Digital Finance and Corporate Cash-Holding Strategy: Organizational Heterogeneity and Strategic Transmission Channels

Abstract

This study examined the impact of digital finance on corporate cash holdings and the influence of organizational structure and corporate strategy, using the example of A-share-listed companies in China from 2011 to 2018. The results showed that digital finance significantly reduced the level of cash holdings of listed companies; compared with Group Holding Company Limited, the impact of digital finance on the cash holdings of independent enterprises was more pronounced; and digital finance reduced the degree of strategic radicalization of listed companies, then reduced the level of cash holdings of listed companies. The reliability of the empirical results was validated using the dynamic panel model, instrumental variable analysis, and other sensitivity tests. The findings of the study have implications for the guidance of digital finance development and dynamic adjustment of corporate cash holdings.

1. Introduction

High levels of cash holdings by listed companies in China have become a common occurrence in recent years, particularly since the 2008 financial crisis. The trend of “cash is king” has become more apparent and has attracted the attention of all sectors of society. Restricted external financing [1], uncertainty in the governance environment [2], and self-interest motivation for investment [3] are fundamental reasons for listed companies holding large amounts of cash. Improving corporate financing environments and strengthening external governance are critical channels for optimizing corporate capital allocation decisions and revitalizing corporate capital stock. China’s digital financial ecosystem is at the development stage as emerging technologies, such as 5G, artificial intelligence, big data, cloud computing, and blockchain, continue to empower the transformation and upgrading of the traditional financial industry [4], establishing a digital financial industry characterized by Internet credit and mobile payment which has overturned the intelligence-gathering method and information-processing capability of the traditional financial system [5,6]. As a result, it is worth investigating whether decision making by listed companies with regard to the level of their cash holdings has changed because of the advancement of digital finance.

From the existing literature, more studies analyzed factors influencing the financial environment of cash holdings and examined the influence of financial development [7], financial intermediation [8], the financial crisis [9], and financial ecology [10]. Studies on digital finance and corporate cash holdings are still relatively rare. Some scholars believe that the development of digital finance has improved the quality and efficiency of financial services, including faster loans and lower financing costs [11], making companies rely on external funding to increase their cash holdings [12,13]. The study is based on a sample of specific types of firms, which affects the generalizability of the findings, and focuses on theoretical aspects with insufficient empirical evidence. Some scholars point out the impact of corporate structure on the cash holding status of reporting entities: the value of cash holdings is significantly higher, and when the board of directors is larger, the proportion of independent directors is higher; the chairman of the board of directors also holds the position of CEO [14]. However, the effect of different organizational forms on the firm’s cash holdings was not analyzed. In market operations, hierarchical organizations and conglomerates all exist to compete with the high transaction costs of other organizations [15], and because digital finance reduces the transaction costs of capital markets, it may impact the relative strengths of organizational structures [16]. Furthermore, corporate strategy, which coordinates the external environment and internal resources, is influenced by both the external industry environment and the macroenvironment, and it acts as a significant guide for the micropolicies of the firm. The interaction between organizational heterogeneity and corporate strategy has not yet been studied in relation to the microeffects of digital finance, but corporate strategy may be a significant channel via which digital finance influences the financial behavior of a firm [17].

The above studies provide important theoretical support and methodological insights for this paper. In this regard, our study systematically examined the impact of digital finance on corporate cash holdings, identified the roles of organizational structure and corporate strategy in the impact mechanism, and empirically tested the theoretical logic using a sample of A-share-listed companies in China from 2011 to 2018. Compared with other research results, the theoretical contribution of this paper is mainly reflected in the following two points. First, it demonstrates that digital finance has a significant impact on corporate cash-holding levels, furthering research into cash-holding decisions and liquidity management. Jimin noted the impact of improvements in the firm’s external institutional environment, such as payment methods in the financial sector, on the firm’s cash holdings [18]. However, this impact is limited to the degree of marketization of financial development and has not yet focused on the context of digital upgrading of the financial industry for corporate cash-holding decisions. The research in this paper provides a theoretical basis for corporate cash-holding decisions in the context of the digital economy. Second, this paper investigates the role of organizational structure and corporate strategy in the impact of digital finance on corporate cash holdings, as well as the heterogeneity of organizational structure and the strategic transmission channels of the microeffects of digital finance. In existing studies on the impact of digital finance on the microbehavior of firms, most scholars have focused on the impact of digital finance development on financing constraints relating to the promotion of entrepreneurship [19], technological innovation [12], investment efficiency [20], green development [21], and investment efficiency and decision making [22], but have neglected the role of organizational and strategic factors in the microeffects of digital finance. This study explores the organizational and strategic factors relating to digital finance microeffects and provides an empirical analysis of the mechanisms with regard to the role of cash holdings, thus extending the research related to the economic consequences of digital finance.

The rest of the paper is organized as follows: Part 2 sets out the theoretical analysis and research hypotheses; Part 3 provides the sample and data sources, empirical model, and variable measures of the study; Part 4 includes the descriptive statistics and empirical findings; Part 5 covers additional analysis and robustness tests; Part 6 is the discussion; and Part 7 presents the conclusions and implications of the study.

2. Theoretical Analysis and Research Hypothesis

2.1. Digital Finance and Cash-Holding Strategy

According to the principal-agent theory, the high-scale cash-holding behaviors manifest the management, and the majority stockholders expropriate the interests of the corporation, and the minority stockholder [23] firms tend to hold high amounts of cash compared to fixed assets in order to satisfy the self-interested needs of management for day-to-day spending and overinvestment, the need for major shareholders to access capital [24], etc.

The pecking order theory explains corporate capital structures by differentiating internal funds from external funds according to macroeconomic conditions, asymmetric information and conditional on financial constraints [25]. It suggests that, as a result of information asymmetry, firms prefer internal financing channels to external debt financing and external equity financing because of the low transaction costs for funding investment projects through cash holdings. Constrained financing in the traditional financial environment produces the tendency for firms to hold more cash to prevent investment opportunities and avoid liquidity risk [26,27].

Trade-off theory argues that the cash-holding behavior of firms stems from the imperfect development of financial markets. There are both benefits and costs associated with holding cash, and managers weigh these benefits and costs to dynamically adjust the target cash holdings to maximize corporate value [28]. The existing studies generally identify three motives for cash holding: transactional, precautionary, and profit-related [29]. According to the motivation for corporate cash holding, the theoretical research results suggest that development and change in the financial markets will change the external financing environment of enterprises, reduce the financing constraints caused by information asymmetry, and finally influence enterprises to adjust their own cash-holding levels and alleviate agency costs, provide more efficient financial services than the traditional financial services. The high costs of manual due diligence, low efficiency of project evaluation, and long loan decision-making chains in China’s traditional bank-based financial system have resulted in problems such as long financing approval cycles and high financing costs. When financial markets change, such as carbon risk, which impacts carbon emissions on the cost of debt financing for non-financial European industries [30]. companies adjust their cash-holding levels. As an important supplement to the traditional financial service system, digital finance, with the help of big data, cloud computing and other technologies, improves the structure of the traditional financial market and its systems, facilitates the daily transactions of corporate cash holdings in a more accurate, efficient, and convenient way, reduces financial risks, seizes investment opportunities to deliver benefits, and reduces management and opportunity costs [31,32]. Digital transformation of the financial industry improves the financing environment, reduces information asymmetry between subjects, and narrows the gap between internal and external financing costs, and, as a result, firms adjust their cash-holding strategies. In summary, the development of digital finance may affect the cash-holding decisions of firms in the following ways.

