significantly exist in China's traditional financial markets and argue that financial mismatches exacerbate the level of financing constraints faced by firms, and Dollar (2007) [6] show that the close ties between SOEs and banks in China are one of the major reasons why credit policies are more favorable to SOEs. However, SOEs are still inferior to private firms in terms of average return on capital, thus presenting a significant distortion of financial resources. The creation of digital financial inclusion has brought new opportunities to alleviate the mismatch of financial resources in China. On the one hand, the development of digital financial inclusion affects the banking industry, intensifies the competition in the banking industry, [15]overturns the traditional credit assessment model and credit pricing mechanism, and forces the financial sector to transform and upgrade; at the same time, digital financial inclusion has the characteristics of informatization, wide coverage, and low cost, which can accurately portray user profiles, finely develop risk pricing, and centralize business processes, enhance banks' capital allocation efficiency and risk management capacity, help break the boundary constraints of traditional finance [8], correct the deficiencies of the traditional financial market, reduce the degree of financial mismatch of enterprises, and thus alleviate the financing constraints of SMEs. On the other hand, digital financial inclusion can take advantage of fast data processing, low service costs, and low venue restrictions, which are different from traditional finance, to reduce the space for rent-seeking, allowing financial resources to be rationally allocated and reducing the degree of financial mismatch. SMEs are thus exposed to more access to financing and lower financing costs, which leads to the alleviation of financing constraints [9]. Based on the above analysis, research hypothesis H2 is proposed: digital financial inclusion alleviates the financing constraints of SMEs by reducing the degree of corporate financial mismatch.

Sample selection and data sources
This paper selects 843 SMEs in the Shenzhen Stock Exchange from 2012-2021 as the sample for the study. The data sources are as follows: (1) micro data of SMEs from CSMAR and WIND database; (2) city-level data from China Statistical Yearbook. (3) The digital financial inclusion data were obtained from the Digital Financial Inclusion Index of Peking University from 2011-2021. The raw data are screened and processed in the following ways: (1) Excluding financial companies; excluding ST companies. (2) Excluding companies with missing financial data. (3) In order to eliminate the effect of extreme values, the sample continuous variable data were processed by two-way tailing (Winsorize). The data processing was done using Stata 15.0 software. After these operations, the final balanced panel data of 843 SMEs with a total of 8430 observations from 2012-2021 were obtained.

Variable Selection
(1) Explanatory variables Financing constraints ( , , ); A large number of scholars from domestic and international have conducted research on the specific quantification methods of financing constraints, and there are many differences in different literatures. In this paper, the KZ index is calculated by Kaplan and Zingales (1997) [8]and used as a proxy for the financing constraint of the firm.
(2) Core explanatory variables The digital financial inclusion index ( t Index ); from the Peking University Digital Financial Inclusion Index, which fully reflects the development of digital financial inclusion in each region of China [10], and is logarithmically treated in order to eliminate the variance between data.
(3) Mechanism transmission variables Financial mismatch ( , , ) 。 Referring to the measurement by Shao(2010) [11], the financial mismatch is measured by the deviation of each firm's annual cost of capital use from the average cost of capital use of the industry in which the firm is located in that year. (4) Control variables Referring to existing studies, in this paper, the following control variables are added; firm-level control variables include: firm size (Lnsize), capital intensity (Capital), dual job (Dual), long-term debt (Debt), gearing (Lev), and firm ownership (Ownership), etc. At the regional level, this paper controls for the regional level of economic development (Pgdp) and the regional level of financial development (Fino). Table 1 gives the names of the variables and their measures in the empirical model of this paper.

Model Setting
To test the previous hypothesis 1 and hypothesis 2, a twoway fixed effects model is constructed in this paper, and the specific expressions are shown below.

Benchmark analysis
Columns (1) and (2) of Table 3 show the results of the baseline regressions in this paper, with column (1) controlling only for time and industry fixed effects and column 2 adding firm and region control variables to the model in column 1. The regressions find that the coefficients of the core explanatory variables ( are -3.795 and -2.367 respectively, both of which are significantly negative. Although the coefficients of the control variables are lowered after adding them, the results are still highly significant. The above regression results indicate that digital financial inclusion can play a significant role in alleviating the financing constraints of SMEs. The previous hypothesis 1 can be verified. In addition, considering that digital financial inclusion can correct the shortcomings of traditional financial development, this paper uses dynamic superposition effects to study the timeliness of the role of digital financial inclusion by referring to Tang(2020) [17] The regression results are shown in columns (3) to (5) of Table  3: the coefficient of digital financial inclusion lag 1 (

Robustness tests
In order to ensure the robustness of the baseline regression results, the following robustness tests are done in this paper: (1) Adjustment of measurement method:Referring to Moser and Voena (2012) [13], using the advanced joint fixed effect of "year × industry" and adding the city cluster effect (cluster) to correct the standard error. (2) replace the measurement indicators of the core explanatory variables, and re-run the regression analysis by replacing the statistical caliber and selection of the explanatory variables, using the digital financial inclusion index of the province where the enterprise is located with the selected lagged period of digital financial inclusion as a proxy variable. (3) Replacing the explanatory variables with their measures and using the SA index as a proxy for the explanatory variable (financing constraint), as in Hadlock (2010) [12] , and re-running the regression analysis. The regression analysis was re-run. The three robustness results are listed in

Mechanism Analysis
Through the previous benchmark regressions, it can be confirmed that the development of digital financial inclusion can greatly alleviate the level of financing constraints of SMEs. Further consideration: the development of digital financial inclusion also affects the external market environment faced by SMEs, and can improve the financing environment of SMEs by accelerating the change of traditional financial markets and increasing the efficiency of resource allocation, so the indirect utility generated by digital financial inclusion will be analyzed below.
To enhance the robustness of the results, different measures of model (2) are used in columns (1) to (3) of Table 5, respectively, and the regression results show that the regression coefficients of digital financial inclusion ( t Index ) are all significantly negative, indicating that digital financial inclusion can significantly improve the financial mismatch of enterprises. It shows that digital financial inclusion, on the one hand, can improve the efficiency of traditional financial institutions by intensifying the competition in the banking industry and so on, and on the other hand, it can take advantage of fast data processing, low service costs and low venue restrictions different from traditional finance to reduce the space for rent seeking, so that financial resources can be rationally allocated and the degree of financial mismatch can be reduced. Some scholars have confirmed that the existence of financial mismatch will aggravate the financing constraints of SMEs, mainly through the following ways: first, financial mismatch will aggravate the "crowding out effect" in the credit market. According to Lu (2008) [16] , state-owned enterprises can obtain financial resources at a lower price. In addition, domestic and foreign scholars generally believe that the problem of "scale discrimination" in China's financial market is still significant, and it is easier for large enterprises to obtain credit funds from financial institutions compared with SMEs, thus generating a scale "crowding out effect" [15] [16] . Second, financial mismatch will promote rent-seeking activities. The government's right to manage and control resources induces rent-seeking behavior to a certain extent, which makes it much more costly and difficult for enterprises to obtain funds. Ge (2018) [14] found that when the financing needs of non-state enterprises cannot be met from the traditional financial market, they can only obtain credit resources by investing a lot of extra rent-seeking costs, and the enterprises face a more severe financing environment. Therefore, it can be verified that digital financial inclusion can have an indirect effect, by alleviating the level of financial mismatch faced by enterprises, and thus alleviating the level of financing constraints of SMEs. To further test whether the indirect effect is significant, this paper conducts a bootstrap test for t Index 、 , , and , , are bootstrap tested, and the results are shown in Table 6, both direct and indirect effects do not include 0, indicating that the indirect effect test is passed, and the degree of financial mismatch of firms plays a significant mediating effect in the mechanism. In summary, hypothesis 2 is verified.

Expandability analysis ⑴ Analysis of the heterogeneity of different ownership properties
The main manifestations of financial mismatch in China are "ownership mismatch" and "size mismatch", so the conclusions of hypothesis 1 and hypothesis 2 may have heterogeneity of property rights. According to the nature of property rights, SMEs are divided into two groups: state-owned and non-state-owned, and the regression results are shown in Table 6.
The regression results are shown in Table 7. In the group of state-owned SMEs, the coefficients of the core explanatory variables ( t Index ) in model (1) and (2) are negative, but both are not significant, indicating that the development of digital financial inclusion does not have a significant impact on the level of financing constraints and the degree of financial mismatch of state-owned SMEs, while in the group of non-state-owned SMEs, the coefficients of the core explanatory variables ( t Index ) in model (1) and (2) are -2.453 and -0.692 respectively, which are significantly negative. The results of the bootstrap test for the non-SOE group are shown in Table  7, which indicates that the mediating effect of financial mismatch is significant in the non-SOE group. The resource endowment, policy making and economic development level of each region in China vary greatly, and the conclusions of the previous hypothesis 1 and hypothesis 2 may differ between different regions. In this paper, we divide the eastern and mid-western regions according to the location of enterprises, and test the two groups according to the previous model. The regression results in Table 8 show that the coefficients of the core explanatory variables ( t Index ) in model (1) and (2) are -2.404 and -1.474 respectively in the eastern region group, and both are significantly negative, indicating that the previous hypothesis 1 still holds in the eastern region and that digital financial inclusion can alleviate the financial mismatch faced by enterprises. The coefficients of t Index are -1.839 and -0.867 respectively, which are also significantly negative. This also demonstrates that the previous hypotheses 1 and 2 are still valid in the midwest region, but the absolute values of the coefficients of t Index have decreased compared to those in the eastern region, indicating that the mitigating effect of digital financial inclusion, is stronger in the eastern region, which indicates that the mediating effect of financial mismatch is significant in the western group.

Conclusions and recommendations
Using 843 SMEs listed on the Shenzhen Stock Exchange in China from 2012-2021 as the research sample, this paper examines the relationship between digital financial inclusion and SMEs' financing constraints using a fixed effects model, and explores its mechanism of action to test what role financial mismatch plays in this mechanism. Finally, a heterogeneity test is conducted. The findings of the study are as follows: First, digital financial inclusion can play a significant role in alleviating the financing constraints of SMEs, and this alleviation effect has a time-dynamic effect. Second, the indirect effect of digital financial inclusion affects the "financial mismatch" faced by enterprises, and the mechanism of " digital financial inclusion -financial mismatch -financing constraint" exists. Third, digital financial inclusion mainly alleviates the financing constraints of non-state-owned SMEs, and the indirect effect of financial mismatch is only significant in non-state-owned enterprises. The alleviating effect of digital financial inclusion is greater in the eastern region. Based on the above research findings, this paper puts forward relevant policy recommendations as follows. First, to improve the long-term mechanism of digital financial inclusion to alleviate the financing constraints of enterprises, especially in the stage of China's economic development toward high quality, China should follow the rapid trend of digital economy development and give certain policy support to the digital economy. Therefore, we should continue to promote and develop digital financial inclusion, attach importance to its role in alleviating the financing difficulties of SMEs, and promote digital financial inclusion to better serve the real economy. Second, all kinds of financial institutions should pay attention to the use of digital technology, accelerate digital transformation, expand the service groups of finance, so that more "tail groups" can receive more comprehensive and fast financial services, and the development of digital financial inclusion can, to a certain extent, reduce the degree of financial mismatch and thus alleviate the financing difficulties of SMEs, and help promote the healthy development of the real economy.