Informal Finance and Enterprise Digital Transformation

2.2.1 Information Mechanisms of IF Affecting the Digital Transformation of Enterprises

IF can mitigate the inhibiting impact of inadequate market perception and cognitive bias on enterprise digital transformation. The financial market’s perception of an enterprise is derived from both the enterprise’s own information disclosure and media coverage (Patell & Wolfson, 1982). When enterprises disclose insufficient information, formal financial institutions struggle to comprehend their true circumstances, leading to adverse selection and moral hazard issues that hinder full financing from formal finance (Fazzari & Athey, 1987), thereby impeding their digital transformation efforts. Moreover, enterprises with low information transparency are susceptible to scrutiny and negative speculation by the media, resulting in a tendency for negative news reporting about the enterprise (Jin et al., 2021). This not only triggers investor panic in the capital market and mass stock sell-offs but also instills caution among banks and other indirect financial institutions when evaluating loan applications from these enterprises (Jia et al., 2023). Consequently, credit issuance conditions become more stringent, posing challenges for obtaining formal financial support essential for successful digital transformation. IF possesses the capability to comprehensively comprehend enterprise information (Ayyagari et al., 2010). On the one hand, it excels in acquiring and processing non-quantitative “soft information” of enterprises, enabling it to assess the authenticity of public information objectively and neutrally (Nguyen & Canh, 2021). Consequently, IF serves as a novel financing channel for enterprises with limited disclosure but significant growth potential, facilitating their digital transformation endeavors. On the other hand, IF leverages social networks to access non-public enterprise information, identify high-quality businesses, and provide them with financial assistance (Deng et al., 2019). This support enhances the operational performance of funded enterprises while allowing them to concentrate on digital projects, thereby expediting the process of digital transformation. By addressing these two perspectives collectively, IF mitigates information asymmetry between enterprises and external markets by fostering a deeper understanding among market participants regarding enterprises’ operations. As a result, concerns and panic-induced negative news about companies gradually diminish while simultaneously enhancing their ability to secure financial backing from external markets – ultimately facilitating their realization of digital transformation goals.

It is important to recognize that high interest rates, despite their apparent drawbacks, serve as a critical signaling mechanism within the information framework. High interest rates signal the scarcity and value of capital, encouraging enterprises to exercise greater caution in securing funds and to selectively invest in more promising and rewarding digital technology projects. Additionally, high interest rates motivate firms to optimize capital usage and expedite technological upgrades and digital transformation. Consequently, even in an environment of high interest rates, IF can still support firms’ digital transformation by leveraging these informational mechanisms.

2.2.2 Financing Mechanisms of IF Affecting the Digital Transformation of Enterprises

IF can mitigate the financing constraints faced by enterprises in their digital transformation endeavors. The challenges of difficult and costly financing hinder the realization of digital transformation, which is influenced by both financial difficulties encountered by enterprises and the inherent characteristics of digital transformation itself (Bedenikovic et al., 2017). On the one hand, enterprises grappling with financing difficulties and high costs face issues such as cash flow shortages and elevated financial leverage, often leading to heightened debt service risks, making it more arduous to secure funding in formal financial markets (Raj & Sen, 2015). On the other hand, investments in digital transformation primarily yield intangible assets like technology, resulting in a decline in the proportion of tangible assets (Zhang et al., 2023). Valuation and pricing disagreements surrounding intangible assets introduce additional risks including cash realization difficulties and depreciation concerns. Consequently, enterprises undertaking digital transformation frequently lack effective collateral while further diminishing their ability to raise funds through formal financial channels. In contrast, IF operates based on geographic and kinship networks. Due to interpersonal relationships and lower collateral requirements, enterprises facing financing difficulties can access credit resources from informal financial channels to support their digital transformation initiatives (Charles & Mori, 2016; Wellalage & Fernandez, 2019). Moreover, in the event of a debt-servicing crisis, IF provides more flexibility for negotiation through ex post renegotiation (Madestam, 2014). Unlike formal finance that seeks compensation through legal means when an enterprise defaults, IF allows enterprises to negotiate loan terms, interest rates, and other conditions (Allen et al., 2019). This alleviates the urgency of debt repayment and facilitates the continued operation of established digitalization projects.

