J. Int. Financ. Markets Inst.

Money 91 (2024) 101934

Contents lists available at ScienceDirect

Journal of International Financial Markets,

Institutions & Money
journal homepage: www.elsevier.com/locate/intfin

Role of financial inclusion and digital transformation on bank

credit risk
Fan Yang, Tajul Ariffin Masron *
Universiti Sains Malaysia, Malaysia

A R T I C L E I N F O A B S T R A C T

Keywords: As a result of information asymmetry, banks have generated many non-performing loans, jeop­
Digital transformation ardizing financial stability. This paper investigates the relationship between digital trans­
Commercial Bank formation and credit risk and whether financial inclusion plays an interactive role. The paper
Financial Inclusion
employs the Generalized Method of Moments (GMM) and data from 116 Chinese banks from 2014
Credit Risk
GMM
to 2021 to investigate the issue. The results show that digital transformation significantly and
dynamically reduces bank credit risks, and inclusive finance plays an interactive role. As financial
inclusion increases, this risk-reducing effect will increase, meaning that banks need to combine
financial inclusion to promote bank risk reduction better and maximize risk assessment accuracy.

1. Introduction

According to the United Nations’ “World Economic Situation and Prospects 2023″ report, the global economy is expected to reach
1.9 percent, one of the lowest levels seen in a considerable amount of time (United Nations, 2023). Economists have spent the last
hundred years studying the interplay between the financial sector and the economy in great detail (Levine, 1997, 2005). The expansion
of the financial sector beyond a certain point, according to the contention of some economic analysts, may hurt the health of the
economy as a whole (Lucas, 1988). On the other hand, the financial sector is generally acknowledged as the primary driver of economic
expansion (Berger et al., 2020). Bernanke (2007) states that we should always be mindful of the enormous economic benefits of a
robust and innovative financial sector.
However, financial frictions persistently challenge the global economy (Correa et al., 2022; Hasan et al., 2023). In 1929, the global
economy experienced the Great Depression. From the 1980 s to the 1990 s, the United States suffered substantial losses due to high-risk
loans and bad assets. 1997, the Asian financial crisis emerged, followed by the 2001 Argentine financial crisis and the 2007–2008
subprime mortgage crisis in America. Europe continues to grapple with ongoing sovereign debt issues. In 2023, a rapid outflow of
deposits led to the collapse of regional banks, including Silicon Valley Bank (SVB) and Signature Bank in New York. A week later, Swiss
authorities confirmed the merger of Credit Suisse and UBS Group AG (UBS). Excessive risk-taking by banks resulted in financial
instability (Dewatripont & Freixas, 2012; Kirchler et al., 2018; An et al., 2021).
China has not had a financial crisis in 30 years, but its banking-centric financial system has been amassing significant and growing
vulnerabilities (Rong, 2022). According to the “Economic and Financial Outlook Report” released by the Chinese banking industry in
2023, non-performing loans persist (BANK OF CHINA, 2022). According to the theory of enterprise behavior (Olson, 1971), when there
is a large gap between the current business state and the desired future state, companies are more likely to engage in problem-searching

* Corresponding author.
E-mail address: [email protected] (T.A. Masron).

https://doi.org/10.1016/j.intfin.2023.101934
Received 15 July 2023; Accepted 30 December 2023
Available online 3 January 2024
1042-4431/© 2024 Elsevier B.V. All rights reserved.
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

and risk-seeking to achieve selfish ends. Banks’ transition to digital platforms is, at its core, an example of technologically driven
financial innovation (Du & Liu, 2022). The use of digital technology presents only opportunities for achievement. This will have
potential if it is carried out correctly. It will be much more problematic if it is not done correctly (Huang, 2021). In other words, if
digital transformation is in the right direction, it will help reduce bank risks. If the direction is wrong, it will increase the bank’s risks.
New financial services such as digital payment, networks, enormous technology credit, and robot advisory promote banks’ digital
transformation. Taking Big Tech Credit as an example, digital finance aids commercial banks in more effectively detecting credit risks
and reducing exposure to such risks. It enables greater access to risk-related data, thereby diminishing information asymmetry
(Gomber et al., 2018). However, the development of digital finance has driven up banks’ funding costs, narrowing the interest rate
spread between deposits and loans. To maintain profitability levels, banks may opt for higher-risk assets to offset their losses and lower
credit standards to preserve market share, thereby increasing risk-taking behaviors.
Despite indications that digital transformation may favor risk reduction, it is unsuccessful (Fig. 1). Several factors, such as the
economy, market volatility, and the borrower’s creditworthiness, frequently contribute to the development of credit risk. It is not easy
to put the effects of digital transformation only on the rise or fall of credit risk, among other things. Furthermore, it may take some time
for the benefits of digital transformation to become fully apparent. Credit risk events may occur at different times, making observing
and analyzing their connection more difficult. More importantly, banks are easily affected by the external environment when un­
dergoing digital transformation—for example, external financial technology innovation-inclusive finance (Yang & Masron, 2023).
Through digital transformation, banks may enhance their risk detection skills, reach more customers with high-quality financial
services, and better achieve inclusive financing (Neaime & Gaysset, 2018; Le et al., 2019; Abdulquadri et al., 2021; Ozili, 2021; Hakimi
et al., 2022). This is a view generally shared by everyone. The need for inclusive finance, however, could force financial institutions to
undergo more extensive digital transformation (Yang & Masron, 2023). For example, inclusive finance provides a broad customer base
for digital transformation, allowing banks to collect more customer information, better conduct customer profile analysis, and reduce
the risk of default (Guo et al., 2023). As a result, inclusive finance could influence digital transformation, which might cause dynamic
shifts in credit risk. The core concept of this research revolves around this.
China’s financial system has a strong banking sector, and the nation is also among the world’s top innovators in financial tech­
nology. Since lending is at the heart of banks’ operations, it is crucial to comprehend how digital transformation has altered the credit
risk faced by China’s banking sector (Cao et al., 2022). When studying the effects of risk on commercial banks, most available research
divides banks into two categories: internal and external financial technology. Scholars have examined the influence of digital financial
inclusion on bank risk from an external standpoint (Musau et al., 2018; Demirgüç-Kunt et al., 2020) and also have explored the impact
of digital transformation on bank risk from an internal perspective within banks (Metawa et al., 2023).
However, very few researchers have considered the interactive effect of financial inclusion on digital transformation (Guo et al.,
2023). This study helps to close a knowledge gap and contribute to the overall digital transformation research agenda by investigating
the interaction effect of external financial inclusion development on the impact of digital transformation on bank credit risk. The
following outline will serve as a guide for the rest of this paper. The following section discusses the existing literature studies and the
corresponding hypotheses. Section 3 conducts empirical research on methodology. Section 4 explains the model results. Section 5
summarizes the full text.

Fig. 1. Digital Transformation Index (DIG)a vs NPL (in %) Note: a DIG is obtained through the weighted average of the word frequency statistics of
the strategy, business, and management transformation sub-indicators (Xie & Wang, 2023). Source: Peking University Digital Finance Research
Center. https://www.idf.pku.edu.cn/zsbz/index.htm; [email protected].

2
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

2. Literature review

Consumers’ shifting demands for financial services, along with the cutthroat competition from fintech, have forced traditional
banks to embrace digital transformation in order to stay ahead of the curve (Cuesta et al., 2015). Banks’ transition to digital platforms
is, at its core, an example of technologically driven financial innovation (Du & Liu, 2022). The impact of internal digital transformation
on risk in banks currently needs consensus. On the one hand, the “innovation-driven argument” (Pierri & Timmer, 2020; Cappa et al.,
2021). Some studies have found that financial institutions that adopt digital technologies generally have lower default and non-
performing loan rates (Akhter, 2023; Chen et al., 2023; Wang et al., 2023). Theoretically, digital transformation can improve the
risk diversification capabilities of financial institutions. Banks can expand their business scope more easily through digitalization and
enter new markets and product areas, thus diversifying risks (Cao et al., 2022). Business banks use data mining to gather customer,
asset, and credit information. They arrange these search results using machine learning for analysis. This extension of risk data sources
improves data quality, accuracy, customer identification, and information asymmetry (Malladi et al., 2021; Guo & Zhu, 2022). Banks
can extend loans to multiple regions and industries through digital platforms, reducing the risk of economic fluctuations in the specific
regions or industries. At the same time, digital transformation provides more data and analysis tools to help financial institutions better
identify, measure, and manage risks, allowing financial institutions to monitor risks in real-time and take timely measures to deal with
potential risks (Du & Liu, 2022; Guo & Zhu, 2022).
On the other hand, over-innovation leads to “innovation destruction theory.” Firstly, digital transformation is accompanied by
inherent operational risks due to information and communication technology (Mishchenko et al., 2021), such as integrity concerns,
monetary crime risks (such as terrorist funding), financial crime risks (including data breaches and debt in digital financial services),
and privacy and data legislation (Ebong & Babu, 2020). Secondly, banks must be more attuned to risk in the beginning phases of their
digital transition. The fast iterative growth of digitalization runs counter to the old system of commercial banks; as a result, commercial
banks face tremendous cost pressure due to the early stage’s high labor and capital requirements and the lengthy building cycle, and
make risky investments to increase profits in order to compensate for cost depletion (Metawa et al., 2023). Reasonable and interactive
financial innovation may improve risk management. In contrast, excessive financial innovation disregarding its risk-bearing capability
could lead to a crisis in the functioning of commercial banks and, in extreme cases, insolvency (Fecht et al., 2012).
Hypothesis 1. The impact of the digital transformation of commercial banks on credit risk is generally risk-suppressive but can be affected by
certain factors that make the effect less effective.
Financial inclusion gives marginalized or left out of society access to essential financial services and ensures that vulnerable groups
can get enough credit at a price they can afford (Malladi et al., 2021; Rumbogo et al., 2021). On one hand, Ahamed & Mallick (2019)
show that the level of risk that commercial banks bear will decrease as the use of inclusive financing increases. More low-income
individuals and small companies may get the loans they need thanks to inclusive finance, which boosts the economy and keeps
banks afloat (Demirgüç-Kunt et al., 2020; Marcelin et al., 2022). On the other hand, financial inclusion will also affect commercial
bank risk-taking, the rapid development of external financial technology, and the impact of traditional banks, resulting in banks
willing to take more risks to obtain profitability profits, reduce the threshold for credit, and investment in high-risk projects in order to
compete for more significant profits (Hawes & Chitra, 2016). Inclusive financial development adds more poor people to the financial
system (Bourreau & Valletti, 2015), causing people with low incomes to exceed their repayment capacity (Yue et al., 2022) and even
overconsume (Panos & Wilson, 2020), resulting in household debt crises, credit default risks (Yue et al., 2022). The financial literacy of
Chinese household heads is lower than that of developed countries (Feng et al., 2019), and some poor people do not enjoy inclusive
finance, or those who have enjoyed it lack legal knowledge, giving rise to default risks.
As digital transformation matures and government regulatory frameworks improve, the coordinated development of inclusive
finance and digital transformation is expected to mitigate risks in the financial sector. As proposed by Ding (2015), internet finance and
inclusive finance show inherent coupling under the “long tail effect.” At the same time, the financial inclusion genes inherent in
financial technology can help reduce market frictions such as information asymmetry and transaction costs (Hanelt et al., 2021) and
expand inclusive financial coverage (Su & Wei, 2017) and sustainability. Regarding internal digital transformation and external ad­
vancements in financial inclusion, various kinds of banks may have varying degrees of acceptability (Yang & Masron, 2023). Dif­
ferences in the degree of digital transformation may also have different uptake levels of financial inclusion, thereby affecting credit
risk.
Hypothesis 2. Financial inclusion can better adjust the effect of digital transformation on bank credit risk reduction.
Hypothesis 3. There is heterogeneity in the impact of digital transformation on credit risk in the context of financial inclusion.
Regarding the factors that influence bank risk, two components can be identified. One is research into the impact of external factors
such as the macroeconomy, interest rates, and industry competition on bank risk (Laeven & Levine, 2009). The other internal features
of the bank include liquidity, profitability, and bank risk scale (Dell’Ariccia & Marquez, 2010). Raza & Hanif (2013) find that as the
number of banks grows, so does the level of competition in the market and the risk to which each bank is exposed. Starting from the
research on the internal characteristics of banks, Betz et al. (2020) extrapolate that the likelihood of non-performing loans will
decrease if the bank’s profitability is increased and that high capital ratio, high liquidity risk, low credit quality, and occurrence of non-
performing loans are all related to the bank’s profitability. Chiaramonte et al. (2015) claim that the net interest rate on total assets
positively affects commercial banks’ risk-taking capacity. A bank’s shareholders’ equity ratio to its total assets is known as the
shareholder-equity ratio (NA). A high shareholder’s equity ratio may indicate that a bank must effectively leverage its financial