Firstly, digital finance can reveal more dimensions of corporate information, lower the risk and cost of financial intermediary services, increase the supply of diverse funds in the market, close the cost gap between exogenous and endogenous financing, and enable enterprises to reduce their cash-holding levels. (1) To access better corporate customer group information, financial institutions can first use big data, artificial intelligence, and other financial technologies for more effective use of information and improved cost efficiency for banks [33]. Such information includes, for example, tax information, product transaction information, raw material procurement information, other operational information, industry and industry chain information, and other clustered information, as well as soft information, such as the individual information relating to the actual controller of the enterprise, social information, and so on. Multidimensional information can alleviate information asymmetry between borrowers and lenders, lowering the risk premium associated with exogenous financing for enterprise loans. (2) Digital finance has reduced the risk of commercial banks [34], eliminated the information monopoly of commercial banks, increased competition between commercial banks [35], and reduced the market power premium of the traditional banking sector, and therefore enables companies to access more favorable financial services. Furthermore, the growing use of technologies such as cloud computing, artificial intelligence, and blockchain has increased the efficiency of the operations and maintenance of financial institutions while decreasing the marginal cost of capital financing, i.e., the digital finance environment has reduced the cost of exogenous financing for enterprises [36]. (3) Digital finance increases investors’ access to financial markets, broadens the investment channels for capital suppliers, and expands the supply of capital in the debt market [37]. In the secondary market, the rapid development of information technology has changed the dynamics of financial markets. Algorithmic trading has increased the operation efficiency of a stock market and provides liquidity for stock market participants contributing to friction-free transactions, and the overall market liquidity has become more abundant [38]. Digital finance lowers the costs for firms when marketing their projects to potential investors [39], a new type of supply chain financial platform solves the problem of nontrust among the participants in the supply chain, improves investors’ efficiency of the capital flow and information flow, and reduces costs [40]. It also alleviates the problem of expensive financing caused by an imbalance in capital supply and demand, as well as, to some extent, the structural financing constraints caused by an imbalance in the development of direct and indirect financing.

Second, by continuously innovating financial products and services, digital finance lowers the threshold of financial services and increases the efficiency of financial services while also broadening access to finance, allowing firms to access exogenous finance more quickly, and reducing the incentive to hold cash preventively. (1) Digital finance can improve the efficiency of financial services [41]. Technologies such as neural networks, expert systems, support vector machines, and hybrid intelligence can significantly improve the speed of lending decisions [42], and digital finance also simplifies the supply–demand transaction process and speeds up the credit approval process [43]. Ant Group’s credit business, for example, has reduced traditional bank loan review and disbursement times from months to three seconds. (2) Through financial product innovation, digital finance can provide enterprises with diversified financing options, broaden their financing channels, provide personalized (customized) financial products, and respond quickly to their financing needs [44]. More timely availability of exogenous financing can help companies revitalize their stock resources through financial management activities such as short-term investments, leading to lower levels of cash holdings.

Finally, digital finance can improve credit regulation in the financial sector and alleviate information asymmetry between lenders and borrowers and weaken management’s self-interested incentive to retain cash flow, in order to reduce the level of corporate cash holdings. The use of technologies such as big data, cloud computing, artificial intelligence, and blockchain technology can help financial institutions obtain customer information more quickly, efficiently, and accurately [45], and manage behavior that negatively affects company value through day-to-day expenditure and overinvestment, such as “adverse selection” and “moral hazard” manipulation of demand for the illegal appropriation of funds [46]. It further establishes effective monitoring of the use of internal and external sources of funds, thus modifying to some extent the excessive cash holdings and other behaviors induced by the self-interested motives of managers. Therefore, improved debtor governance and debtor governance in the digital financial environment weakens the self-interest incentive of management to hold corporate cash.

In summary, according to the Pecking order theory, digital finance improves financing efficiency and ease of access to external corporate funds by expanding the delivery methods and breadth and depth of traditional financial services. According to the principal–agent theory, the development of digital finance can increase the acquisition of corporate information, alleviate the information asymmetry between borrowers and lenders, and reduce the self-interested motivation of management to hold cash. According to trade-off theory, firms may tend to reduce cash holdings and use them for other production, investment, and other activities to maximize value. Accordingly, this paper proposes Hypothesis 1:

Hypothesis 1.

An improved digital financial environment will drive companies to reduce capital holdings.

2.2. Digital Finance, Group Control, and Cash Holdings

Independent firms play an important role in China’s economic development. The structural imbalance of the traditional financial industry, the imperfect financial system, and the prudence of risk assessment have led to a lack of equity in the availability of financial resources and afforded a clear advantage to enterprise groups over independent firms in terms of external noneffective capital market transaction costs. Khanna and Palepu argue that business groups can be seen as an organizational response to an imperfect institutional environment within emerging economies [47]. Because reducing transaction costs in capital markets is an important function of digital financial development [44], the relative advantage of the enterprise group as an organizational form may be weakened as digital finance develops.

Compared to group-controlled firms, independent firms tend to disclose less public information [48], choose to hold higher levels of cash for precautionary motives such as lower ownership structure [49], constrained external opportunities, and instability [50]. This is also reflected in the fact that independent enterprises may face greater financing constraints than group-controlled enterprises because of factors such as insufficient collateral and inadequate financial systems, and therefore choose to hold higher levels of cash as a precaution and actively seek alternative financing to meet their needs and cope with internal and external uncertainties. Efficient allocation of decision-making power at the enterprise level seeks to minimize the sum of information and agency costs, and the incentive effect of decentralization is more pronounced in group-controlled enterprises [51], whereas the operational efficiency of independent enterprises may be limited by the disadvantages of their organizational structure. Enterprise groups can optimize the allocation of capital through the cash flow complementarity mechanism of internal capital markets [52], which has the potential to reduce precautionary cash holdings while also avoiding underinvestment resulting from financing constraints. There is a clear risk-sharing effect of the organizational form [53]. The emergence and development of digital finance provides new ways in which independent enterprises can resolve the financing dilemma. Digital finance complements and improves the traditional financial system, significantly lowers the threshold of access to financial services, and provides new financing channels such as network lending, equity crowdfunding, and Internet funds which broaden the available funding sources, reduce information asymmetry, and enable independent enterprises to obtain financing at lower cost and in a more diversified manner. In addition, the advantages of digital technology in terms of innovative financial products, service processes, and models, facilitation of risk management and credit, simplification of loan procedures, and alleviation of the organizational defects of independent enterprises, improve financing efficiency and reduce transaction costs for independent enterprises. Diversified, low-cost, and efficient sources of financing significantly increase the availability of external funds, which can reduce the incentive of independent firms to hold cash. Accordingly, Hypothesis 2 is proposed:

Hypothesis 2.

The impact of digital finance on the level of cash holdings is more significant for independent firms than for group-controlled firms.

2.3. Digital Finance, Strategic Positioning, and Cash Holdings

The existing literature usually studies the impact of controlling shareholders’ investment choices [54] and management’s shareholding on cash distribution decisions from the perspective of corporate governance [55], but the actual cash-holding strategy is a financial consequence of a company’s corporate operations. Corporate governance indirectly affects cash-holding decisions through the firm’s business behavior, where there is business behavior, there is corporate strategy, and companies implementing different types of strategy show significant differences in their business behavior and forms of control allocation [56].

As an important artefact coordinating the external environment and internal resources, corporate strategy is influenced by both the external industrial environment and the macroenvironment, as well as provides important guidance for corporate micropolicies. The elements of a company’s strategy are generally categorized as defensive, analytical, and offensive [57]. This classification covers the full range of market positioning approaches according to the degree of aggressiveness. Offensive strategies have the highest degree of aggressiveness, followed by analytical strategies and defensive strategies, which have the lowest degree of aggressiveness. Defensive strategies emphasize product quality and price, usually using cost leadership or market focus business models. Offensive strategies emphasize innovation and environmental adaptability, generally using market development or product development business models. R&D investment, company operations, and internal management systems are different from other strategy types. The difference in the degree of strategic aggressiveness contributes to different internal and external environments, business risks, and agency costs. The development of digital finance has profoundly changed the external financing and information environments of companies [58], and corporate strategies tend to select the level of strategic aggressiveness appropriate to these external environments in a power-variable manner and their precautionary cash holdings are reduced in line with operational risk. Additionally, corporate strategy is an important starting point for business and financial decisions, and financial management policies are significantly different when the corporate strategy is heterogeneous because of the difference in the resulting agency costs. Holding excessive cash leads to intensification of corporate agency conflicts and lower value levels. The development of digital finance alleviates the corporate agency cost problem by reducing external financing pressures, which, in turn, influences the cash-holding strategy.

China’s listed companies have a high threshold for financing through issuing shares, and corporate financing channels are more dependent on debt financing [59]. Digital finance enables financial institutions to identify corporate operating risks in a more comprehensive and timely manner, and an aggressive strategy may cause a mismatch between loan returns and corporate capital use risks, thus preventing listed companies from adopting a more robust strategic position. Aggressive strategies tend to increase corporate financing needs, and listed companies with aggressive strategies tend to maintain higher liquidity in order to exploit future investment opportunities, i.e., aggressive strategies make listed companies hold larger amounts of cash. Based on the above analysis, the development of digital finance influences the more aggressive corporate strategies and reduces the precautionary and agency motivations for holding cash. Therefore, this paper proposes Hypothesis 3:

Hypothesis 3.

Digital finance reduces the strategic aggressiveness of listed companies, which, in turn, reduces their cash holdings.

3. Study Design

3.1. Sample and Data

The initial research sample of this paper was A-share-listed companies in Shanghai and Shenzhen from 2011 to 2018, and some abnormal samples were excluded according to the following criteria: companies in the financial sector, listed companies under delisting risk warnings (ST), listed companies with missing key data, and listed companies with an asset-to-liability ratio greater than 1. After the exclusions, 15,982 observations were obtained. The data at the listed company level were obtained from the CRSMAR database, and the digital finance index data were obtained from the China Digital Inclusive Finance Report published by Peking University Digital Finance Research Center (https://idf.pku.edu.cn/), which compiled the “Digital Inclusive Finance Index” covering 2011–2021. The “Digital Inclusive Finance Index”, which covers all provinces (cities and counties) in mainland China and summarizes the development status of digital inclusive finance in the various regions [60], provided the data for this empirical study. To exclude possible perturbations of the empirical results by outliers, Winsorize was applied to all continuous variables at the 1% and 99% quartiles.

3.2. Definition of Variables

3.2.1. Digital Finance (Index) Metrics

Drawing on the financial inclusion indicator system constructed at the 2013 G20 Summit in St. Petersburg, Russia [61], and the research results of Guo Feng et al. [62], this study selected the provincial-level China Digital Inclusive Finance Index published by the Digital Finance Research Centre of Peking University as a proxy variable for the level of digital finance development in the region in which the listed companies are located. The China Digital Inclusive Finance Index includes indicators in the following three dimensions: breadth of digital financial coverage (Breadth, characterized by digital financial account coverage), depth of digital financial use (Depth, characterized by the frequency or number of digital financial services such as payments and credit), and digitization of inclusive finance (Digi, characterized by the cost and convenience of digital financial services). In this study, we explored the impact of the secondary dimensions of digital financial development on the cash-holding strategies of listed companies.

3.2.2. Definition of Group Control (Group)

Drawing on Khanna et al. [63], and He et al. [53], the criteria for determining whether a listed company was affiliated with an enterprise group were established as follows: if the listed company’s actual controller was a group company or a company that acts as a group company, the value was 1; otherwise, it was 0. The listed company was considered an independent enterprise if it had no actual controller or if the actual controller was the State-owned Assets Supervision and Administration Commission, the State-owned Capital Investment and Operation Company, or any other government agency at any level of government.

3.2.3. Measurement of Cash Holdings (Cash)

Drawing on the measures used by Opler et al. [64] and Han et al. [65], the ratio of cash and cash equivalents to total assets was used as an indicator of the size of cash holdings. In addition, in the robustness test section, we re-examined the impact of digital finance on cash holdings by using the ratio of cash and cash equivalents to net assets and the ratio of cash-to-cash equivalents to total assets adjusted for industry averages as proxy indicators of the size of the cash holdings of the firms.

3.2.4. Degree of Strategic Aggressiveness (Stra)

Drawing on the study by Bentley et al. [66], R&D investment (R&D expenditure as a percentage of sales revenue over the past five years), growth capacity (historical growth rate of sales revenue over the past five years), employee turnover (number of employees as a percentage of sales revenue over the past five years), and company operations (selling and administrative expenses as a percentage of sales revenue over the past five years) were used to construct the index of strategic aggressiveness of member-listed companies (fixed assets as a percentage of total assets over the past five years). The index of the degree of strategic aggressiveness of member-listed companies was constructed in six aspects, including fixed assets as a proportion of total assets over the past five years. The above variables were divided into five equal groups by “year–industry”, with the first five variables assigned values from 0 to 4 from small to large, and the last variable assigned values from 0 to 4 from large to small and then summed up. The index of the aggressiveness of each company’s strategy was obtained for each year, with a larger index implying a more aggressive strategy.

3.2.5. Other Control Variables

To mitigate the effect of omitted variables on causal inference, drawing on Bates et al. [67], this study controlled for influencing factors related to the cash holdings of listed companies. These were: listed company size (Size, the natural logarithm of a listed company’s year-end total assets), company age (Age, the natural logarithm of a company’s years of listing plus one), profitability (ROA, return on total assets), solvency (Lev, gearing ratio), growth (Growth, growth rate of operating income), and net operating cash flow (FCF, the ratio of net operating cash flow to year-end (Big4, annual report audited by Big 4). Corporate governance variables were equity concentration (Concen, shareholding of the largest shareholder), dual role (Dum, chairman and general manager are the same person), board size (Direct, number of directors), and board independence (Ratio of the number of independent directors to the number of directors). Detailed descriptive statistics are shown in

Table 1. Descriptive statistics.

Variables	Sample Size	Average Value	Standard Deviation	Minimum Value	Lower Quartile	Median	Upper Quartile	Maximum Value
Cash	15,982	0.174	0.120	0.0171	0.0902	0.142	0.226	0.598
Index	15,982	220.9	83.68	29.74	165.6	228.8	282.2	377.7
Breadth	15,982	201.9	81.48	15.33	150.0	209.8	267.4	353.9
Depth	15,982	223.5	86.28	36.28	161.2	219.9	282.9	400.4
Digi	15,982	279.3	110.2	15.71	230.7	309.3	373.8	440.3
Group	15,982	0.450	0.497	0	0	0	1	1
Stra	9533	17.96	3.793	6	15	18	21	29
Cash	15,982	0.174	0.120	0.0171	0.0902	0.142	0.226	0.598
Size	15,982	22.28	1.271	19.95	21.37	22.09	23.00	26.16
Age	15,982	2.260	0.674	0.693	1.792	2.398	2.890	3.258
ROA	15,982	0.0371	0.0515	−0.181	0.0134	0.0338	0.0616	0.186
Lev	15,982	0.438	0.206	0.0551	0.273	0.434	0.595	0.883
Growth	15,982	0.189	0.432	−0.529	−0.0133	0.113	0.276	2.796
FCF	15,982	0.0432	0.0725	−0.496	0.00530	0.0420	0.0826	0.876
Big4	15,982	0.0620	0.241	0	0	0	0	1
Concen	15,982	0.352	0.148	0.0942	0.234	0.333	0.452	0.750
SOE	15,982	0.427	0.495	0	0	0	1	1
Dual	15,982	0.241	0.427	0	0	0	0	1
Direct	15,982	8.691	1.720	5	7	9	9	15
Ind	15,982	0.374	0.0533	0.333	0.333	0.333	0.429	0.571

.

3.3. Model Design

Firstly, to test the impact of digital finance on the cash holdings of firms, the following model was constructed:

$$Cas{h}_{it}={\alpha }_0+{\alpha }_{1}Inde{x}_{it}+{\displaystyle \sum _{j=2}^{11}{\alpha }_j}Control{s}_{it}+YEAR+INDUSTRY+\epsilon$$

(1)

Cash in the model is the measure of cash holdings and Index is the digital financial inclusion index of the province in which the listed company is located. The control variables in the model (1) were as described above, and the empirical tests controlled for industry and time effects by default, using OLS regression estimates with robust standard errors.

Secondly, we constructed model (2) to test the moderating effect of whether a firm is controlled by a group on the impact of digital finance on its cash holdings, which was modeled as follows:

$$\begin{array}{l}Cas{h}_{it}={\alpha }_0+{\alpha }_{1}Inde{x}_{it}+{\alpha }_{2}Group+{\alpha }_{3}Inde{x}_{it}\ast Grou{p}_{it}+{\displaystyle \sum _{j=4}^{11}{\alpha }_j}Control{s}_{it}\\ +YEAR+INDUSTRY+\epsilon \end{array}$$

(2)

Group in the model is a dummy variable for whether the listed company is controlled by a group. The control variables in the model (1) were as described above, and the empirical tests controlled for industry and time effects by default, using OLS regression estimates with robust standard errors. In addition, to further verify the effect of Group control on the main causal relationship in this study, we divided the sample into two groups according to whether they were controlled by a Group and used group regression to test the heterogeneity of the effect of digital finance on the cash holdings of listed companies.

Finally, drawing on the test for mediating effects set out by M J Valente et al. [68], we constructed the following model to test the mediating effect of the degree of strategic aggressiveness in the impact of digital finance on cash holdings of listed companies.

$$\begin{array}{l}Cas{h}_{it}={\alpha }_0+{\alpha }_{1}Str{a}_{it}+{\displaystyle \sum _{j=2}^{11}{\alpha }_j}Control{s}_{it}+YEAR+INDUSTRY+\epsilon \\ Cas{h}_{it}={\alpha }_0+{\alpha }_{1}Inde{x}_{it}+{\alpha }_{2}Str{a}_{it}+{\displaystyle \sum _{j=3}^{11}{\alpha }_j}Control{s}_{it}+YEAR+INDUSTRY+\epsilon \end{array}$$

(3)

Stra is a measure of strategic aggressiveness in the model, and the empirical tests controlled for industry and time effects by default, using OLS regression estimates with robust standard errors.

4. Analysis of the Empirical Results

4.1. Descriptive Statistics

Table 1 presents the results of the descriptive statistics for the main variables. In general agreement with the cash holdings of listed companies in China described in the literature such as Shanyue [14], the cash-holding level (Cash) of listed companies in our study sample averaged 0.174, with a standard deviation of 0.12 and a minimum value of only 0.0171, but a maximum value of 0.598, indicating a large variation in the cash-holding strategies of listed companies. The digital finance index (Index) averaged 220.9 had a minimum value of 29.74 and a maximum value of 377.7, with a standard deviation of 83.68, indicating a high degree of dispersion in the level of digital finance development between different provinces or in the same province in different years, which provided good material for our empirical research, while the differences in the level of cash holdings of firms are greater than the differences in digital finance. After adding group control variables, the dummy variable of whether the company was controlled by a group had a mean value of 0.45, indicating that 45% of A-share-listed companies in China were controlled by a group, 55% of listed companies are independent, further demonstrating the prevalence of independent enterprises as a form of organization in our economic development. The degree of strategic aggressiveness (Stra) averaged 17.96 had a minimum value of 6 and a maximum value of 29, with a high degree of dispersion and variability in the strategic choice of listed companies. Other control variables are shown in Table 1. The results of the descriptive statistics of the remaining control variables are generally consistent with existing studies [58].

4.2. Baseline Regression Results

This study examines the impact of digital finance on corporate cash holdings. Column (1) of

Table 2. Baseline regression results.

	(1)	(2)	(3)	(4)
	Explained Variable: Cash
Index	−0.0001 *** (−12.0995)
Breadth		−0.0001 *** (−11.7216)
Depth			−0.0001 *** (−10.4553)
Digi				−0.0001 *** (−12.6258)
Size	−0.0037 *** (−3.6209)	−0.0038 *** (−3.6896)	−0.0044 *** (−4.3276)	−0.0039 *** (−3.8248)
Age	−0.0048 *** (−2.9598)	−0.0048 *** (−3.0085)	−0.0053 *** (−3.2805)	−0.0043 *** (−2.6648)
ROA	0.2327 *** (11.0499)	0.2349 *** (11.1581)	0.2425 *** (11.5560)	0.2227 *** (10.5220)
Lev	−0.1722 *** (−26.2448)	−0.1720 *** (−26.1886)	−0.1690 *** (−25.9155)	−0.1711 *** (−26.2232)
Growth	−0.0038 * (−1.8030)	−0.0038 * (−1.7951)	−0.0039 * (−1.8097)	−0.0042 ** (−1.9615)
FCF	0.1686 *** (11.5944)	0.1677 *** (11.5330)	0.1660 *** (11.4392)	0.1720 *** (11.8127)
Big4	−0.0059 (−1.6231)	−0.0054 (−1.4923)	−0.0052 (−1.4247)	−0.0076 ** (−2.0878)
Concen	0.0349 *** (5.6949)	0.0354 *** (5.7906)	0.0361 *** (5.8918)	0.0343 *** (5.6162)
Dual	0.0044 ** (2.0349)	0.0043 ** (2.0229)	0.0041 * (1.8886)	0.0038 * (1.7742)
Direct	0.0018 *** (3.1144)	0.0019 *** (3.1801)	0.0021 *** (3.5391)	0.0019 *** (3.2572)
Ind	0.0226 (1.2704)	0.0230 (1.2908)	0.0234 (1.3117)	0.0260 (1.4644)
Intercept	0.2977 *** (14.1250)	0.2957 *** (14.0190)	0.3045 *** (14.4728)	0.3009 *** (14.3082)
Industries	YES	YES	YES	YES
Annual	YES	YES	YES	YES
N	15982	15982	15982	15982
R2	0.2205	0.2201	0.2186	0.2213

Note: T values in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the regression results of digital finance (Index) on the level of cash held by listed companies (Cash), whereas columns (2) to (4) report the regression results of the three breakdowns of digital finance: breadth of digital finance coverage (Breadth), depth of digital finance use (Depth), and digitization of financial inclusion (Digi), respectively, on the level of cash held by listed companies (Cash). The empirical results in column (1) show that the coefficient of digital finance (Index) as an explanatory variable whose regression coefficient is −0.0001, was significantly negative at the 1% significance level, indicating that the development of digital finance in the region reduced the level of cash holdings of listed companies in the region, and the regression results effectively support H1. Columns (2) to (4) examine the relationship between the different dimensions of digital finance development and corporate cash holdings, respectively. The empirical results show that the coefficients of Breadth (breadth of digital finance coverage), Depth (depth of digital finance use), and Digi (digitization of inclusive finance) were all significantly negative at the 1% significance level, indicating that the development of each dimension of digital finance led to a decrease in the level of cash holdings of the listed companies. The breadth of digital financial coverage (Breadth) expands market credit resources, alleviates information asymmetry between borrowers and lenders, and improves the level of supervision; the depth of digital financial (Depth) deeply innovative products and services to broaden corporate financing channels and improve the efficiency of financing services, and enterprises can obtain exogenous financing in a more timely manner, while digitization of inclusive finance (Digi) makes corporate financing more convenient and efficient, reduces corporate financing costs, and weakens the relative advantages of group enterprises through endogenous financing, and the results of regression coefficients of three indicators further verify hypothesis 1 (an improved digital financial environment will drive companies to reduce capital holdings).

4.3. Test Results on the Moderating Effect of Group Control

Different internal governance environment, internal control system and other factors, enterprises choose different levels of cash holdings based on preventive and other motives, the operation of the internal capital market of group enterprises, relative to independent enterprises to ease the financing constraints, strengthen capital financing, group enterprises hold lower levels of cash. The creation and development of digital finance has weakened the relative advantages of this organizational form of group enterprises compared to independent enterprises [69]. Therefore, we analyzed the cash holding levels of group enterprises and independent enterprises, and the regression results are shown in the

Table 3. Results of the heterogeneity analysis distinguishing group control.

	(1) Full sample	(2) Group = 1	(3) Group = 0
	Explained Variable: Cash
Index	−0.0001 *** (−13.0428)	−0.0000 * (−1.7892)	−0.0002 *** (−14.2389)
Group	−0.0330 *** (−6.23)
Index*Group	0.0000 *** (5.4662)
Size	−0.0042 *** (−4.1458)	−0.0055 *** (−4.3420)	−0.0026 * (−1.7580)
Age	−0.0066 *** (−3.9838)	−0.0005 (−0.2141)	−0.0091 *** (−4.2672)
ROA	0.2327 *** (11.0481)	0.3487 *** (11.9074)	0.1531 *** (5.6730)
Lev	−0.1729 *** (−26.3716)	−0.1224 *** (−14.7830)	−0.2062 *** (−25.5212)
Growth	−0.0033 (−1.5677)	0.0022 (0.7804)	−0.0065 ** (−2.2980)
FCF	0.1682 *** (11.5382)	0.1318 *** (7.3479)	0.1962 *** (10.8107)
Big4	−0.0048 (−1.3312)	−0.0015 (−0.3096)	−0.0090 (−1.5328)
Concen	0.0288 *** (4.5909)	0.0233 *** (2.6879)	0.0394 *** (4.4525)
Dual	0.0053 ** (2.4887)	0.0043 (1.2704)	0.0042 (1.5739)
Direct	0.0017 *** (3.0125)	0.0050 *** (6.3286)	−0.0019 ** (−2.0455)
Ind	0.0257 (1.4475)	0.0287 (1.1548)	−0.0168 (−0.6267)
Intercept	0.3151 *** (14.7867)	0.2505 *** (9.1242)	0.3646 *** (11.1713)
Industries	YES	YES	YES
Annual	YES	YES	YES
Suest		79.52 (p = 0.000)
N	15982	7191	8791
R2/Adj_R2	0.2219	0.2146	0.2415

Note: T values in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

.