It is important to recognize that although IF often involves high interest rates, these elevated rates also carry significant positive implications for financing mechanisms. High interest rates serve as a reflection of the market’s risk pricing mechanism and effectively filter out enterprises with genuine potential. Firms willing to accept higher borrowing costs typically possess greater risk-taking capacity and growth potential, making them more likely to succeed in digital transformation. Consequently, despite the presence of high interest rates, IF can still play a crucial role in facilitating the digital transformation of enterprises through its financing mechanisms.

Based on the above analysis, research hypothesis 1 is proposed.

Hypothesis 1: The advancement of IF facilitates the facilitation of enterprises’ digital transformation.

3.2.1 Explained Variables

Enterprise digital transformation (DLTN). In this study, we refer to the works of Hu et al. (2023) and Chen et al. (2023) to analyze the frequency of digitized keywords in corporate annual reports. Specifically, we employ text recognition techniques to quantify the occurrence of digitized transformation keywords and calculate their proportion relative to the total number of words in these reports.

3.2.2 Core Explanatory Variables

Informal financial development (IF) is quantified in this study by calculating the ratio of microfinance loan balances to the local area size within each province, with a subsequent division by 1,000 to normalize the scale (Tan & Tian, 2021). While microfinance companies represent only a fraction of IF, they serve as an indicator of standardized management under governmental guidance. Given data limitations on other forms of IF, employing this ratio remains justifiable for assessing resource intensity and regional disparities in informal financial development.

3.2.3 Control Variables

Considering that the digital transformation of enterprises is influenced not only by the development of IF but also by the enterprise’s developmental status, this study selects control variables at the enterprise level, specifically including: enterprise size (size, logarithm of total assets), enterprise income status (sale, logarithm of total operating income), book-to-market value ratio (BM, ratio of book value to market value), shareholding concentration (Concen, proportion held by largest shareholder), age of establishment (age), Tobin’s Q value (Tobing, market value to replacement cost ratio), concurrent appointment status of shareholders and executives (dual; 1 if chairman serves as general manager and 0 otherwise) and audit opinion type (audit; 1 if no qualified opinion exists and 0 otherwise).

4.2.1 Consideration of Lag Effects

The empirical results in Table 2 demonstrate the relationship between IF and firms’ digital transformation over time. Columns (1)–(3) examine the impact of lagged IF on firms’ digital transformation, while columns (4)–(6) investigate the effect of IF on the antecedent term of firms’ digital transformation. In all columns, the core explanatory variables exhibit positive coefficients that are statistically significant at least at a 5% confidence level. For columns (1)–(3), the coefficients for IF lagged by 2, 3, and 4 periods are gradually increasing to 0.394, 0.456, and 0.492, respectively. Similarly, in columns (4)–(6), there is also a highly significant effect of IF on enterprise digitization front after a delay of 2, 3 and 4 periods. Therefore, although the driving force from informal financial development towards firms’ digital transformation shows marginal incremental characteristics, this cumulative effect indeed accelerates the process of firms’ digital transformation over a minimum time span of four years.