3
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

resources for growth. In contrast, a low ratio may indicate that a bank has taken on excessive debt and needs more net worth to sustain
external shocks.

3. Methodology

3.1. Model specification

According to the long-tail theory proposed by Anderson (2004), long-term accumulation and expansion may cause the market for
low-volume products to become more significant than mainstream products, which benefits competing businesses. Innovation theory
is introduced by Schumpeter & Opie (1934), who highlights the significance of innovation to economic expansion. The notion of
information asymmetry predicts that the financial system would be volatile (Stiglitz & Weiss, 1981). According to the theory of en­
terprise behavior (Olson, 1971), when there is a wide gap between the existing state of the company and the intended future state,
organizations are more prone to participate in issue-finding and risk-seeking. These concepts may provide a solid theoretical grounding
for connecting digital transformation investment and danger. The basic outline of the theory is as follows:
CRISK = f (DIG, DIG*FINDEX, FINC, C) (1)
where CRISK is Credit Risk, DIG is Digital Transformation Index, and C is an indicator of all the other influencing variables. FINC is
the Digital Financial Inclusion Index (China City). DIG*FINC indicates that businesses are not autonomous, and their external envi­
ronment may affect company innovation (Yang & Masron, 2023). However, whether financial inclusion in the external environment
fosters digital growth as it does traditional business innovation has yet to be addressed appropriately. Current research often assumes
technological advancement is consistent across areas. While the digital financial business model described in this article has the
potential to serve customers throughout a wide swath of the globe, its expansion will be influenced by regional differences in de­
mographics, infrastructure, and other geospatial factors (Guo et al., 2020). Using the methodologies outlined in the literature, we will
include an interactive project into the CRISK equation: DIG*FINC. We hypothesize that the interaction item determines the effect on
risk when commercial banks undergo digital transformation. Therefore, we are going to look at it. We get the following model:
CRISK = f (DIG, DIG*FINC, FINC, SIZE, NA, BD, DL, CAR, ROE, HHI) (2)
where SIZE, NA, BD, DL, CAR, ROE, HHI represent that Bank Size, Shareholders Equity, Loan-to-Deposit Ratio, Deposit-to-Loan
Ratio, Capital Adequacy Ratio, Return on Equity, and Herfindahl Index, respectively. The current mainstream bank risk measure­
ment indicators are as follows: Delis & Kouretas (2011) and Guo & Zhu (2021) utilize the bank’s risk-taking as measured by the ratio of
risk-weighted assets to total assets, but keep in mind that this metric is an ex-ante indicator used to assess the capacity and degree to
which a bank can withstand risks in its investments. Papadopoulos (2019) uses the non-performing loan ratio to measure the bank’s
risk-taking ability reversely to describe better the risk borne by the bank. The higher the loan rate, the weaker the risk-taking capacity.
Since the non-performing loan ratio can reflect how much risk assets the bank will absorb when carrying out loan business under the
impact of digital financial development, this article chooses the non-performing loan ratio (NPL) as the proxy variable for bank credit
risk.

NPLi,t = ∂ + β1 DIGi,t + β2 DIGi,t ∗FINDEX i,t + β3 FINDEX i,t + β4 SIZEi,t + β5 NAi,t + β6 BDi,t + β7 DLi,t + β8 CARi,t
(3)
+β9 ROEi,t + β10 HHI i,t + ℷt + ui + εi,t

∂ and βs are parameters to be estimated. i refers to the bank and t denotes time. ℷt and ui stand for controlling time effect and
individual effect. εi,t is a random error item. This study uses a “panel estimation” method to figure out Eq (3). Panel data models are
superior in terms of their ability to predict outcomes accurately (Baltagi, 2008). Phan et al. (2020) develop a dynamic panel model
using the first-order lag items of the proxy variables that assess the performance of commercial banks as explanatory variables. This
results in the model having a more accurate representation of reality. The logarithm reduces the data’s range, making them easier to
compare and visualize. In certain instances, this might help show trends and patterns. DIG and FINC values are in the hundreds, so
logarithms are taken to narrow the gap with other values.
NPLi,t = φ + σNPLi,t− 1 +β1 lnDIGi,t + β2 lnDIGi,t ∗lnFINCi,t + β3 lnFINCi,t (4)

+ β4 SIZEi,t + β5 NAi,t + β6 BDi,t + β7 DLi,t + β8 CARi,t +β9 ROEi,t + β10 HHI i,t + ℷt + ui + εi,t

On the right side of the dynamic panel model are dependent variables whose values have evolved through time. φ are parameters to
be estimated. Where σ is the coefficient of the lagged dependent variable.

3.2. Estimation technique

The Generalized Method of Moments (GMM) regression model will be used in this article. In a number of different economic
models, explanatory variables could be connected with error terms, resulting in endogeneity problems (Subramaniam & Masron,
2021). Therefore, the decisions that are made about digital transformation are impacted by unobservable factors that can have an
effect on credit risk. The Generalized Method of Moments (GMM) is a reliable estimation method that may effectively resolve
endogeneity concerns. If instrumental variables (IV) are included, in that case, it may be possible to overcome the endogeneity

4
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

problem, which would then make it possible to get a more accurate evaluation of the impact that digital transformation has had on
credit risk (Yang & Masron, 2023). The data for this research is in the form of dynamic panel data. Our data collection encompasses
many observation units and time intervals in its entirety. The GMM method is a very reliable way to estimate. It is able to handle single
fixed effects in panel data and allows the explanatory factors to become lagged dependent variables (Yang & Masron, 2023). This
capability will allow it to handle panel data more effectively. When analyzing panel data, GMM models are often effective at over­
coming challenges such as heteroskedasticity and serial correlation (Yang & Masron, 2023). Estimates derived from GMM models
might potentially be more reliable if these considerations are taken into account. The GMM estimator may be implemented in either a
one-step or two-step process, and users have the choice of which method to choose. Because it uses optimum weighting matrices, the
two-step estimator has the potential to provide findings that are more accurate than those produced by the one-step estimator. This is a
theoretical possibility. Because of this, we use a two-stage procedure (Subramaniam & Masron, 2021). The reliability of the GMM
estimator is evaluated using two kinds of specification tests: the Hansen test and the serial correlation test. The instruments are
trustworthy, and the model is correctly characterized if the null hypothesis is not rejected by the Hansen test for over-identifying limits.
Furthermore, first-order serial correlation, often known as AR(1), is tolerated by the GMM estimator. Second-order serial correlation,
or AR(2), is fatal to the GMM estimator’s internal consistency (Subramaniam & Masron, 2021). Systematic GMM two-step regression
was run using Stata 17.

3.3. Data Source

The balanced panel data of 116 commercial banks in China from 2014 to 2021 is used as the study sample in this work. The
commercial bank digital transformation index and inclusive finance index are from the Peking University Financial Research Center.
The digital transformation index of this work constructs an index system using the three commercial bank dimensions (Xie & Wang,
2023). Strategy digitalization index (DIGS), Business digitalization index (DIGB), Management digitalization index (DIGM). DIGS
focuses on the degree to which a bank strategically considers digital technology. DIGB examines the degree to which banks integrate
digital technology into their financial services. DIGM examine how banks use digital technology in their governance frameworks and
organizational management (Xie & Wang, 2023). Other indicator data are from CSMAR databases, the official website of the China
Banking and Insurance Regulatory Commission, the China Money Network Rating Report, and Annual reports from commercial banks.

