Risks
2021
,
9
, 99
20 of 27
Regression results show that, whether in city commercial banks or rural commercial
banks, there is a significant negative correlation between the level of regional Fintech
development and bank risk-taking. This means that the development of Fintech reduces
the risk-taking of banks, and it also exists in urban commercial banks and rural commercial
banks. Further comparative analysis shows that the regression coefficient of Fintech
development level of urban commercial banks is
−
0.0350, while that of rural commercial
banks is
−
0.0236. This suggests that while Fintech has a dampening effect on overall
bank risk, it is more pronounced among city commercial banks. The possible reason is
that the scale of urban commercial banks is generally larger than that of rural commercial
banks, and the customer groups faced by urban commercial banks are better. Therefore,
the financial technology of urban commercial banks has a stronger effect on curbing bank
risks than rural commercial banks.
7.2. Mechanism Inspection
Referring to the practice of
Gu and Yang
(
2018
), the intermediary effect test procedure
is used to examine the transmission mechanism of Fintech development affecting bank
risk-taking.
7.2.1. Based on the Intermediary Effect Test within the Bank
This paper uses the internal interest margin (NIM) and the governance cost (Gov-
ernance) as the intermediary variables to test the conduction path. Table
13
shows the
regression results of the mediation effect. Column (1) is not included in the intermediate
variable NIM. The results show that the total effect of Fintech 1 (Fintech level) on Z-Score
(bank’s risk-taking) is
−
0.0249, and it is significant at the level of 1%. Column (2) shows that
Fintech 1’s effect on the mediating variable NIM is
−
0.463, and it is significant at the level
of 10%. After the median variable NIM is included in column (3), the coefficients of Fintech
1 and NIM are significant. The direct effect of Fintech 1 on Z-Score is
−
0.0240, and the direct
effect of NIM on Z-Score is 0.00170. Comparing the columns (1) and (3), after including the
intermediary variables, the regression coefficient of the core explanatory variable Fintech
1 decreased, but both were significant. Therefore, the bank’s net interest margin is part
of the intermediary factor that Fintech affecting bank risk, and the intermediary effect
accounts for 3.614%. Similarly, Columns (4), (5), and (6) are listed as the regression results
of the inclusion of Governance (governance costs). The coefficients before and after the
inclusion are significant, and the management cost is also a part of the mediation factor
that influences bank risk in Fintech. Moreover, the mediation effect accounts for 11.217%.
Therefore, the results of this paper support hypothesis 2.
Dostları ilə paylaş: