Regional digital finance and corporate investment efficiency in China

ABSTRACT

Digital finance has a substantial effect on macroeconomics and plays an important role in corporate investment behaviour. However, few studies examine how digital finance affects corporate investment efficiency. We use the data of Chinese A-share listed companies for 2011–2017 and the provincial digital finance index developed by Peking University to document that digital finance significantly improves corporate investment efficiency. Our findings are supported by extensive robustness tests. In addition, we identify two mechanisms by which digital finance may affect corporate investment efficiency: by reducing financing constraints and stimulating corporate innovation. We find that digital finance has a more pronounced effect on non-state-owned enterprises, small firms, firms in the central and western regions of China, firms with fewer loans, and firms with a higher dependence on external financing, indicating that digital finance increases the inclusiveness of financial markets. Finally, our economic consequences test shows that digital finance increases the total factor productivity of firms. Overall, this study provides insights for developing countries. In particular, it suggests that digital finance can increase firms’ resource allocation efficiency.

KEYEORDS:

• Digital finance
• investment efficiency
• financial constraints
• innovation

JEL CLASSIFICATION:

• D21
• G20
• G31

Acknowledgment

We thank seminar participants at Zhongnan University of Economics and Law, and Central China Normal University for their helpful suggestions. We also acknowledge the financial support from the Major Project of National Social Science Foundation of China (21ZDA010), the National Natural Science Foundation of China (71991473; 71772178), and the Innovation and Talent Base for Digital Technology and Finance (B21038). All errors are our own.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ First, we consider different concepts. Fintech and digital finance are not exactly the same; the former is more technology-oriented, whereas the latter is more neutral in its orientation. That is, compared with Fintech, digital finance puts more emphasis on the use of digital technology to provide financial services, which is especially important in a highly regulated financial market such as the Chinese market.

Second, we use a different method. Lv and Xiong (2022) use the method of Richardson (2006) to measure corporate investment efficiency. As this method focuses on the residual item, which varies with the variables and estimation methods, it is not robust. In addition, while Richardson’s method is suitable for examining corporate investment efficiency in developed countries, its applicability to China is unknown. We use the sensitivity of investment to investment opportunities to measure corporate investment efficiency (Chen et al. 2011), as this is used in many Chinese papers and is more suitable than Richardson’s method for examining China’s current corporate investment situation. In addition, the corporate investment (INV) and investment opportunity (TQ) values included in our model are relatively fixed and cannot be changed arbitrarily, so they are relatively robust.

Third, we consider different mechanisms. We consider two channels – financing constraints and corporate innovation – as mechanisms, whereas Lv and Xiong (2022) consider corporate governance as a mechanism. However, other studies show that inefficient investment is caused by financial frictions, primarily financing constraints. Overall, digital finance increases corporate investment efficiency by easing financing constraints and stimulating corporate innovation, and our mechanism seems to be more closely linked than corporate governance to this topic.

Fourth, we perform a different heterogeneity analysis. We consider five types of heterogeneity – firm ownership, firm size, firm location, firm loans, and firm external financing dependence – to verify the inclusive characteristics of digital finance, and the results are in line with our prediction. However, Lv and Xiong (2022) use regional urbanization rate and institutional environment to test heterogeneity, which obviously lack heterogeneity at the level of firm characteristics.

Fifth, we add an analysis of economic consequences. The literature states that increasing investment efficiency is the key to increasing firm total factor productivity (Greenwood and Jovanovic 1990). We add to this by proving that digital finance increases firms’ total factor productivity, which complements the findings of Lv and Xiong (2022).

² This index has been used many times by previous studies (Li, Wu, and Xiao 2020; Ding, Gu, and Peng 2022), see Guo et al. (2020) for detailed index measurement method and the meaning of sub-index.

³ We only report the results of the second stage in Table 5. The regression results of the first stage are available upon request.

⁴ Furthermore, to establish a causal direction of the relationship between the Broadband China policy and corporate investment efficiency, we conduct a placebo test by randomly assigning cities to the Broadband pilot cities category. Following Zhang, Tao, and Nie (2022), we perform 1,000 baseline regressions. The results of the regression are distributed around 0 and have an approximately normal distribution, indicating that other non-observed factors do not affect the baseline results. Due to space limitations, relevant results are available upon request.

⁵ The reason for this is to remove the effect of heteroskedasticity in the model.

⁶ Referring to Kaplan and Zingales (1997), KZ1 = −1.001909 * Cash +3.139193 * Leverage − 39. 3678 * Dividends − 1.314759 * Ch +0.2826389 * TQ. Dividends is cash dividends divided by total assets, and Ch is cash holdings divided by total assets. See Appendix A for the definitions of Cash, Leverage, and TQ. Referring to Wei, Zeng, and Li (2014), KZ2 = −6.315 * Cash − 39.356 * Dividends − 3.494 * Ch +3.291 * Leverage +0.460 * TQ. It is worth mentioning that the two indicators have a high correlation, indicating that the financing constraint indicator is relatively stable.

⁷ We divide the sample into two groups using the annual median of firm size.

⁸ We divide the firm’s registered location in the east, middle and west of China to divide the region.

⁹ We divide the sample into two groups using the annual median of firm loans.

¹⁰ We divide the sample into two groups using the annual median of external financing dependence.