Internet Finance, Financing of Small and Micro Enterprises and the Macroeconomy

ABSTRACT

With regard to lending to small and micro enterprises (SMEs), the cost of traditional finance is mainly marginal. By contrast, the cost of Internet finance is mainly fixed, and the average cost of Internet finance is largely reduced. From the perspective of cost structure and size changes, we build dynamic equilibrium models. We prove that the transition from traditional to Internet finance will increase the number of SMEs obtaining loans and the aggregate output. We also conduct a quantitative analysis with the result revealing that Internet finance will increase China’s credit-access SMEs by 619.91% and aggregate output by 2.72%.

KEYWORDS:

• Internet finance
• financing of small and micro enterprises
• macroeconomy
• dynamic general equilibrium model

JEL:

• E23
• G21

Acknowledgments

The authors thank Paresh Kumar Narayan (the editor), James Saunoris (the editor), and the two anonymous reviewers for numerous constructive comments and suggestions that greatly improved the article. All errors remain the authors’ own.

Disclosure statement

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

Supplementary Material

Supplemental data for this article can be accessed on the publisher’s website.

Notes

1. It should be pointed out that not all digital platforms make good use of Internet technologies and big data. The failure of some operators to make good use of Internet technologies and big data (e.g., some P2P platforms) cannot deny the ability of the entire Internet finance industry to use Internet technologies and big data to provide loans to SMEs.

2. In practice, the loan limit relies on the previous cash flow. Some examples are as follows. MYbank sets higher loan limits for SMEs that receive more money through QR codes (Zhang 2019). In the interaction between bank and tax, the loan amount hinges on the tax paid. The POS loan limit is based on the transaction flow of the POS machine (Wu 2019). In this model, we assume that the loan limit depends only on the cash flow in the previous period. More generally, assuming that the loan limit depends on the cash flow in past multiple periods does not affect the steady-state analysis and results below.

3. Note that q_t is exogenous to the family head, but Q_t is not; the latter can be adjusted by changing capital.

4. The meaning of Eq. (1)(1) $-{\eta }_t+\beta {\eta }_{t+1}\left(1-\delta +\alpha {x_{t+1}}^{\alpha -1}\frac{\mathrm{\partial }{x}_{t+1}}{\mathrm{\partial }{k}_{t+1}}\right)+{\beta }^2{\eta }_{t+2}\alpha {x_{t+2}}^{\alpha -1}\frac{\mathrm{\partial }{x}_{t+2}}{\mathrm{\partial }{k}_{t+1}}=0$(1) is as follows. Consider a unit of goods in period t. If it is used for current consumption, the utility increment will be ${\eta }_t$. If it is saved, in period t + 1, goods will increase by $\alpha {{\mathrm{x}}_{t+1}}^{\alpha -1}\frac{\mathrm{\partial }{x}_{t+1}}{\mathrm{\partial }{k}_{t+1}}$, consumption will increase by $1-\delta +\alpha {x_{t+1}}^{\alpha -1}\frac{\mathrm{\partial }{x}_{t+1}}{\mathrm{\partial }{k}_{t+1}}$, and utility will increase by $\beta {\eta }_{t+1}\left(1-\delta +\alpha {x_{t+1}}^{\alpha -1}\frac{\mathrm{\partial }{x}_{t+1}}{\mathrm{\partial }{k}_{t+1}}\right)$; in period t + 2, due to the increase of the loan limit, goods and consumption will increase by $\alpha {x_{t+2}}^{\alpha -1}\frac{\mathrm{\partial }{x}_{t+2}}{\mathrm{\partial }{k}_{t+1}}$, and utility will increase by ${\beta }^2{\eta }_{t+2}\alpha {x_{t+2}}^{\alpha -1}\frac{\mathrm{\partial }{x}_{t+2}}{\mathrm{\partial }{k}_{t+1}}$. The utility increments in the two cases should be the equal.

5. Note that q_ss′ is larger than q_ss, and Q_ss′ is smaller than Q_ss. That is, capital is allocated from entrepreneurs with productivity in the interval q_ss–q_ss′ to entrepreneurs with productivity in the interval Q_ss–Q_ss′. Therefore, the resource allocation is improved.

6. We also conduct sensitivity analyses in Appendix S4 in the Supplementary Material and find that results are robust for reasonable values of β_c, σ, μ, and λ. In addition, in Appendix S5 in the Supplementary Material, we discuss the relationship between this work and the literature on barriers to Internet adoption.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [Grant number 71903081] and the Innovation and Entrepreneurship PhD Program of China’s Jiangsu Province.