Table 3 presents the results of the regressions of digital finance (Index) on the level of cash holdings (Cash) of listed companies, distinguished according to whether the listed companies in the sample are group-held or not. Column (1) shows the results of the moderating effect test with the inclusion of the interaction term, and columns (2) and (3) show the results of the group regression of digital finance (Index) on the level of cash held by listed companies (Cash). The regression results in column (1)show that the coefficient result of Digital Finance shows significantly negative at the 1% level, and the coefficient on the interaction term between digital finance and group control (Index*Group) was significantly positive at the 1% level of significance, indicating that digital finance can reduce a firm’s cash holdings and the effect of digital finance on the cash holdings of independently listed companies was more significant and that the development of digital finance weakened the comparative advantage of capital markets within corporate groups. The results of the subgroup test indicated that the coefficient of digital finance (Index) was −0.0000 and −0.0002 in the samples of independently listed companies and group-held listed companies, respectively. The results of the Suest test indicated that there was a significant difference in the coefficient of digital finance (Index), and the results of the group test further verify the conclusion that independently listed companies obtain greater benefit from the development of digital finance than group-controlled companies, thus verifying Hypothesis 2.

4.4. Results of Testing for the Mediating Effect of Strategic Positioning

Companies are influenced by their internal and external governance environment to choose different strategies. Companies have different strategies and different levels of strategic aggressiveness, and the development of digital finance affects the internal and external environment of companies, which affects the level of strategic aggressiveness and thus the level of cash holdings of companies.

Table 4. Results of the test for mediating effects of the degree of strategic aggressiveness.

	(1)	(2)
	Explained Variable: Stra	Explained Variable: Cash
Index	−0.0076 *** (−4.1132)	−0.0000 *** (−3.0784)
Stra		0.0005 * (1.8196)
Size	0.5815 *** (13.0900)	−0.0083 *** (−6.3883)
Age	−1.1683 *** (−12.4854)	0.0176 *** (6.8109)
ROA	8.3655 *** (9.5896)	0.2577 *** (10.6094)
Lev	−1.3968 *** (−5.3457)	−0.1207 *** (−15.4964)
Growth	1.4617 *** (17.1315)	−0.0024 (−0.9220)
FCF	−0.0000 *** (−7.5376)	0.0000 *** (5.9021)
Big4	−0.0506 (−0.2619)	−0.0129 ** (−2.4699)
Concen	−2.7849 *** (−10.1646)	0.0308 *** (4.1239)
Dual	0.6811 *** (6.9347)	0.0019 (0.7389)
Direct	−0.0623 ** (−2.4734)	0.0029 *** (4.2574)
Ind	−0.2173 (−0.2811)	0.0346 (1.5991)
Intercept	10.5594 *** (11.0061)	0.2856 *** (10.2873)
Industries	YES	YES
Annual	YES	YES
N	9533	9533
R2	0.1066	0.1525

Note: T values in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the results of the mediating effects of strategic aggressiveness (Stra) on digital finance (Index) and the level of cash holdings (Cash) of listed companies. Column (1) shows the regression results for the degree of strategic aggressiveness (Stra) on digital finance (Index), and column (2) shows the regression results for the level of cash holdings (Cash) of listed companies on digital finance (Index) and the degree of strategic aggressiveness (Stra). The regression results in column (1) show that the coefficient on digital finance was significantly negative at the 1% significance level, indicating that listed companies in regions with higher levels of digital finance development were less strategically aggressive. The regression results in column (2) show that the coefficient on digital finance (Index) was significantly negative and the coefficient on strategic aggressiveness (Stra) was significantly positive at the 10% significance level, indicating that the degree of strategic aggressiveness plays an important role of transmission path in the mechanism of digital finance affecting the level of cash holdings of listed companies, i.e., digital finance weakens the level of cash holdings of listed companies by reducing the degree of strategic aggressiveness of listed companies, supporting H3.

5. Robustness Tests

Baseline regression results may be confounded by endogeneity issues as well as other issues. Endogeneity issues may disrupt the causal inference of digital finance on microfirm cash holdings, causing the explanatory variable (digital finance) to correlate with the perturbation term, resulting in omission bias in the results of ordinary least squares regressions. Furthermore, issues with variable measures in benchmark regressions, as well as sample selection, may result in empirical results that are not unbiased and consistent estimates. This section mitigates potential endogeneity issues using both dynamic panel models (DPD) and the instrumental variables approach (IV) and replaces variable measures and sample selection to re-examine the impact of digital finance on the cash holdings of listed companies.

5.1. Mitigation of Endogenous Problems

5.1.1. Estimation of Dynamic Panel Models

The current cash-holding level of a listed company is the result of dynamic adjustment of the previous period’s cash-holding level according to changes in the internal and external environment of the company [70]. Therefore, the previous period’s cash-holding level is an important determinant of the cash-holding level of listed companies, and the omission of this variable may lead to omitted variable bias in the estimation results of the ordinary least squares method. In this regard, this section incorporated the first-order lagged term (L.Cash) and the second-order lagged term (L2.Cash) of cash-holding level into the model to estimate the impact of digital finance on listed companies using a dynamic panel model.

Table 5. Empirical results based on dynamic panel models.

	Explained Variable: Cash
L.Cash	0.4621 *** (20.9229)
L2.	0.0221 (1.5548)
Index	−0.0001 *** (−4.0636)
Size	0.0356 *** (4.9622)
Age	−0.0125 (−1.1984)
ROA	−0.0161 (−0.5143)
Lev	−0.1488 *** (−7.4122)
Growth	−0.0035 (−1.2027)
FCF	0.2788 *** (13.8941)
Big4	−0.0177 * (−1.9164)
Concen	−0.0267 (−0.8338)
Dual	0.0030 (0.7758)
Direct	0.0005 (0.3648)
Ind	0.0679 * (1.8173)
Intercept	−0.7678 *** (−4.8664)
Industries	YES
Annual	YES
AR(1) p-value	0.0000
AR(2) p-value	0.1498
Sargan p-value	0.148
N	10263
Wald chi2	941.24 ***

Note: T values in parentheses; * p < 0.1, *** p < 0.01.

reports the results of estimating the dynamic panel model using the systematic GMM. The results show that the coefficient on digital finance (Index) was significantly negative at the 1% level of significance, indicating that the baseline regression results remained robust after using the dynamic panel model to mitigate the endogeneity problem. The results of the autocorrelation test show that AR(1) was significant and AR(2) was not significant at the 10% significance level, indicating that there was first-order but not second-order autocorrelation in the error terms, satisfying the conditions for estimation of the error terms in the systematic GMM. The results of the overidentification test showed a Sargan p value of 0.148, indicating that the instrumental variables selected in the estimated model satisfied the homogeneity condition.

5.1.2. Instrumental Variables Approach

The omission of important variables that are not observable at the listed company or regional level, such as regional culture, may affect both financial development and corporate cash holdings [71,72], and their omission in the control variables may lead to biased OLS estimation results. Therefore, drawing on domestic research in China, we used Internet penetration in each province as the instrumental variables to overcome the endogeneity problem caused by omitted variables. The instrumental variables approach is a model that can theoretically mitigate the endogeneity problem completely [73], and the instrumental variables also satisfy the correlation and homogeneity requirements. Although the popularity of the Internet has driven the development of financial innovations such as local digital finance [74], there was no direct causal link between Internet penetration and microfirm cash-holding levels.

Table 6. Empirical results based on the instrumental variables approach.

	(1)	(2)
	Explained Variable: Index	Explained Variable: Cash
Index		−0.0001 *** (−4.8072)
Net	3.7766 *** (81.46)
Size	15.6881 *** (22.61)	−0.0072 *** (−5.8600)
Age	11.3656 *** (11.50)	−0.0060 *** (−3.6190)
ROA	−120.462 *** (−9.75)	0.2776 *** (13.4353)
Lev	−65.9458 *** (−17.37)	−0.1563 *** (−23.7133)
Growth	3.67104 *** (2.76)	−0.0048 ** (−2.1775)
FCF	0.0000 (−0.82)	0.0000 *** (6.3920)
Big4	−16.7968 *** (−5.56)	−0.0125 ** (−2.4865)
Concen	−29.1545 *** (−7.21)	0.0339 *** (5.0180)
Dual	4.38779 *** (3.23)	0.0059 *** (2.5970)
Direct	−5.1533 *** (−12.67)	0.0029 *** (4.1552)
Ind	−53.6725 *** (−4.34)	0.3507 *** (13.6876)
Intercept	−252.7875 *** (−16.38)	0.3451 *** (13.5123)
Industries	YES	YES
Annual	YES	YES
F-value (Phase 1)	309.12 (0.0000)
Hausman		6.21 (0.0127)
N	12738	12738
Adj-R2	0.4123	0.2010

Note: T values in parentheses; ** p < 0.05, *** p < 0.01.

presents the results of the instrumental variables test, with columns (1) and (2) showing the results of the first- and second-stage regressions of the instrumental variables method for digital finance on cash holdings of listed companies. The instrumental variables test showed that in the first stage regression results the coefficient on Internet penetration (Net) was significantly positive at the 1% level with an F-value significantly greater than 10, indicating that the instrumental variables satisfied the condition of good correlation. The coefficient on Digital Finance (Index) was significantly negative at the 1% level, in agreement with the results of the previous analysis, indicating that the baseline regression results remained robust after use of the instrumental variables approach to mitigate the endogeneity problem.

5.2. Other Robustness Procedures

5.2.1. Substitution of Explanatory Variables

To test the sensitivity of the main findings of this paper to the variable measures, we retested the causal relationship between digital finance and the cash holdings of listed companies using the ratio of cash and cash equivalents to net assets (Cash_Equ), and the ratio of cash and cash equivalents to total assets adjusted for industry averages (Cash_Adj), as alternative indicators of the level of cash holdings (Cash).

Table 7. Empirical results from replacement of the explanatory variables.

	(1)	(2)
	Explained Variable: Cash_Equ	Explained Variable: Cash_Adj
Index	−0.0002 *** (−6.7180)	−0.0000 * (−1.8681)
Size	−0.0107 *** (−3.6746)	−0.0036 *** (−3.5452)
Age	−0.0012 (−0.2763)	−0.0036 ** (−2.2632)
ROA	0.3896 *** (7.7604)	0.2204 *** (10.5134)
Lev	0.4689 *** (20.8214)	−0.1689 *** (−25.7721)
Growth	−0.0052 (−1.0172)	−0.0043 ** (−2.0595)
FCF	0.3081 *** (7.1144)	0.1561 *** (10.7401)
Big4	−0.0152 * (−1.7085)	−0.0071 * (−1.9520)
Concen	0.0708 *** (4.6683)	0.0320 *** (5.2318)
Dual	0.0100 ** (2.1215)	0.0038 * (1.7825)
Direct	0.0051 *** (3.2973)	0.0016 *** (2.7529)
Ind	0.0500 (1.1461)	0.0249 (1.4139)
Intercept	0.2641 *** (4.7217)	0.1368 *** (6.4974)
Industries	YES	YES
Annual	YES	YES
N	15982	15982
R2	0.1709	0.1450

Note: T values in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

reports the regression results of cash holdings on digital finance. The empirical results showed that the coefficient on digital finance (Index) remained significantly negative after replacing the measures of cash holdings, indicating that the empirical findings of the negative relationship between digital finance and the level of cash holdings of listed companies were highly robust.

5.2.2. Exclusion of Sample Selection Perturbations on Baseline Regression Results

Some studies have suggested that year-specific or region-specific samples may interfere with the robustness of the empirical findings [58]. For example, the 2015 stock market crash may have affected both the development of digital finance and cash holdings of listed companies, and the development of digital finance and the level of cash holdings of listed companies in China’s municipalities may also be affected by the heterogeneity of the particular local administrative regions. Ignoring the uniqueness of these two types of samples may lead to biased or inconsistent estimation results. Therefore, we removed the 2015 sample, and the sample of municipalities directly under the Central Government to re-examine the impact of digital finance on listed companies’ cash holdings, and the empirical results are shown in

Table 8. Empirical results of excluding outlier samples.

	(1)	(2)
	Explained Variable: Cash
Index	−0.0001 *** (−11.9590)	−0.0001 *** (−12.1480)
Size	−0.0034 *** (−3.0938)	−0.0035 *** (−2.9934)
Age	−0.0068 *** (−3.9645)	−0.0067 *** (−3.7008)
ROA	0.2231 *** (9.7699)	0.2045 *** (9.0280)
Lev	−0.1794 *** (−25.5179)	−0.1626 *** (−22.4850)
Growth	−0.0045 * (−1.9458)	−0.0030 (−1.2773)
FCF	0.1736 *** (11.2326)	0.1555 *** (9.7947)
Big4	−0.0070 * (−1.7996)	−0.0088 * (−1.9364)
Concen	0.0369 *** (5.6022)	0.0292 *** (4.3106)
Dual	0.0051 ** (2.2389)	0.0071 *** (3.0871)
Direct	0.0020 *** (3.2081)	0.0025 *** (3.7984)
Ind	0.0199 (1.0406)	0.0386 * (1.9136)
Intercept	0.2957 *** (13.0477)	0.2843 *** (11.8052)
Industries	YES	YES
Annual	YES	YES
N	13888	12738
R2	0.2289	0.2067

Note: T values in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01.

. Column (1) shows the regression results with the 2015 sample removed, and column (2) shows the regression results with the municipality sample removed. The digital finance (Index) results were both significantly negative at the 1% significance level, indicating that the baseline regression results of this study remained robust after considering the sample selection issue.

6. Discussion

Factors influencing on cash holdings, scholars have noted and analyzed the impact of financial development [7], financial intermediation [8], financial crisis [9], financial ecology [10], and other factors on the level of corporate cash holding. Traditional finance has many problems, such as difficult and expensive financing. Companies will choose to hold more cash in order to allocate financial resources across time and prevent the objective requirement of internal and external uncertainty [65,75], which is also used to restrain high external financing costs and perform a reservoir function [76,77]. Digital finance, an emerging product of the contemporary digital economy, offers a new research perspective on corporate cash holdings.

First, these results are consistent with previous studies according to the financing constraints. The innovative model triggered by the development of digital finance has broken the boundaries of traditional financial services, realized the efficient allocation of resources [78],and alleviated the financing constraints faced by enterprises, reduced cash holdings of enterprises, and promoted high-quality economic development [79].The development of digital finance can improve the external financing environment [80], reduce information asymmetry, lower financing costs and ease financing constraints [81]. solves corporate financing dilemmas, guides investment behavior, stimulates corporate innovation [12,13]. The development of digital finance reduces the level of cash holdings of public companies, it supplements the scholars’ research on the constraints of digital finance on the financing of small and medium-sized enterprises [82], and the paper’s hypothesis was supported. Further comparison with the previous results, the regression results of our study from three indicators, which are the breadth of digital finance coverage (Breadth), depth of digital finance use (Depth), and digitization of financial inclusion (Digi), more comprehensively confirm that the development of digital finance can reduce the cash holdings of A-share-listed companies and that refute the view of Fuster et al. that digital finance development firms increase cash holdings [11].

Second, when comparing our results with previous ones, it is important to note that it distinguishes between organizational forms of enterprises. Identifying conglomerates and independent companies among Chinese listed companies from the perspective of ultimate controllers, the development of digital finance significantly reduces the level of cash holdings of independent companies in the overall analysis. Scholars have found that the cash flow sensitivity of firms under the influence of governance and non-corporate governance factors is stronger when financing is constrained [83,84,85]. Previous studies have focused on the relative advantages of cash holdings of group firms when external financing is constrained [86], and investigated the impact of political connections and business group affiliation on the cash holdings of firms listed on main Chinese stock exchanges. It also concluded that business groups were negatively related to cash holdings [87]. This study analyzes the centralized management of funds in group companies based on a focus on cash-holding motives [88], overinvestment [89], internal capital market effectiveness [85], etc., on the basis of differences in cash-holding levels from independent companies. We focus on the impact of digital finance development on the level of cash holdings of independent firms, the development of digital finance provides low-cost and diversified funding sources for independent firms, reduces the level of cash held by independent firms, and provides empirical evidence for financial control and cash-holding decisions of listed firms.

Lastly, differential strategy as an important influencing factor for cash holdings, on the one hand, the greater the difference in corporate strategies compared to conventional industry strategies, the greater the difference in performance and the greater the degree of cash flow volatility [90]. On the other hand, the degree of strategic deviation can exacerbate information asymmetry [91], information asymmetry causes excess corporate cash holdings [92]. The greater the degree of strategic variation, the greater the risk to the business and the higher the level of audit and tax concern [93,94]; enterprises will increase the level of cash-holding due to uncertainty. Previous studies have shown that the strategic difference is positively correlated with the level of cash holdings [95]. This article examines the mediating effects of strategic positioning. From the perspective of the development of digital finance, it explores how digital finance can change the internal and external environment of enterprises, adjust the strategic choices of enterprises, reduce the strategic aggressiveness of listed companies, and then reduce the level of cash holdings of enterprises.

7. Conclusions and Implications of the Study

7.1. Research Conclusions

This study investigated the impact of digital finance on corporate cash holdings, identified the roles of organizational structure and corporate strategy in the impact mechanism, and empirically tested the theoretical logic using a sample of A-share-listed Chinese companies from 2011 to 2018. Through theoretical analysis and empirical tests, the main findings of this study were as follows: (1) the development of digital finance significantly reduces the level of cash holdings of listed companies, and the breadth of coverage (Breadth), depth of use (Depth), and digitization (Digi) of digital finance have the same effect, which also indicates that the positive effect of digital finance cannot be achieved without the synergy of its various dimensions; (2) the impact of digital finance development on the cash holdings of independent enterprises is more pronounced than its impact on group-controlled enterprises, because it provides diversified access to financing, improves financing efficiency, reduces transaction costs, and mitigates the organizational deficiencies of independent enterprises; (3) by reducing the strategic aggressiveness of public companies, digital finance undermines their precautionary incentive to invest in opportunities through cash holdings. The baseline regression results remained same after using dynamic panel model and instrumental variables strategy to reduce potential endogeneity issues. Sensitivity testing procedures, such as replacing key variables and excluding specific samples, all indicated satisfactory robustness of the empirical results. Our findings have important guidance implications for corporate liquidity management, organizational design, and strategic decision-making.

7.2. Insights and Recommendations

Based on the trade-off theory, the agency theory, and the pecking order theory, we explain the motivation of high cash holdings of listed companies based on three types of motives for cash-holding decisions and provide an in-depth analysis of the impact of digital finance on the level of corporate cash holdings in terms of both organizational form and strategic positioning.

The main implications of this study are as follows:

First, companies should accelerate the development of digital finance and adjust their cash-holding strategies according to changes in the external financing environment in a timely manner in order to access external funds efficiently and at low cost and reduce their level of cash holdings. According to the research results, because of improvements in the corporate financing and governance environments brought about by digital finance, the cash-holding levels of listed companies in regions with a higher degree of digital finance development were significantly lower. In order to improve the efficiency of stock capital holding funds, utilization, enterprises should adjust their cash-holding level in real time by taking into account property rights attributes, industry competition, and the level of development of digital finance.

Second, for organization design in the context of the digital economy, enterprises should construct appropriate organizational structures. The reduction in transaction costs brought about by digital finance reshapes the relative advantages of enterprise groups and independent enterprises. Enterprises should assess the transaction costs borne by organizations such as markets, independent enterprises, enterprise groups, and strategic alliances in the digital finance environment and design organizational structures that are adapted to the digital economy.

Finally, for strategic corporate decisions in the context of the digital economy, firms should adopt a level of strategic aggressiveness appropriate to their capital structure and capital risk appetite. This study found that digital finance reduced risk-taking by enterprises because of the maturity of digital finance developments such as payment and credit in China. Enterprises should assess the supply of capital and risk preferences of capital providers in the context of digital finance and make strategic decisions that meet the overall interests of stakeholders in order to achieve cocreation, coproduction, and sharing of corporate value.

7.3. Limitations and Future Research

First, due to the confidentiality of data and other reasons, data for non-listed companies are not available. This paper only takes China’s A-share-listed companies as research objects and does not involve non-listed companies. The scope of data sample selection is limited. Future scholars can analyze research on cash holdings of unlisted companies based on a digital finance perspective. Second, in terms of measuring the level of corporate cash holdings, this paper uses the overall cash holding level variable and does not distinguish cash holding levels based on three motives: transactional, precautionary, and profit-related. Digital finance may have different effects on cash holding levels based on different motives, and the results may vary. This research provides a new way of thinking for future research on the metrics of cash holdings. Lastly, the effect of digital finance on the cash holding level of enterprises is complex, this paper only focuses on two aspects, namely, the organizational form and strategic positioning of enterprises, subsequent scholars should go further and consider other micro effects on the level of cash holdings to obtain more reliable and comprehensive conclusions (Supplementary Materials).