4.2.2 Deletion of Special Samples

The level of informal financial development and the digital transformation behavior of enterprises are closely intertwined with the domestic and international economic and financial environment. Therefore, it is essential to consider economic and financial factors that may impact both aspects in the regression analysis. In this study’s research sample, two significant financial shock events are incorporated: one being the global financial crisis in 2008, which had a lasting influence on trends; and the other being China’s stock market crash in 2015. These shocks not only disrupt borrowing and lending activities in the informal financial market but also distort firms’ willingness to undergo digital transformation. Due to their inherent nature, these types of shocks cannot be effectively controlled through variable construction methods alone; hence it is reasonable to re-test the samples after excluding data during such periods of financial shock events. Although this article does not include data from the 2008 international financial crisis, column (1) presented in Table 3 excludes data from 2010, as crises tend to have long-lasting effects over time. Under these circumstances, we find that the coefficient for informal financial development (L.IF) is estimated at 0.316 with a statistically significant result at a significance level of 1%. Consequently, our core conclusion regarding how informal financial development drives enterprise digital transformation remains unchanged. Column (2) eliminates the sample of the 2015 Chinese stock market crash based on column (1), which also excludes the samples from 2016 and 2017 to account for the trend effect of the stock market crash. At this juncture, the coefficient of informal financial development (L.IF) exhibits a positive sign and passes the significance test at a level of 1%. Furthermore, there are significant disparities in economic levels between municipalities and ordinary regions, suggesting that characteristics related to informal financial market development, management, as well as enterprise digital transformation may possess distinct attributes. Consequently, column (3) removes samples associated with municipalities; nevertheless, it still yields a significantly positive coefficient for IF development (L.IF).

4.2.3 Replacement of Variable Measures

The core variable measures in Table 4 are replaced to re-evaluate the influence of IF on firms’ digital transformation. Columns (1)–(3) employ three distinct methodologies for assessing firms’ digital transformation: the first based on the ratio of digitization-related intangible assets to total assets (DLTN1), the second based on the logarithmic value of keyword frequency related to digital transformation in the firm’s annual report (DLTN2), and the third is the number of patents filed by a firm for inventions in digital technologies (DLTN3). In columns (1)–(3), regression coefficients for L.IF with respect to DLTN1, DLTN2, and DLTN3 exhibit positive values, passing a significance test at least at a 5% level. Columns (4) and (5) measure informal financial development through two indicators: microfinance institution intensity (IF1), which is the ratio of the number of microfinance institutions to district area, and paid-in capital intensity (IF2), calculated as the ratio of paid-in capital of microfinance institutions to district area. The regression results in columns (4) and (5) indicate that the coefficient estimates for L.IF1 and L.IF2 are 0.025 and 0.030, respectively, both significant at the 1% level. These findings provide robust evidence supporting the role of informal financial development in driving business digital transformation.

4.2.4 Placebo Test

In our study, enterprise digital transformation may be influenced by certain stochastic factors that could introduce bias in causality identification. To mitigate this issue, we employ a counterfactual analysis through a placebo test. Specifically, we randomly assign the values of IF to firms, thereby constructing a spurious sample linking IF and firm digital transformation, and re-run the regression tests. If the regression coefficients and t-values from the spurious causality are smaller than those from the true results, it suggests that the stochastic factors have minimal interference with the baseline findings. Figures 1 and 2 depict the distribution of regression coefficients and t-values over 1,000 iterations, respectively. The means of these coefficients and t-values approximate zero, indicating a normal distribution, and both are significantly lower than the true regression outcomes (as indicated by the black vertical lines in Figures 1 and 2). Consequently, the placebo test results affirm that the stochastic factors do not substantially interfere with the baseline relationship, thus supporting the robustness of our conclusion that IF facilitates the digital transformation of firms.

Figure 1

Distribution of coefficients for placebo test.

Figure 2

Distribution of t-values for placebo test.

4.3.1 Instrumental Variables Method

To address the endogeneity in the regression model, we employed two instrumental variables and applied two-stage least squares estimation. The first instrumental variable is DIA, representing dialect diversity, calculated as the ratio of the number of distinct dialects in the region to the resident population. Higher values of DIA indicate greater linguistic diversity (Zhang & Fan, 2018). The second instrumental variable is SO, representing the development level of social organizations, measured as the ratio of the number of social groups to the resident population. Higher SO values suggest more developed social organizations (Wu et al., 2020).

Both DIA and SO satisfy the criteria for valid instrumental variables. First, greater dialect diversity implies more varied local customs and norms, reducing the likelihood of social capital formation and making private lending more challenging, thereby potentially hindering the informal financial market’s development. Conversely, more developed social organizations facilitate social capital formation, promoting the informal financial market’s growth. Second, neither dialect diversity nor social organization development directly influences corporate decision-making, ensuring the exclusion restriction for instrumental variables.

Table 5 presents the regression results using instrumental variables. Column (1) indicates that dialect diversity has a significant negative impact on informal financial development, with the p-value of the Kleibergen–Paap rk LM test being 0.002, confirming that the instrumental variables are valid at the 99% confidence level. The second-stage regression results in column (2) show that the coefficient of L.IF on DLTN is significantly positive. Furthermore, columns (3) and (4) confirm that social group serves as an appropriate instrumental variable, and the baseline relationship remains robust after addressing endogeneity through two-stage least squares.

4.3.2 Sensitivity Analysis of Omitted Variables

Although we have controlled for key factors that may influence the digital transformation, it remains challenging to account for all potential variables. To address the endogeneity issue caused by omitted variables, we employ a sensitivity analysis to evaluate the extent to which these variables impact the baseline relationship. Specifically, we select treatment variables and examine whether the coefficients and t-values of the baseline regression change when the magnitude of the omitted variables reaches multiples of the treatment variables. As illustrated in Figure 3, when the omitted variable reaches three times the magnitude of the treatment variable (size), the regression coefficient of IF on the digital transformation remains at 0.30 with no change in sign. In Figure 4, even when the intensity of the omitted variable is tripled relative to Size, the corresponding t-value for the regression coefficient still exceeds 1.96, passing the 5% significance test. These findings suggest that the endogeneity problem due to omitted variables does not substantially affect the baseline relationship.

Figure 3

Sensitivity analysis of omitted variables based on regression coefficients.

Figure 4

Sensitivity analysis of omitted variables based on t-value.

4.4.1 Characteristics of Enterprise Property Rights

The results of the heterogeneity test based on the micro perspective are presented in Table 6. Concerning the nature of property rights, IF plays a significant role in fostering the digital transformation of non-state-owned enterprises (with an L.IF coefficient of 0.311 and passing the significance test at 1%, as indicated in column (2)). Conversely, it does not exert a substantial impact on the digital transformation of state-owned enterprises (with a t-value of only 0.81, failing to pass the significance test at a confidence level of 10%, as shown in column (1)). This disparity arises from state-owned enterprises’ ability to access ample credit through formal financial channels due to their reliance on government creditworthiness, rendering IF gains insignificant for them. In contrast, non-state-owned firms face vulnerability regarding exclusion from formal financial markets (Allen et al., 2019) and encounter difficulties in securing complete financing. IF partially bridges this financing gap for non-state-owned enterprises and yields noteworthy marginal benefits for their digital transformation.

4.4.2 Enterprise Technology Characteristics

From a technological perspective, IF can effectively facilitate the digital transformation of high-tech enterprises (with an L.IF coefficient of 0.233 and passing the significance test at a 5% level, as indicated in column (3)). However, it does not contribute significantly to the digital transformation of non-high-tech enterprises (with a t-value of only −0.06 and failing to pass the significance test at a 10% confidence level, as shown in column (4)). Technology serves as the core attribute of high-tech enterprises; however, due to their substantial intangible assets and limited access to formal financial support, they encounter challenges in obtaining necessary funding. The lenient requirements for collateral within informal financial channels provide an opportune source of funds for high-tech enterprises, enabling them to leverage their technological advantages and increase investments in research and development (R&D). Consequently, this enhances their R&D capabilities while bolstering the success rate of digitalization projects – ultimately contributing to successful digital transformations. In contrast, non-high-tech enterprises lack a strong technological foundation; even if they secure funds from informal financial markets, these resources are unlikely to be effectively translated into tangible technological outputs. Instead, such endeavors may result in wastage of credit resources – thus rendering IF insignificant for their digital transformation.

