TYPE Original Research

PUBLISHED 11 January 2024

DOI 10.3389/fphy.2023.1297912

Bank digital transformation, bank

OPEN ACCESS competitiveness and systemic risk
EDITED BY
Matthieu Garcin,
Pôle Universitaire Léonard de Vinci, France Kaiwei Jia* and Xinbei Liu
REVIEWED BY School of Business Administration, Liaoning Technical University, Huludao, Liaoning, China
Peter Cincinelli,
University of Bergamo, Italy
Sunil Kumar,
University of Delhi, India
The aim of this paper is to analyze the impact of the digital transformation of
*CORRESPONDENCE
Kaiwei Jia, banks on their systemic risks. We find that the digital transformation of
1980jkw@163.com commercial banks can significantly inhibit the systemic risk of banks, and this
RECEIVED 20 September 2023 conclusion is still valid after considering the endogeneity of the model. The bank’s
ACCEPTED 22 December 2023 digital transformation reduces its systemic risk by increasing its own
PUBLISHED 11 January 2024
competitiveness. Further analysis shows that the reduction of banks’ marginal
CITATION costs due to digital transformation is a key factor in promoting banks’
Jia K and Liu X (2024), Bank digital
transformation, bank competitiveness and
competitiveness as the mechanism by which digital transformation reduces
systemic risk. banks’ systemic risk. The role of bank digital transformation in reducing
Front. Phys. 11:1297912. systemic risk is heterogeneous, which is more obvious in large commercial
doi: 10.3389/fphy.2023.1297912
banks, commercial banks that have not established financial technology
COPYRIGHT subsidiaries, and systemically important banks.
© 2024 Jia and Liu. This is an open-access
article distributed under the terms of the
Creative Commons Attribution License (CC BY). KEYWORDS
The use, distribution or reproduction in other
forums is permitted, provided the original digital transformation, commercial banks, systemic risk, bank competitiveness, China
author(s) and the copyright owner(s) are
credited and that the original publication in this
journal is cited, in accordance with accepted
academic practice. No use, distribution or
1 Introduction
reproduction is permitted which does not
comply with these terms. Information technology has become the primary engine for social and economic
development as a result of the third scientific and technological revolution. Nowadays,
the world has entered the era of the digital economy. Along with the disruptive impact of
digital technology on all industries, how to combine with digital information technology,
how to stand firm in the new round of technological revolution, how to realize their own
sustainable development, and other issues have become the focus of research in all
industries of society. The financial industry is no exception [1], and digital finance and
fintech are precisely the products of the deep integration of traditional finance and
technological innovation [2].
As an essential foundation and significant component of the financial industry,
commercial banks have become the earliest industry in the financial industry to face
the opportunities and challenges brought by digital technologies [3]. On the one hand, new
technologies such as block chain, the Internet of Things, and intelligent investment can
bring innovative value to commercial banks [4]. On the other hand, the rapid development
of fintech innovation accelerates market diversion, resulting in lower profits for commercial
banks [5]. In order to better develop in the wave of digitization and reduce certain profit
losses brought by fintech to banks, commercial banks are actively carrying out digital
transformation work. For example, in China’s banking industry, since November 2015,
when China’s Industrial Bank set up the CIB Digital Financial Services Co., Ltd., by the end
of 2021, 17 banks in China have set up fintech subsidiaries, which include five large state-
owned commercial banks, eight joint-stock commercial banks, two city commercial banks,
as well as Shenzhen Rural Commercial Bank and Guangxi Rural Credit Union. Meanwhile,
nearly one-half of the commercial banks have set up fintech and digital finance business
units. On top of that, more than half of China’s listed commercial banks have reached

Frontiers in Physics 01 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

strategic cooperation agreements with headline Internet companies We combined relevant data from 2013 to 2021 to conduct an
such as Jingdong, Tencent, and Ali. empirical study. We create a system of indicators for bank digital
Digital transformation is the central theme of the global transformation in three dimensions: cognitive, organizational, and
economy [6]. The digital transformation of commercial banks product. We empirically investigated the effect of banks’ digital
has changed the traditional forms of services and business transformation on systemic risk and its mechanism of action on the
practices in the banking sector [7], which has led to a better basis of measuring banks’ digital transformation. It is discovered
ability to adapt to market competition and customer consumer that the digital transformation of commercial banks successfully
preferences in the digital era [8], which contributes to the sound lowers systemic risk, and this finding holds true after considering the
operation and sustainable development of banks. The use of fintech endogeneity of the model and replacing variables. Consider that
can not only bring innovations in payment systems, credit markets, competitiveness is the basis for banks’ long-term survival and
and insurance to commercial banks [9], but also bring sustainable development. And it has always been important to
improvements in business models to banks by introducing investigate whether banks’ level of competition affects their
specialized platforms, reaching neglected customer segments, systemic risk and, in turn, the stability of the entire financial
improving customer selection, reducing banks’ operating costs, market. Therefore, regarding the study of the mechanism of
and optimizing banks’ business processes [10]. All of these help action, we mainly focus on banks’ competitiveness levels. The
keep banks running smoothly and lessen the possibility of bank mechanism analysis shows that digital transformation → increase
risks. Of course, digital transformation does have problems, such as in bank competitiveness (Lerner index) → decrease in systemic risk
increased inter-institutional correlation and homogenization of is the mechanism of action of bank digital transformation to reduce
transformation, which exacerbate banks’ risks. The operational systemic risk. Further decomposition of the Lerner index into bank
risks of banks can easily trigger bank failures, and the resulting pricing and marginal cost reveals that the decline in marginal cost
systemic financial risks can even lead to the outbreak of a financial due to bank digital transformation is the main reason for attenuating
crisis. This will have extremely strong negative impacts on the systemic risk. In addition, the impact of commercial banks’ digital
sustainable development of the banking industry and even the transformation on systemic risk has heterogeneity regarding bank
whole society and economy. Since the financial crisis in 2008, size, transformation mode, etc., and its effect is more substantial for
systemic risk has been the focus of academic research [11], and large commercial banks, commercial banks that have not set up
the prevention and resolution of systemic financial risk are even fintech subsidiaries in the process of transformation, and
more fundamental to the smooth development of a country’s systemically important banks.
economy. Yet traditional banks are again major players in The remainder of the text is structured as follows: In order to
systemic financial risk [12]. Thus, the intriguing question: Does determine the connection between bank digitalization and financial
the digital transformation of commercial banks contribute to the risk, we first review the pertinent literature. Then, based on existing
reduction of the banks’ systemic risk and, as a result, ensure the studies, we conduct a theoretical analysis to explore the mechanism
stable operation and long-term growth of the banking industry? In by which banks’ digital transformation affects systemic risk and put
order to further analyze the impact of banks’ digital transformation forward corresponding hypotheses. Subsequently, the research
on systemic risk, this study will address the following questions: design was carried out, and we collected relevant data from
First, how does systemic risk in banks change as a result of digital 31 listed commercial banks in China, detailing the variable
transformation? Second, what are the mechanisms underlying the measurement, model setup, and data sources of this paper. On
impact if the digital transformation of banks has an effect on this basis, an empirical analysis was conducted, including
systemic risk for banks? Third, does this effect vary for banks benchmark regression, a mechanism test, heterogeneity analysis,
with various qualities? and robustness check. Finally, our research conclusions and insights
The Chinese market is the primary focus of this study for the are presented.
following reasons: First of all, China is a leader in the development
of international fintech [3]. Meanwhile, due to the enormous
market in China, the rapid development of emerging 2 Literature review
technologies such as big data, block chain, and artificial
intelligence has pushed China’s fintech level to continuously After reviewing the previous research, academics have
enhance and have a significant impact on China’s financial concentrated on the micro and macro levels when examining the
industry [13]. Second, the Chinese government has permanently reasons behind systemic risk in banks. At the micro level, bank size
attached great importance to the stability of the financial market. It [16, 17], linkages among banks and other financial institutions
has never been lax in preventing systemic financial risks [14]. [18–20], the size of shadow banking [21], and the financial
Third, the structure of China’s financial system has its condition of banks and other financial institutions [22, 23] and
characteristics, and the dominant role of banks in indirect other factors all have an impact on banking systemic risk. Systemic
financing makes them occupy a significant position in the risk is also influenced at the macro level by variables like monetary
financial market. This bank-oriented financial market makes the policy [24], economic policy uncertainty [25], [26], and the pro-
vast majority of risks in China’s financial system concentrated in cyclicality of the economy [27]. However, with the advent of the
the banking system [15]. Therefore, choosing China as the research digital economy, data has become a new factor of production,
object to investigate the effects of digital transformation on leading to changes in the factors that constitute bank risk.
commercial banking systemic risk is very important and Digital transformation is a new concept that has emerged as the
representative. world enters the era of the digital economy, and a unified

Frontiers in Physics 02 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

understanding of it has not yet been formed in the academic commonality, in-creased interbank competition prompts banks to
community. Since the Financial Stability Board of the prefer portfolio diversification and increase interbank holdings of
United States put forward the concept of “digital transformation common loans [40], and banks’ common behavior is enhanced,
of commercial banks is the use of new technologies in the banking which exacerbates the accumulation of systemic risk [41]. According
industry to form innovations in products and services, etc.” in 2016, to Xu, Eva [42] China’s banking sector has high barriers to entry, and
more scholars have referred to the process of introducing digital competition among banks may increase banks’ future demand for
technologies in the industry as the process of digital transformation counterparty banking products, which aggravates the accumulation of
[28]. From past research, scholars usually analyze the research systemic risk. Some scholars have also argued that monopolistic behavior
related to bank digitalization from the perspective of banks’ use due to increased bank competitiveness can exacerbate the systemic risk
of digital technologies such as fintech. Regarding the issue of the of banks, which are characterized by “competitiveness-vulnerability”,
relationship between digitization and bank risk, most of the existing i.e., “competition-stability”. For European countries, the negative impact
studies focus on the impact of digital technology on banks’ of bank competition on systemic risk becomes more pronounced when
individual risk, especially bank risk-taking. There is a relative there is foreign investment in banks [43]. In terms of the Chinese
lack of analysis on the impact of bank digitization on their banking system, there is also a reduction of systemic risk due to banking
systemic risk. competition [44], [45].
Some scholars believe that using fintech by banks can effectively It is not difficult to find from the existing research that the research
reduce banks’ risk-taking. Utilizing digital technologies, such as fintech, on the impact of the bank’s digitalization on its risk focuses more on the
can help banks be better at determining the creditworthiness of their bank’s individual risk, and there is a relative lack of research on the
clients [13], cut down on their percentage of non-performing loans [29], impact of systemic risk. The analysis of the corresponding role of the
[30], and increase their competitiveness [31] to lower banks’ level of risk- mechanism is also not perfect. Although it has been concluded that the
taking and improve the stability of the banking sector. Li, He [32] found development of fintech by banks can improve their own
that bank fintech innovation can effectively reduce bank risk-taking from competitiveness, the online operation mode also helps banks to
four perspectives: bank operating income, capital adequacy ratio, expand the scope of their own services and attract more customers
operating performance, and risk control ability. Some scholars also to form a competitive strategy and gain a competitive advantage [46].
believe that the use of fintech can exacerbate the level of risk-taking by However, there is still a gap in the research on whether banks’
banks. Wang, Liu [33] found that fintech overall exacerbated the risk- competitiveness will be a factor affecting the overall systemic risk of
taking behaviors of banks after analyzing the relevant data in China from banks in the digitalization process. Since there are different views on the
2011 to 2018. And Chen, Yang [3] concluded that with the continuous impact of bank competitiveness on systemic risk, we believe that it is of
development of fintech, financial risk shows a changing trend of research interest to consider competitiveness as a mechanism of bank
increasing and then decreasing. digital transformation on systemic risk. At the same time, considering
Although there are individual scholars’ studies involving the that banks’ digital transformation is more about the reconstruction of
impact of digital technology on banks’ systemic risk, they only self-worth, there is still a particular difference from fintech, which uses
analyze it from the perspectives of risk-taking, inter-bank cutting-edge technology to innovate products and services. Therefore,
correlation, and banks’ leverage. Wang, Liu [34] found that the unlike the existing studies analyzing the systemic risk of fintech on banks,
increase in risk-taking propensity and inter-bank correlation the article focuses more on banks’ digital transformation. It conducts
brought about by the improvement in the level of bank fintech empirical research on the aforementioned question based on the fixed
pushes up the likelihood of the occurrence of systemic risk. Dong, effects and mediation effects models from the perspective of banks’ own
Wu [35] found that from the perspective of capital, after using competitiveness. The answer to this question is of great value in
leverage ratio as a decomposition index of capital adequacy ratio, the understanding the systemic risk formation mechanism in the context
use of fintech can effectively increase the level of leverage ratio of of digital transformation, improving the path of banks’ digital
banks, thus inhibiting the transfer of risk and reducing the systemic transformation, and maintaining financial stability.
risk of banks. However, there is a gap in research on the
competitiveness perspective of banks themselves.
While digital transformation can increase the level of 3 Theoretical analysis and research
competitiveness of banks. However, scholars have also not come to a hypothesis
consensus on the topic of how competitiveness among banks affects
systemic risk. The level of competitiveness of banks themselves is 3.1 Digital transformation and systemic risks
weakened by competition in the banking industry [36]. Some of commercial banks
scholars believe that bank competitiveness helps to reduce systemic
risk and that banks are characterized by “competitiveness-stability”, Information asymmetry is an important cause of bank systemic risk
i.e., “competition-fragility”. According to Beck, Demirgüç-Kunt [37], a [47]. The information asymmetry that existed between banks and their
bank’s financial position is more likely to be stable and less likely to result clients prior to banks going digital had the following effects on banks:
in systemic risk the higher the market concentration, or how competitive Initially, before loans were granted, applicants would frequently conceal
the bank is. From the perspective of risk-taking, scholars analyzed the their negative information. At this time, banks affected by information
relationship between competition and risk in Indonesian banks [38] and asymmetry not only find it challenging to discover the applicant’s
Vietnamese banks [39], and found that banks tend to take on more risk financial problems in a timely manner but also find it difficult to
when they face increased competition, which is not conducive to the accurately identify high-quality customers, which reduces the quality
stability of the financial system. From the perspective of bank of the bank’s credit. Second, the bank found it challenging to monitor the

Frontiers in Physics 03 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

lender’s cash flow after the loans were issued because of the information franchise license and loss of the additional income it would have received.
asymmetry issue. This raises the likelihood of the bank having non- Banks are compelled to adopt more cautious business strategies as a result
performing loans, lowers the quality of the bank’s credit, and aggravates of the high cost of bankruptcy, which also deters them from pursuing
the bank’s insolvency risk. Depositors will become suspicious of high-risk investments and lowers the likelihood of systemic risks.
associated banks when one exhibits bankruptcy-like behavior, which The reasons why digital transformation improves the
will generate serious group run problems and increase the possibility of competitiveness of banks are as follows: First, there is a lot of
systemic risk. However, banks’ digital transformation can effectively information. Digital transformation has broadened the channels for
solve the adverse selection and moral hazard problems caused by commercial banks to obtain information and enhanced the ability of
information asymmetry, thus effectively reducing banking systemic banks to collect information [53]. The comprehensive and dynamic
risk. Applying digital technologies such as big data and block chain grasp of borrower information increases the bank’s own competitiveness.
has broadened the scope of information collection by banks. The Second, there are many customers. On the one hand, digital
standardized and visualized information processing mode has transformation has brought offline business online, broadened the
improved the utilization rate of information and can effectively bank’s business coverage, extended customer acquisition channels,
alleviate the problem of information asymmetry in the traditional and expanded the bank’s total customer base. On the other hand,
management model of banks [48], [49], thereby reducing the intelligent services brought by digital technology optimize customer
systemic risk of banks. Specifically, the use of digital technology can experience, improve service quality, further enhance customer stickiness,
help banks form an accurate portrait of customer behavior. Before and enhance the market power of banks. Third, there are too many
lending, multi-dimensional information identification and intelligence products. The application of digital technology has accelerated the
risk identification improve the accuracy of banks’ credit evaluation of update speed of banking products and services, and promoted the
borrowers, help banks better identify high-quality customers, and diversification of banking business and the refinement and
enhance the forward-looking nature of risk supervision, thereby granulation of services [29], enhanced the diversity and difference of
improving credit quality and reducing the probability of default risk. products, and enhanced the competitiveness of banks themselves in the
After the loan, the supervision mode of dynamically tracking the flow of industry. Fourth, high efficiency and low cost. Online payment has
funds will help the bank detect suspicious behaviors of loan customers in changed the traditional business model of banks, which not only
a timely manner and quickly take remedial measures to enhance the increases the convenience of operation for customers but also
timeliness of risk tracking and the accuracy of risk treatment, effectively increases the efficiency of the bank’s services, reduces the cost of
reducing the possibility of bank risks and reducing its contribution to business, and increases the level of competitiveness of the bank itself.
systemic risk. With the continuous enhancement of the bank’s own
The long-tail theory shows that traditional banks frequently give competitiveness level, the performance level is further increased,
priority to the needs of the top 20% of large customers while neglecting which effectively promotes the bank’s capital growth. Capital
or even ignoring the needs of the remaining 80% of long-tail customers. accumulation can effectively alleviate the contagion effect of individual
Digital transformation enhances the ability of bank information risks and further reduce the possibility of bank systemic risks [15]. First of
collection, organizing, and processing comprehensively through bulk all, the more competitive banks have greater market power, they have
data processing, which not only realizes the accurate identification of higher cost plus or lower marginal costs, and can obtain high returns
potential high-quality long-tail customers and improves the profitability through higher loan interest rates or lower marginal costs, effectively
of the bank but also enhances the risk prevention ability of the bank itself promoting the bank’s capital. Growth, improving its own capital buffer
[50]. Improvements in profitability and risk control levels help to reduce capacity, effectively enhancing the bank’s own ability to resist risks, and
banks’ risk-taking behavior and bankruptcy risk due to profit-seeking reducing the possibility of systemic risks. Secondly, banks with stronger
motives [51]. In turn, this lessens the degree of spillovers from individual competitiveness can form more stable long-term lending relationships
bank risk and the level of bank systemic risk contribution. Based on the with customers, which helps to alleviate the information asymmetry
above analysis, the following research hypotheses are proposed in between the two, effectively reduces the probability of default risk, and
this paper: thus weakens the contribution of banks to systemic risks. Based on the
above analysis, this paper proposes the following research hypotheses:
H1: Under the condition that other conditions remain unchanged,
the digital transformation of commercial banks will reduce their H2: The digital transformation of commercial banks will reduce
systemic risk. their contribution to systemic risk by enhancing the competitiveness
of banks.

3.2 Bank digital transformation, bank

competitiveness, and systemic risk 4 Research design
Bank digital transformation reshapes its own competitiveness and 4.1 Definition and measurement of variables
affects the franchise value of banks. The theory of bank franchises holds
that enhancing one’s own competitiveness is the only way to increase the 4.1.1 Bank systemic risk
intrinsic value of bank franchises [52], and the self-discipline effect of Consistent with most of the existing literatures, this paper is
franchise values has an obvious restraint effect on bank risk behaviors. based on the ΔCoVaR theory proposed by Adrian and
Due to the existence of franchise value, bank bankruptcy will be more Brunnermeier [54] and adopts DCC-GARCH modified ΔCoVaR
expensive because it will result in double losses from the loss of the to measure systemic risk. Additionally, as proxy variables for

Frontiers in Physics 04 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

systemic risk of banks in the robustness test, the marginal expected Pit − MC it
Lerner it
(1)
loss (MES) proposed by Acharya, Pedersen [55] and the SRISK Pit
proposed by Brownlees and Engle [56] are used. The specific
Among them, Pit is the price of bank output, which is measured
measurements are presented in the Supplementary Appendix.
by the ratio of total income to total assets. MCit is the bank’s
marginal cost, which is calculated by transcending the
4.1.2 Digital transformation of commercial banks
logarithmic cost function. The Lerner index indicates the ability
This paper draws on the construction ideas of the Internet
of a company to obtain excess profits. The larger the Lerner index,
transformation index of the Internet Finance Research Center of
the stronger the ability of commercial banks to obtain excess profits,
Peking University and the digital transformation index measured
and the higher the level of bank competitiveness.
by Zhang, Guo [57], and based on the characteristics of the digital
transformation of the bank, the features of the annual reports of
listed companies are crawled with the help of the Python crawler
4.1.4 Control variables
According to the studies that are currently available, banks’ systemic
function, and the corresponding word frequency statistics and
risk is influenced by micro characteristics like their own assets and
principal component analysis are carried out. Specifically, the
liabilities. Accordingly, the following six micro-level control variables are
digital transformation index of commercial banks is constructed
added to the empirical model with reference to Wang, Liu [34], Shi, Sun
from the three dimensions of cognition, organization, and
[60], Denis, Sami [61], and Tian and Wang [62]: i) the size of the bank is
product. 1) Cognitive level, taking into account the ongoing
represented by the natural logarithm of its total asset size; ii) since the
advancement of Chinese digital technology and the
core capital adequacy ratio is able to effectively dampen the systemic risk
characteristics of the banking sector’s digital transformation.
of banks [62], the core capital adequacy ratio (CAR) is therefore included
Based on the research of Zhang, Guo [57], this paper adds
in the model; iii) the ratio of total assets to shareholders’ equity (LEV) is
seven words such as “technology”, “online”, “artificial
introduced into the model in the paper, taking into account that higher
intelligence (AI)”, “5G”, “financial technology”, “biometrics”
leverage is more likely to trigger systemic risk [12]; iv) the bank’s deposit
and “automation”, and counts the frequency of relevant words
and loan ratios are expressed through the ratio of total loans to total
appearing in the bank’s annual report constitutes the sub-
deposits, which measures the bank’s own liquidity; v) the ratio of the
indicator D-cognition of the cognitive level of the bank’s
bank’s net profit to total assets is used to express the bank’s return on
digital transformation through principal component analysis.
total assets (ROA), which reflects the bank’s profitability level; and (vi)
2) At the organizational level, with reference to the
the difference between the firm’s year of incorporation and the statistical
classification of the digital transformation of commercial
year (age) is included in the model in view of the fact that the firm’s age
banks from the organizational perspective, information is
may also have a certain impact on the systemic risk. Additionally, the
sorted out from five perspectives: planning and promotion at
following three macro-level control variables are introduced with
the leadership level, setting up new science and technology
reference to the studies conducted by Wang, Liu [34] and Lim, Costa
departments, setting up online services and other business
[63]: i) A nation’s GDP growth rate is a good indicator of its economic
departments, setting up financial technology subsidiaries, and
health; the stronger the macroeconomic environment, the more it
scientific and technological personnel. And finally, the sub-
contributes to lowering bank systemic risk. As a result, the regression
indicator D-organization at the organizational level of the
model in this paper includes the GDP growth rate (GGDPr) at the
digital transformation of commercial banks is synthesized. 3)
national level; ii) deposit reserves are included because, as a
At the product level, this paper manually searched for
national macro-prudential tool, they can effectively reduce
information on whether the bank has launched WeChat
systemic risk [63]; and iii) excessive price level growth in a
banking, mobile banking, online banking, remote banking, and
country may result in domestic pass-through inflation, which
open banking through channels such as commercial bank official
can then cause systemic risk. As a result, the model incorporates
websites and commercial bank annual reports, and finally
the national CPI growth rate as a macro-level control variable.
obtained the sub-indicator D-product at the product level.
The specific variables and definitions are shown in Table 2.
After the sub-indices are formed, the comprehensive index
Digitalize of digital transformation is synthesized by principal
component analysis, and it is used as a proxy variable for the
digital transformation of banks. The specific construction
4.2 Model setting
methods of relevant indicators are shown in Table 1. At the
This paper builds the following regression model to further test
same time, drawing on the measure of Xie and Wang [58], this
the influence of banks’ digital transformation on banks’ systemic risk
paper uses the total bank digitalization index (D_T) as a proxy
based on measuring the indicators of banks’ digital transformation
variable for the digital transformation of commercial banks in the
and systemic risk:
robustness test.

4.1.3 Mediating variables Delta CoVaR dcci,t
α0 + α1 Digitalizei,t +
βi Controli,t + μi

In order to explore the mediating effect of banks’ own + γi + εi,t
competitiveness in the process of digital transformation’s impact on (2)
systemic risk, this paper draws on the research method of Angelini and
Cetorelli [59] and measures the level of competitiveness of commercial Among them, i and t are the bank and the year respectively;
banks by Lerner index. Calculated as shown in (Eq. 1): Delta_CoVaR_dcci,t is the systemic risk of bank i in year t;

Frontiers in Physics 05 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

TABLE 1 Construction and processing methods of the digital transformation index.

Level Based on Processing method

Cognition Internet, Digitization, Electronics, Big Data, Blockchain, Cloud Computing, ①Collect and standardize the frequency of words in the annual report
Internet of Things, Artificial Intelligence (AI), 5G, Fintech, Intelligence,
Biometrics, Automation, Technology, Online ②Extract principal components

③Principal component analysis and synthesis

Organize Leadership planning and promotion, new technology department, online ① Whether to set up the digital Transformation Committee, Fintech
service business department, fintech subsidiary, staffing Development Committee, Fintech Office and other relevant management
institutions in the board of directors or senior management

② Whether to open financial Science and Technology Department, Software/

Technology Development Center, Data Center, Data Management Department,
Information Technology Department, Information Technology Department,
Financial Technology Research Institute, data banking Department and other
new science and technology departments

③ Whether to open the network finance department, online center, retail finance
department, network gold and wealth management department, direct banking
department and other business departments

④ Whether to establish a fintech subsidiary

⑤ Whether the annual report emphasizes the existence of scientific and

technological research and development team, the introduction of scientific and
technological talents, information technology personnel, information technology
personnel and other information

⑥ Synthetic tissue level subindex: the maximum value is 1, and the minimum
value is 0

Product WeChat Banking, Mobile Banking, online banking, remote banking, open ① Manual search of bank annual report and bank official website to determine
banking whether the bank has the above five products

② Synthetic product level subindex: the maximum value is 1, and the minimum
value is 0

TABLE 2 Variable definition and description.

Variable type Variable Variable name Variable specification

symbol
Explained variable Delta_CoVaR_dcc Bank systemic risk Conditions calculated based on the DCC-GARCH model

Explanatory variable Digitalize Bank digital transformation Principal component analysis and synthesis, and the detailed methods are shown in
index Table 1

Intermediary variable Lerner Level of bank competitiveness The Lerner Index, see Eq. 1 for details

Micro control variables Size Bank size Logarithm of total asset size

age Enterprise age The difference between the years of establishment of the enterprise and the statistical
year

car Capital adequacy Core capital adequacy ratio

LEV Leverage ratio Total assets/Shareholders’ equity

LDR Loan to deposit ratio Total Loans/Total Deposits

ROA Return on total assets Net profit/Total assets

Macro control GGDPr Gross domestic product Year-on-year growth rate of GDP
variables
GCPIr Rate of inflation Year-on-year growth rate of CPI

rd Deposit reserve ratio Deposit reserve requirements for large financial institutions

Digitalizei,t is the degree of digital transformation of bank i in year t. error. In addition, in order to alleviate the influence of
Controli,t is the control variable at the micro and macro levels; μi is heteroscedasticity, the empirical results all adopt clustering robust
the banks fixed effects; γi is the time fixed effect; εit is the random standard errors.

Frontiers in Physics 06 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

TABLE 3 Descriptive statistics of the variables.

Variable name Mean Median Maximum Minimum Standard deviation Number of observations
Delta_CoVaR_dcc 9.720 9.828 22.525 3.087 3.972 188

Digitalize 2.137 1.941 4.087 0.176 1.226 188

Lerner 0.496 0.517 0.873 0.086 0.243 188

Size 28.669 28.928 31.138 25.418 1.542 188

age 30.176 25.000 108.000 8.000 19.855 188

car 10.091 9.605 14.020 7.990 1.579 188

LEV 13.842 13.425 18.328 10.641 1.903 188

LDR 79.002 77.179 115.985 47.426 13.672 188

ROA 0.888 0.876 1.390 0.505 0.189 188

GGDPr 8.752 8.533 13.385 2.742 3.377 188

GCPIr −0.238 −0.389 0.784 −1.561 0.764 188

rd 14.968 14.500 20.000 11.500 2.938 188

4.3 Samples and data sources maximum value of 0.873 and a minimum value of just 0.086,
showing that there are clear differences in the level of
As of 2021, there are 41 A-share listed commercial banks in competitiveness among banks.
China. In terms of data selection, due to the short data interval, this To observe the correlation between the variables more
paper excludes four banks that were listed in 2021, Ruifeng Bank, intuitively, in Table 4, we analyze the correlation of the variables
Qilu Bank, Shanghai Rural Commercial Bank and Bank of involved in this paper. The correlation coefficient between systemic
Chongqing. At the same time, since the digital transformation risk and banks’ digital transformation is −0.597, as shown by the
indicators are constructed through text analysis, the availability correlation coefficient matrix in Table 4. This negative correlation is
of the original disclosure data is critical, so six companies with consistent with Hypothesis H1, which serves as the foundation for
missing information, including Changshu Rural Commercial Bank, the empirical research that follows in this paper.
Bank of Beijing, China Everbright Bank, Xiamen Bank, Jiangsu Zijin
Rural Commercial Bank and Jiangsu Suzhou Rural Commercial
Bank, were excluded bank. Finally, 31 A-share listed banks were 5 Empirical tests
retained, including 6 state-owned commercial banks, 8 joint-stock
commercial banks, 12 city commercial banks, and 5 rural 5.1 Digitalization of commercial banks and
commercial banks. The text data that digital transformation relies banking systemic risk
on is obtained through manual searches such as bank annual
reports, web page information, and bank official websites. Stock The baseline regression results for the impact of commercial
return data comes from Choice Financial Terminal (https://choice. banks’ digital transformation on banking systemic risk are shown in
eastmoney.com/). Other variables mainly come from Wind database Table 5, where columns (1) and (2) show the results without and
(https://www.wind.com.cn/) and CSMAR database (https://global. with control variables, respectively. The empirical findings
csmar.com/). Considering that 2013 is the first year of digital finance demonstrate that the regression coefficient’s sign direction is
in China, this paper selects 2013–2021 as the sample period. consistent both before and after the introduction of a number of
This paper performed a 1% winsorize on the continuous control variables. It demonstrates the robustness of this model.
variables to lessen the impact of outliers on the empirical Specifically, the regression coefficient of bank digital trans-
analysis, and in the end, it was able to produce the formation is significantly negative at the 5% level after all control
descriptive statistics of the unbalanced panel data in Table 3. variables are included in column (2), showing that the likelihood of
Among them, the mean, maximum, and minimum values of bank systemic risk gradually decreases with increasing bank digital
banking systemic risk are, respectively, 9.720, 22.525, and 3.087. transformation. For every unit of bank digital transformation, its
This demonstrates that there were differences in the systemic contribution to systemic risk decreases by 25%. This confirms the
risk among various banks and a certain polarization trend in the anticipated conclusion drawn from the earlier theoretical analysis
banks’ systemic risk levels during the observation period. The that the digital transformation of commercial banks can lower the
bank digital transformation index has average, maximum, and systemic risk of banks. The specific cause may be that during the
minimum values of 2.137, 4.087, and 0.176, respectively, digital transformation process, the use of digital technologies like big
reflecting the varying degrees of digital transformation data and cloud computing, on the one hand, lowers the cost of
experienced by various commercial banks. The Lerner index, information acquisition for banks, effectively resolving the issue of
which measures the level of bank competitiveness, has a information asymmetry between banks and enterprises, and, on the

Frontiers in Physics 07 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

TABLE 4 Correlation coefficients of variables.

Delta_CoVaR_dcc Digitalize Size age car LEV LDR ROA GGDPr GCPIr rd
Delta_CoVaR_dcc 1.000

Digitalize −0.597 1.000

Size 0.071 0.383 1.000

age 0.018 0.296 0.590 1.000

car 0.006 0.123 0.193 0.344 1.000

LEV 0.127 −0.382 −0.115 −0.300 −0.606 1.000

LDR −0.022 0.323 0.282 0.106 −0.020 −0.499 1.000

ROA 0.185 −0.148 0.384 0.301 0.314 0.105 −0.266 1.000

GGDPr −0.456 0.221 0.017 0.012 −0.003 0.040 0.013 0.077 1.000

GCPIr 0.231 −0.124 0.008 0.021 −0.002 0.100 −0.105 0.097 −0.359 1.000

rd 0.219 −0.364 0.181 0.083 −0.145 0.591 −0.361 0.539 0.077 0.212 1.000

TABLE 5 Benchmark model regression results.

Variable Delta_CoVaR_dcc Delta_CoVaR_dcc

(1) (2)
Digitalize −0.216* −0.250**

(−1.84) (−2.10)

Control variable Not control Control

Banks fixed effects Yes Yes

Time fixed effect Yes Yes

R2 adjusted 0.955 0.957

F 292.398 295.725

N 188.000 188.000

Note: t values are in brackets; *** ** * mean significance at 1%, 5%, and 10% levels, respectively. Similarly hereinafter.

other hand, enhances the ability to prevent and control risk before Delta CoVaR dcci,t
α0 + α1 Digitalizei,t + Σθi Controli,t + μi
lending. On the other hand, the bank’s service system has been + γi + εi,t
expanded to include high-quality long-tail groups, which has
(3)
increased its profitability and decreased the likelihood of systemic
risks. This result is consistent with the previous theoretical analysis M i,t
φ0 + γ1 Digitalizei,t + Σθi Controli,t + μi + γi + εi,t (4)
and verifies that hypothesis H1 of this paper is established. Delta CoVaR dcci,t
β0 + β1 Digitalizei,t + β2 M i,t
(5)
+ Σθi Controli,t + μi + γi + εi,t

5.2 Digital transformation of commercial The intermediary variable among them is Mi,t, and the other
banks, bank competitiveness, and bank variables are the same as in formula (2). Table 6 displays the results
systemic risk of the regression.
From the results in Table 6, we can see that in column (1), the
From the results of the above benchmark regression, it can be seen impact of the digital transformation of commercial banks on the
that the digital transformation of banks has a negative inhibitory effect Lerner index is positive at the 10% significance level because the
on their systemic risk. So, what is the mechanism for the impact of digital Lerner index reflects the level of competitiveness of commercial
transformation on the systemic risk of commercial banks? Is bank banks. Therefore, it can be concluded that the digital transformation
competitiveness an important factor in commercial banks’ digital of commercial banks has significantly improved the competitiveness
transformation to reduce their systemic risks? In order to better of banks themselves at a level of 10%. It can be seen in column (2)
explore related issues, this paper uses the mediating effect model for that the regression coefficient of the digital transformation of
analysis. The specific model is shown in Eqs 3–5 below: commercial banks is negative at the 10% significance level, and

Frontiers in Physics 08 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

TABLE 6 Mediating effect test.

Variable Bank competitiveness (Lerner index) Price(P) Marginal cost (MC)

(1) (2) (3) (5) (6)

Lerner Delta_CoVaR_dcc P MC Delta_CoVaR_dcc

Digitalize 0.023* −0.222* 0.0002 −0.001* −0.216*

(1.79) (−1.85) (1.17) (−1.75) (−1.79)

M −1.201** 32.080*

(−2.54) (1.90)

Control variable Control Control Control Control Control

Banks fixed effects Yes Yes Yes Yes Yes

Time fixed effect Yes Yes Yes Yes Yes

2
R adjusted 0.719 0.958 0.540 0.651 0.958

F 80.081 420.227 220.879 64.205 365.826

N 188.000 188.000 188.000 188.000 188.000

Note: Table 5 has given the regression results of the first step of the mediation effect test. In Table 5, there is a significant negative correlation between the digital transformation of commercial
banks and systemic risk. The first step of the mediating effect is established, and the second and third steps can be tested. At the same time, in order to avoid repeated output of results, this form
only reports the test results of Steps 2 and 3. The t values in the brackets ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

the regression coefficient of the Lerner index is negative at the 5% following aspects: First of all, from the perspective of banks
significance level. All the above steps are in line with the test providing services, on the one hand, although the basic
conditions of the mediation effect model, indicating that the investment involved in digital transformation is relatively large,
bank’s own competitiveness level plays an intermediary role in however, with the use of digital technology, the online business
the impact of the digital transformation of commercial banks on model has led to a continuous reduction in service labor costs and
systemic risk and that the bank’s digital transformation reduces store operating costs. On the other hand, digital technology brings a
bank systemic risk by affecting competitiveness. The mediation large number of high-quality long-tail customers into the bank’s
effect is 0.028. This shows that the bank’s digital transformation service scope, increasing the total number of customers. At the same
can reduce systemic risk by improving its own competitiveness. This time, the exclusive customized service model brought by big data
is consistent with the previous theoretical analysis and verifies the technology has greatly increased the quality and efficiency of
hypothesis H2 of this paper. banking services, further enhancing customer stickiness. It can be
Considering that the Lerner index, which measures the level of seen that with the continuous deepening of digital transformation,
bank competitiveness, is composed of two parts: price and marginal banks have not only achieved a reduction in service costs but also
cost, Therefore, after decomposing the Lerner index into price (P) expanded the overall number of high-quality customers. This has led
and marginal cost (MC), it is further analyzed to determine which to a gradual decline in the marginal cost of servicing each customer
index is the key to promoting the level of bank competitiveness to for banks. Secondly, from the perspective of information collection
become the mechanism of commercial bank digital transformation and processing, on the one hand, after the transformation, the
to reduce bank systemic risk. The empirical results are shown in application of digital technology can help banks collect more
columns (3) through (5) of Table 6. extensive customer-related information and provide them with
Since the price (P) in this article is measured by the ratio of total diversified and targeted information after accurately analyzing
income to total assets, its essence is the turnover rate of total assets. customer behavior. Diversified products and services help to
For banks, although digital transformation can prompt them to expand the bank’s advantages of economies of scope, which are
expand their business scope and improve service efficiency. consistent with the reduction of marginal costs emphasized in the
However, from the perspective of China’s monetary policy, since theory of economies of scope. On the other hand, the batch
2013, China has entered a cycle of interest rate cuts, and the processing of data and the continuous improvement of
narrowing of interest rate spreads has weakened the profitability information systems have also made the marginal cost of bank
of banks, which has reduced the asset turnover rate of banks to a information processing lower and lower. Thirdly, from the
certain extent. At this time, banks are more likely to ensure stable perspective of risk management, with the optimization of the
operations by reducing costs, thereby mitigating systemic bank risks. bank’s overall business chain in the digital transformation
To sum up, the mechanism for digital transformation to increase process, the bank’s internal risk management system is becoming
bank competitiveness by increasing price P to reduce bank systemic more and more perfect. The information tracking of borrowers
risk is not significant. before and after lending has achieved full coverage of business risk
Regarding marginal cost (MC), the reduction effect of bank management. Moreover, the decentralization of the block chain
digital transformation on its marginal cost is mainly reflected in the increases the accuracy of information and reduces the marginal cost

Frontiers in Physics 09 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

of bank risk management. On the one hand, the capital financial technology subsidiaries, and cooperating with
accumulation brought about by reducing marginal costs helps to Internet or technology companies. And the difference in
increase the bank’s risk prevention and control capabilities. On the transformation mode may have different impacts on its
other hand, it increases the liquidity level of banks, reduces the systemic risk. Therefore, this paper conducts a heterogeneity
bank’s dependence on interbank lending, reduces the degree of test based on the model of bank digital transformation.
financial correlation between banks, and thus reduces the possibility Considering that all sample banks in the sample period have
of systemic risks. set up digital or technical electronic services and cooperate with
From the empirical analysis results in Table 6, we can see that the Internet or technology companies to a certain extent. So, the
regression coefficient in column (3) does not meet the condition of a heterogeneity analysis focuses on whether to establish financial
mediating effect. When marginal cost is used as an intermediary technology subsidiaries.
variable, the coefficient of the core explanatory variable in column Columns (3) and (4) in Table 7 report the test results of
(4) is significantly negative at the 10% level, indicating that the establishing a fintech subsidiary, and not establishing a fintech
digital transformation of banks helps reduce the marginal costs of subsidiary. Among them, the regression result of column (3) is
banks. In column (5), the coefficient of bank digital transformation not significant, and the regression result of column (4) is
is negative at a significance level of 10%, while the coefficient of significantly negative at the 1% level. This result suggests that
marginal cost is positive at a significance level of 10%. It shows that the systemic risk reduction effect of banks’ digital
marginal cost plays an intermediary role in the impact of a transformation is more pronounced among banks without
commercial bank’s digital transformation on systemic risk and establishing fintech subsidiaries, under equal conditions of
further shows that the reduction of the marginal cost of banks having opened digital or technology-enabled e-services and
brought about by digital transformation is the key factor that having partnered with internet or technology firms. The
promotes the competitiveness of banks to become the possible reasons are as follows: First, cooperation with
mechanism of digital transformation to reduce the systemic risk Internet or technology companies can effectively make up for
of banks. It is consistent with the above theoretical analysis. the bank’s own technical shortcomings. While making full use of
the technological resources of the partner company, it can
effectively reduce the innovation cost and innovation risk in
6 Heterogeneity analysis the process of the digital transformation of the bank, weakening
the bank’s possibility of contagion. Secondly, compared with
6.1 Scale heterogeneity building a financial technology subsidiary by itself, it is easier to
obtain returns in the short term by cooperating with enterprises,
Because China’s commercial banks have their own which reduces the motivation of banks to pursue high-risk assets
characteristics in terms of scale and type, in the process of digital in order to make up for the high investment costs of digital
transformation, the impact of commercial banks with different transformation, and the probability of systemic risks has also
characteristics on their systemic risk may also be different. decreased. Therefore, the reduction effect of bank digital
Therefore, referring to existing research methods, this paper transformation on systemic risk is more obvious in banks
classifies state-owned commercial banks and joint-stock that have not established financial technology subsidiaries.
commercial banks as large commercial banks and analyzes urban For banks to set up financial technology subsidiaries on their
commercial banks and rural commercial banks as small and own. On the one hand, although the establishment of financial
medium-sized commercial banks. Empirical results are shown in technology subsidiaries by banks is a means for them to promote
columns (1) and (2) of Table 7. their own digital transformation based on the advantages of the
From the results of columns (1) and (2) in Table 7, it can be seen parent Company. In the long run, the establishment of a fintech
that the regression coefficient of the digital transformation of large subsidiary will help improve the bank’s overall digitalization
commercial banks is significantly negative at the level of 1%. In level, enhance the bank’s own core competitiveness, and
comparison, the negative impact of the digital transformation of effectively reduce its systemic risk. However, Chinese
small and medium commercial banks on systemic risk only holds at financial technology subsidiaries appeared relatively late.
the 5% significance level. This shows that compared with small and Except for a few banks established in 2015 and 2016, most
medium-sized commercial banks, large banks have advantages in financial technology subsidiaries were established between
terms of capital scale, talent allocation, and risk management. 2018 and 2020. Due to the short establishment time, it is very
Therefore, it is easier to reduce its systemic risk through digital likely that there will be problems, such as an insufficient risk
transformation. management model and a single business operation. In the case
of an unsound banking supervision system, it is easier to induce
systemic banking risks. On the other hand, these banks that have
6.2 Heterogeneity of digital established financial technology subsidiaries are also carrying
transformation models out digital transformation through cooperation with Internet or
technology companies, and cooperation with Internet or
Through the collection of relevant data, it can be seen that technology companies can effectively reduce the contribution
the digital transformation models of the Chinese banking level of banking systemic risk. Therefore, in general, the impact
industry at this stage are mainly divided into three types: of digital transformation on the systemic risk of banks may not
setting up its own digital service departments, establishing be significant.

Frontiers in Physics 10 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

TABLE 7 Heterogeneity test.

Variable Scale heterogeneity Transformation Systemic importance

model heterogeneity

(1) (2) (3) (4) (5) (6)

Digitalize −0.327*** −0.491** −0.200 −0.401*** −0.346*** −0.161

(−3.02) (−2.03) (−1.23) (−2.68) (−2.87) (−0.57)

Control variable Control Control Control Control Control Control

Banks fixed effects Yes Yes Yes Yes Yes Yes

Time fixed effect Yes Yes Yes Yes Yes Yes

2
R adjusted 0.982 0.918 0.986 0.955 0.973 0.923

F 303.375 32.909 108.754 141.292 234.838 23.139

N 112.000 76.000 44.000 144.000 129.000 59.000

6.3 The heterogeneity of systemically 7 Robustness test

important banks and non-systemically
important banks 7.1 Endogeneity test

Compared with non-systemically important banks, This paper’s analysis suggests that banks’ digital transformation can
systemically important banks play a more important role in lower their systemic risk; however, more discussion is required to arrive
the generation and infection of banking systemic risks. at this conclusion. This is due to the possibility that the two
Therefore, the impact of digital transformation on different variables—bank digital transformation and bank systemic risk—have
systemically important banks may be different. According to a mutually causal endogeneity relationship. In order to reduce the
the list of Chinese systemically important banks in 2022 issued possible endogeneity problem in the model, this paper replaces the
by the People’s Bank of China and the China Banking and estimation method and uses a dynamic panel model (Generalized
Insurance Regulatory Commission, this paper divides the Method of Moments) for the test. Table 8’s column (1) illustrates
sample banks into two categories: systemically important and that the p-value for both the Sargan and AR (2) tests is greater than
non-systemically important for empirical analysis. The 0.1, indicating the absence of second-order autocorrelation in the
empirical results are shown in Table 7 (5) and (6). Among residual terms and the instrumental variables are appropriately
them, the regression coefficient of digital transformation of selected, confirming that the selection of the model is justified. From
systemically important banks is significantly negative at the the regression results, the number of estimated coefficients after the
1% level; the regression coefficient of non-systemically lagged first order of systemic risk of banks is positive at a 5% significance
important banks is negative but not significant. This shows level, indicating that systemic risk has a continuous effect. And the effect
that the reduction effect of digital transformation on bank of banks’ digital transformation on systemic risk is significantly negative,
systemic risk is more obvious in systemically important which is consistent with the results of the previous benchmark
banks. The possible reasons for this result are as follows: regression, which indicates that the results of this paper that banks’
First, systemically important banks have strong financial digital transformation can reduce systemic risk are reliable.
strength and obvious advantages such as scientific and
technological talents and R&D capabilities, making it easier
for such banks to form economies of scale in the process of 7.2 Other robustness tests
digital transformation. Under the influence of economies of
scale, the profitability of banks has gradually increased, their In order to ensure the robustness of the benchmark model, this
risk management and control capabilities have improved, and paper further analyzes the robustness from the following two
their ability to prevent systemic risks has become stronger. perspectives. First, the explanatory variables are replaced. On the
Secondly, because systemically important banks occupy an basis of the baseline model selecting ΔCoVaR to measure systemic
important position in Chinese banking system, they are risk, in the robust-type test, marginal expected loss (MES) proposed
subject to stricter financial supervision, prompting banks to by Acharya, Pedersen [55], and SRISK proposed by Brownlees and
choose more prudent decisions in the process of digital Engle [56] are selected to be used as proxies for systemic risk of
transformation. In addition, its decision-making execution banks. The exact formula is presented in the Supplementary
ability is strong, and the speed of adapting to the market is Appendix, and the regression results are presented in column (2)
relatively fast. All of these help to reduce the possibility of and column (3) of Table 8. Second, the core explanatory variables are
systemic risks in banks. replaced. In this paper, we draw on the measurement method of Xie

Frontiers in Physics 11 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

TABLE 8 Robustness test.

Variable Endogeneity test Other robust tests

(1) (2) (3) (4)

L.Delta_CoVaR_dcc 0.300**

(2.05)

Digitalize −0.607*** −0.891** −4.743**

(−4.15) (−2.06) (−2.48)

D_T −3.030***

(−4.83)

Control variables Control Control Control Control

Banks fixed effects Yes Yes Yes Yes

Time fixed effect Yes Yes Yes Yes

AR (2) −0.39

(0.694)

Sargan 83.16

(0.118)
2
R adjusted 0.765 0.247 0.825

F 93.331 18.769 127.712

N 157.000 188.000 188.000 188.000

Note: AR (2): z-value outside parentheses, p-value inside parentheses; Sargan: chi2 value outside parentheses, p-value inside parentheses.

and Wang [58] to reconstruct the bank digital transformation index, about by digital transformation is the key factor that promotes
and the bank digitalization index (D_T) is used as a proxy variable the competitiveness of banks to become the mechanism of digital
for digital transformation of commercial banks for the robustness transformation to reduce the systemic risk of banks. In addition,
test, and the regression results are shown in Table 8, column (4). the digital transformation of commercial banks has a
The estimates of the digital transformation coefficients for banks heterogeneous effect on reducing systemic risk. The digital
in columns (2) through (4) are all negative, as shown in Table 8. The transformation of large commercial banks, commercial banks
estimates of columns (2) and (3) among them are significant at a that have not established financial technology subsidiaries, and
significance level of 5%, whereas the estimate of column (4) is systemically important banks is more helpful in reducing the
significantly negative at a significance level of 1%. This suggests that systemic risk level of banks.
the digital transformation of banks is effective in reducing banks’
systemic risk. The regression results of the robustness tests are
similar to the previous benchmark regression, which again indicates 8.2 Recommendations
that the estimation results of this paper are more robust and further
validate the previous hypotheses. We can better understand the overall direction of banking digital
transformation and the development strategy of risk prevention and
control by examining the overall impact, mechanism, and
8 Conclusions and recommendations heterogeneity of banking digital transformation on banking
systemic risk. The future development strategy for the digital
8.1 Conclusions transformation of Chinese commercial banks should be
continually enhanced based on further strengthening risk
In the digital era, more and more companies are starting to prevention and control. In this regard, this paper puts forward
enhance their competitiveness through digital transformation. the following policy recommendations:
Does the digital transformation of the banking industry, a major Enhancing awareness of cooperation and promoting mutual
player in systemic financial risk, affect systemic risk? Does the benefit and win-win results: The improvement of the level of
digital transformation of banks make them more competitive, competitiveness will help to increase the bank’s own capital
which lowers systemic risk? This study demonstrates that accumulation and reduce the possibility of bank systemic risk.
commercial banks’ digital transformation boosts their own Moderate cooperation also has certain advantages for the
competitiveness and lowers systemic risk. Further research rational allocation of financial resources. Therefore, the
finds that the reduction of the marginal cost of banks brought digital transformation of commercial banks should carry out

Frontiers in Physics 12 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

the reciprocal cooperation while improving their own Author contributions

competitiveness, and avoiding the vicious and homogeneous
competition in the cooperation, so as to promote the mutual KJ: Conceptualization, Formal Analysis, Project administration,
benefit and win-win situation among commercial banks. For Supervision, Writing–review and editing. XL: Data curation, Formal
small and medium-sized banks, they can alleviate the problem of Analysis, Investigation, Methodology, Writing–original draft.
being disadvantaged due to scale and capital constraints in the
process of digital transformation through intra-industry alliance
and cooperation, and enhance their competitiveness level by Funding
embracing the group to increase their attenuating effect on
systemic risks. The author(s) declare that no financial support was received for
Clarifying the path of transformation and adhering to differentiated the research, authorship, and/or publication of this article.
development: At present, the digital transformation of Chinese banks is
still in the exploratory stage. This is a long-term and continuous
improvement process. Commercial banks should give full Conflict of interest
consideration to their own capital scale and research and
development level, formulate differentiated transformation strategies The authors declare that the research was conducted in the
in line with their own development, not blindly follow, and effectively absence of any commercial or financial relationships that could be
respond to the lack of competitiveness in the digital process, so as to construed as a potential conflict of interest.
form personalized services with their own characteristics.
Enhancing risk awareness and innovating risk management: In
the early stages of digital transformation, the disequilibrium between Publisher’s note
input and output may prompt commercial banks to disregard risks
in order to maintain profitability, increase the allocation of risky All claims expressed in this article are solely those of the authors
assets, and affect bank risks. Therefore, commercial banks should and do not necessarily represent those of their affiliated
strengthen their awareness of risk prevention. In the process of organizations, or those of the publisher, the editors and the
digital transformation, we should always keep in mind the priority of reviewers. Any product that may be evaluated in this article, or
risk prevention and control and adhere to digital innovation claim that may be made by its manufacturer, is not guaranteed or
behaviors based on risk prevention and control. endorsed by the publisher.

Data availability statement Supplementary material

Publicly available datasets were analyzed in this study. This data The Supplementary Material for this article can be found online
can be found here: https://choice.eastmoney.com/, https://www. at: https://www.frontiersin.org/articles/10.3389/fphy.2023.1297912/
wind.com.cn/, and https://global.csmar.com/. full#supplementary-material

References
1. Brandl B, Hornuf L. Where did fintechs come from, and where do they go? The 10. Virginia M, Oona V, Emil S. The digital transformation and disruption in business
transformation of the financial industry in Germany after digitalization. Front Artif models of the banks under the impact of fintech and bigtech. Proc Int Conf Business
Intelligence (2020) 3:8. doi:10.3389/frai.2020.00008 Excell (2020) 14:294–305. doi:10.2478/picbe-2020-0028
2. Lee SH, Lee DW. Fintech - conversions of finance industry based on ict. J Korea 11. Benoit S, Colliard J-E, Hurlin C, Pérignon C. Where the risks lie: a survey on
Convergence Soc (2015) 6:97–102. doi:10.15207/JKCS.2015.6.3.097 systemic risk*. Rev Finance (2017) 21(1):109–52. doi:10.1093/rof/rfw026
3. Chen B, Yang X, Ma Z. Fintech and financial risks of systemically important 12. Cincinelli P, Pellini E, Urga G. Systemic risk in the Chinese financial system: a
commercial banks in China: an inverted U-shaped relationship. Sustainability (2022) panel granger causality analysis. Int Rev Financial Anal (2022) 82. doi:10.1016/J.IRFA.
14(10):5912 .doi:10.3390/su14105912 2022.102179
4. Chen MA, Qinxi W, Baozhong Y. How valuable is fintech innovation? Rev 13. Hu D, Zhao S, Yang F. Will fintech development increase commercial banks risk-
Financial Stud (2019)(5) 5. doi:10.1093/rfs/hhy130 taking? Evidence from China. Electron Commerce Res (2022). doi:10.1007/s10660-022-
09538-8
5. Zeng L. Application of financial technology innovation in commercial banks: a case
study of bank of China. Int J New Dev Eng Soc (2021)(3) 5 doi:10.25236/ijndes.2021. 14. Jiang H, Zhang J. Discovering systemic risks of China’s listed banks by covar approach
050306 in the digital economy era. Mathematics (2020) 8(2):180. doi:10.3390/math8020180
6. Furr N, Ozcan P, Eisenhardt KM. What is digital transformation? Core tensions 15. Yang Z, Li D. Research on the systemic financial risk of Chinese banks
facing established companies on the global stage. Glob Strategy J (2022) 12(4): — application analysis based on the "go to one method. Econ Res J (2018) 53(08):
595–618doi:10.1002/gsj.1442 36–51.
7. Halemani SSM Importance of banks in cashless transactions under digitalization 16. Varotto S, Zhao L. Systemic risk and Bank size. J Int Money Finance (2018). doi:10.
system. Journal of Trend in Scientific Research and Development(2018), 188–190. 1016/j.jimonfin.2017.12.002
doi:10.31142/ijtsrd18702
17. Laeven L, Ratnovski L, Tong H. Bank size, capital, and systemic risk: some
8. Cuesta C, Ruesta M, Tuesta D, et al.The digital transformation of the banking international evidence. J Banking Finance (2016) 69:S25–34. doi:10.1016/j.jbankfin.
industry. Oriental Art (2015). doi:10.1109/TMAG.2013.2243424 2015.06.022
9. Thakor AV. Fintech and banking: what do we know? J Financial Intermediation 18. Sun L. Financial networks and systemic risk in China’s banking system. Finance
(2020) 41:100833. doi:10.1016/j.jfi.2019.100833 Res Lett (2020) 34. doi:10.1016/j.frl.2019.07.009

Frontiers in Physics 13 frontiersin.org

Jia and Liu 10.3389/fphy.2023.1297912

19. Mistrulli PE. Assessing financial contagion in the interbank market: maximum 40. Hirata W, Ojima M. Competition and Bank systemic risk: new evidence from
entropy versus observed interbank lending patterns. J Banking Finance (2011) 35(5): Japan’s regional banking. Pacific-Basin Finance J (2020)(C) 60. doi:10.1016/j.pacfin.
1114–27. doi:10.1016/j.jbankfin.2010.09.018 2020.101283
20. Liao M, Sun T, Zhang J. China’s financial interlinkages and implications for inter- 41. Silva-Buston C. Systemic risk and competition revisited. J Banking Finance (2019)
agency coordination. IMF Working Pap (2016) 16(181):1. doi:10.5089/ 101. doi:10.1016/j.jbankfin.2019.02.007
9781475530490.001
42. Xu F, Eva L, Yajun X. Wealth management products, banking competition, and
21. Bellavite Pellegrini C, Cincinelli P, Meoli M, Urga G. The contribution of (shadow) stability: evidence from China. J Econ Dyn Control (2022) 137. doi:10.1016/J.JEDC.
banks and real estate to systemic risk in China. J Financial Stab (2022) 60:101018. 2022.104346
doi:10.1016/j.jfs.2022.101018
43. Faia E, Laffitte S, Ottaviano GIP. Foreign expansion, competition and Bank risk.
22. Cincinelli P, Pellini E, Urga G. Leverage and systemic risk pro-cyclicality in the J Int Econ (2019) 118. doi:10.1016/j.jinteco.2019.01.013
Chinese financial system. Int Rev Financial Anal (2021) 78. doi:10.1016/J.IRFA.2021.
44. Morelli D, Vioto D. Assessing the contribution of China’s financial sectors to
101895
systemic risk. J Financial Stab (2020) 50. doi:10.1016/j.jfs.2020.100777
23. Beltratti A, Stulz RM. The credit crisis around the globe: why did some banks
45. Wei SC, Meng Q, Abbas RSK, Muhammad U. Bank competition in China: a
perform better? Journal of Financial Economics (2012). 105 (1), 1–17. doi:10.1016/j.
blessing or a curse for financial system? Econ Research-Ekonomska Istraživanja (2021)
jfineco.2011.12.005
34(1):1244–64. doi:10.1080/1331677x.2020.1820361
24. Gang J, Qian Z. China’s monetary policy and systemic risk. Emerging Markets
46. Kolodiziev O, Krupka M, Shulga N, Kulchytskyy M, Lozynska O. The level of
Finance & Trade (2015) 51(4):701–13. doi:10.1080/1540496x.2015.1039895
digital transformation affecting the competitiveness of banks. Banks Bank Syst (2021)
25. Yi F, Yanru W, Qi W, Yang Z. Policy uncertainty and Bank systemic risk: a 16(1):81–91. doi:10.21511/bbs.16(1).2021.08
perspective of risk decomposition. J Int Financial Markets, Institutions Money (2023)
47. Summer M. Banking regulation and systemic risk. Open Economies Rev (2003)
88. doi:10.1016/J.INTFIN.2023.101827
14(1):43–70. doi:10.1023/a:1021299202181
26. Xite Y, Qin Z, Haiyue L, Zihan L, Qiufan T, Yongzeng L, et al. Economic policy
48. Yongfang A, Guanglin S, Tao K. Digital finance and stock price crash risk. Int Rev
uncertainty, macroeconomic shocks, and systemic risk: evidence from China. North Am
Econ Finance (2023) 88. doi:10.1016/J.IREF.2023.07.003
J Econ Finance (2024) 69. doi:10.1016/J.NAJEF.2023.102032
49. Tobias B, Valentin B, Ana G, Manju P. On the rise of fintechs: credit scoring using
27. Bakoush M, Gerding EH, Wolfe S. Margin requirements and systemic liquidity
digital footprints. Rev Financial Stud (2020) 33(7):2845–97. doi:10.1093/rfs/hhz099
risk. J Int Financial Markets, Institutions Money (2018) 58:78–95. doi:10.1016/j.intfin.
2018.09.007 50. Guo Y, Liang C. Blockchain application and outlook in the banking industry.
Financial Innovation (2016) 2(1):24. doi:10.1186/s40854-016-0034-9
28. Riedl R, Benlian A, Hess T, Stelzer D, Sikora H. On the relationship between
information management and digitalization. Business Inf Syst Eng (2017) 59(6):475–82. 51. Liu Z. Research on the influence of internet finance on the risk taking of
doi:10.1007/s12599-017-0498-9 commercial banks. Finance Trade Econ (2016)(04) 71–85+115. doi:10.19795/j.cnki.
cn11-1166/f.2016.04.007
29. Cheng M, Qu Y. Does Bank fintech reduce credit risk? Evidence from China.
Pacific-Basin Finance J (2020) 63. doi:10.1016/j.pacfin.2020.101398 52. Xu G, Shi Q. An empirical study on the self-regulation effect of the franchise value
of listed banks in China. China Soft Sci (2009)(01) 20–7+40.
30. Yin F, Jiao X, Zhou J, Yin X, Ibeke E, Iwendi MG, et al. Fintech application on
banking stability using big data of an emerging economy. J Cloud Comput (2022) 11(1): 53. Zhu C. Big data as a governance mechanism. Rev Financial Stud (2019) 32(5):
43. doi:10.1186/s13677-022-00320-7 2021–61. doi:10.1093/rfs/hhy081
31. Li X, Yang P. Financial technology, market power, and banking risk. Mod Econ Sci 54. Adrian T, Brunnermeier MK. Covar. Am Econ Rev (2016) 106(7):1705–41. doi:10.
(2021) 43(01):45–57. 1257/aer.20120555
32. Li C, He S, Tian Y, Sun S, Ning L. Does the bank’s fintech innovation reduce its 55. Acharya VV, Pedersen LH, Philippon T, Richardson M. Measuring systemic risk.
risk-taking? Evidence from China’s banking industry. J Innovation Knowledge (2022) Rev Financial Stud (2017) 30(1):2–47. doi:10.1093/rfs/hhw088
7(3):100219. doi:10.1016/j.jik.2022.100219
56. Brownlees C, Engle RF. Srisk: a conditional capital shortfall measure of systemic
33. Wang R, Liu J, Luo H. Fintech development and Bank risk taking in China. Eur risk. Rev Financial Stud (2017) 30(1):48–79. doi:10.1093/rfs/hhw060
J Finance (2021) 27(4-5):397–418. doi:10.1080/1351847x.2020.1805782
57. Zhang Q, Guo L, Ou Y. Has the digital transformation improved the banks’ ability
34. Wang D, Liu Y, Xu Y, Liu L. Fintech, macroprudential supervision and systemic to serve the real economy? —— based on the economy of China’s listed commercial
risk in China’s banks. China Finance Econ Rev (2022) 11(4). doi:10.1515/CFER-2022- banks verified by the document. Wuhan Finance Monthly (2022)(04) 29–39.
0025
58. Xie X, Wang S. Digital transformation of China Commercial Bank: measurement,
35. Dong X, Wu Z, Chen Q. The impact of fintech development on the risk prevention process and impact. Q J Econ (2022) 22(06):1937–56. doi:10.13821/j.cnki.ceq.2022.06.06
and control of commercial banks —— is based on the demonstration of 176 commercial
59. Angelini P, Cetorelli N. The effects of regulatory reform on competition in the
banks in China analyse. Jiangsu Soc Sci (2023)(01) 84–94. doi:10.13858/j.cnki.cn32-
banking industry. J Money, Credit Banking (2003) 35(5):663–84. doi:10.1353/mcb.2003.
1312/c.20230207.016
0033
36. Cihak M, Wolfe S, Schaeck K. Are more competitive banking systems more stable.
60. Shi Q, Sun X, Jiang Y. Concentrated commonalities and systemic risk in China’s
J Money, Credit Banking (2009) 06:41–4. doi:10.5089/9781451864038.001
banking system: a contagion network approach. Int Rev Financial Anal (2022) 83.
37. Beck T, Demirgüç-Kunt A, Levine R. Bank concentration, competition, and crises: doi:10.1016/J.IRFA.2022.102253
first results. J Banking Finance (2006) 30(5):1581–603. doi:10.1016/j.jbankfin.2005.
61. Denis D, Sami V, Sara Y. Bank liquidity creation and systemic risk. J Banking
05.010
Finance (2021) 123. doi:10.1016/J.JBANKFIN.2020.106031
38. Wibowo IGBE, Wibowo B. The effect of competition levels and banking
62. Tian J, Wang QB. Capital constraints, bank risk spillovers and macrofinancial risk.
concentration on systemic risks: Indonesia’s case. Indonesian Capital Market Rev
Finance Trade Econ (2015)(08) 74–90. doi:10.19795/j.cnki.cn11-1166/f.2015.08.008
(2019) 9(2). doi:10.21002/icmr.v9i2.7138
63. Lim CH, Costa A, Columba F, Kongsamut P, Otani A, Saiyid M, et al.
39. Nguyen TH, Tran HG. Competition, risk and profitability in banking system
Macroprudential policy: what instruments and how to use them? Lessons from
— evidence from vietnam. Singapore Econ Rev (2020) 65(06):1491–505. doi:10.1142/
country experiences. Rochester, NY, USA: Social Science Electronic Publishing (2011).
s0217590820500137

Frontiers in Physics 14 frontiersin.org