4.4.3 Enterprise Life Cycle Characteristics

The formal financial sector typically assesses the creditworthiness of potential clients based on factors such as assets and profitability, often posing challenges for firms to secure adequate credit during critical stages of their life cycle. Consequently, enterprises in the most promising growth stage face financial exclusion (Allen et al., 2019). In contrast, the informal financial sector lacks such a preference and can fully support the development of these enterprises (as shown in columns (5)–(6)). Start-ups, with limited internal resources, can effectively alleviate their financial pressure through IF. Moreover, given their highly malleable production and business models, start-ups tend to embrace digitization and embark on the path of digital transformation when provided with sufficient financial resources (e.g., increased supply of IF). However, due to the firm’s size constraints, the impact of IF on digital transformation is weaker for start-ups compared to mature firms (0.234 < 0.297). Mature-stage enterprises possess broad financing channels and market shares while having comprehensive strategic plans for development. With support from IF, mature-phase enterprises can unleash their full development potential and enhance overall strength – laying a solid foundation for successful digital transformation. On the other hand, IF cannot support digital transformation efforts during decline periods for enterprises (t-value only reaches 0.78), failing to pass significance tests at a 10% confidence level (as shown in column (7)). The weakened innovation and R&D capabilities of enterprises in recession, coupled with a lack of visionary strategic thinking among management and a limited subjective willingness to embrace innovation and transformation, hinder the effective utilization of IF for digital development. Consequently, IF fails to facilitate the digital transformation of enterprises in recession.

4.4.4 Enterprise Geographic Location Characteristics

Differences in economic development and the degree of marketization across different regions in China have resulted in a more efficient allocation of formal financial resources in the east-central part of the country compared to the west. Therefore, can IF provide enhanced support to enterprises in the economically less developed western region? Table 7 presents group regressions from a macro perspective, geographically dividing China into East, Central, and West. The empirical findings demonstrate that IF has a facilitating effect on enterprise digital transformation both in the eastern and central regions (as indicated by columns (1)–(2)). However, this facilitating effect is significantly stronger in the central region than in the eastern region (0.607 > 0.246). Conversely, IF exhibits no impact on enterprise digital transformation within the western region (t-value of 0.79), failing to pass significance tests at a 10% level. The reason behind IF’s ability to promote digital transformation lies within its optimization role for resource allocation; as such, it effectively supports those enterprises genuinely requiring funds for their transformative endeavors. The role of IF in promoting the digital transformation of enterprises is comparatively weaker in the eastern region than in the central region due to the more developed formal finance sector in the former. Enterprises in the eastern region can obtain comprehensive support from formal financial institutions, thereby reducing their reliance on IF for digital transformation initiatives. Conversely, the central region exhibits less development and market orientation within its formal finance sector, resulting in a significant imbalance in resource allocation and a larger credit gap for enterprises. Consequently, IF plays a more pronounced role in facilitating digital transformation efforts among enterprises located within this region. However, despite expectations based on this logic, IF does not play a significant role in promoting digital transformation within the Western region. This disparity can be attributed to regulatory barriers, infrastructure limitations, and business culture differences. Firstly, inadequate regulatory effectiveness hinders the development of IF in the West. Insufficient regulatory policies often lead to a lack of standardization, making it difficult for enterprises to obtain efficient and stable financial services required for digital transformation. Secondly, underdeveloped infrastructure poses significant challenges. Inadequate network coverage, limited application of digital technologies, and substandard hardware equipment reduce the likelihood of enterprises leveraging informal financial tools for their digital upgrades. Lastly, variations in business culture play a crucial role. Traditional business practices in the western region tend to perpetuate existing models, resulting in a lack of awareness and initiative towards digital transformation. Additionally, the relatively rigid business environment and inefficient market resource allocation further impede the digital transformation of enterprises.