Table 1
List of variables definition and source.
Variables Definition/measurement Source

NPL Non-performing loan ratio = provision ratio/provision coverage ratio × 100 % CSMAR: https://www.gtarsc.com
DIG The principal component analysis method is used to determine the weight of each Peking University Digital Finance Research Center.
indicator, the linear efficacy function method is used to perform dimensionless [email protected]； [email protected]
processing of the data, and the weighted average is graded from bottom to top. The
total transformation index is obtained through the weighted average of the
transformation sub-indexes (Xie & Wang, 2023).
DIGB Measured through three dimensions: digital channels, digital products, and digital
R&D. Among them, digital channels are measured by whether the bank has launched
mobile banking and WeChat banking that year; digital products are characterized by
the launch of Internet financial management, Internet credit, and e-commerce; digital
research and development is constructed by the application information of banks’
digital technology patents (Xie & Wang, 2023).
DIGM They are measured through three dimensions: digital architecture, digital talents, and
digital cooperation. The digital architecture is measured by whether the bank has
adjusted its organizational structure. Adjustments such as adding an Internet Finance
Department, Digital Finance Department, and FinTech Department are made within
the bank’s organizational structure. Adjustments made outside the bank’s
organizational structure, such as establishing a financial technology subsidiary, are
measured. Digital talent is measured by the proportion of executives and directors with
information technology backgrounds in the bank’s executive team and board of
directors; Digital cooperation is measured by the bank’s cooperation with external
financial technology companies (Xie & Wang, 2023).
DIGS Text learning method (Hassan et al., 2019), a total of 124 keywords were identified in
6 categories: artificial intelligence, blockchain, cloud computing, big data, online, and
mobile (Xie & Wang, 2023).
FINC Peking University Digital Financial Inclusion Index (China City).
DL Loan-to-deposit ratio = Total bank loans/total deposits × 100 %
ROE Return on Equity = net profit/shareholders’ Equity CSMAR / Bank official website/ China Money Network
CAR Capital adequacy ratio = Capital/Risk Assets × 100 % Rating Report: https://www.chinamoney.com.cn/
SIZE The logarithm of total bank assets
NA Shareholder Equity Ratio = Owner’s equity/total assets
BD Loan Ratio = Allowance Balance/Total Loans × 100 %
HHI Using the financial license information of banking institutions from the China Banking State Financial Supervision and Administration Bureau:
Regulatory Commission, we calculated the number of branches of each bank in each https://xkz.cbirc.gov.cn/jr/
city each year. Then, we constructed the Herfindahl-Hirschman Index (HHI) of the
banking industry in each city (Degryse et al., 2009).

5
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

DIG*FINC has been decentralized to prevent multicollinearity problems. For 0 values, 1 has been added to take the logarithm. Table 1
summarizes the variables’ measurements and sources.

4. Empirical results

Descriptive statistics for this investigation’s data are shown in Table 2. As can be seen, there is a wide range of values for the digital
transformation degree index overall. DIG may range from 0 to 197.10; the smallest value is 0, and the highest is 197.10. The range of
DIGS ranges from 0 to 532,90, inclusive. Some banks have gone too far in their digital transformation, while in other areas, it needs to
go farther.
The correlation analysis in Table 3 shows that the correlation between CAR and NA is 0.720. Therefore, this study used the variance
inflation factor (VIF) to evaluate possible multicollinearity among variables. The results showed that the maximum value of VIF was
2.99, which was far less than the critical value of 10. In addition, the DIG, DIGM, and DIGS indicators of the bank’s digital trans­
formation index are negatively correlated, indicating that the digital transformation index specifically reduces credit risks. This ac­
cords with the findings of relevant literature (Ahamed & Mallick, 2019).
Using Eq. (4), we conducted regression analysis employing both GMM (columns (1) to (4)) and FE (columns (5) to (8)) models.
Table 4 presents the results of our analysis.
The interaction between digital transformation and financial inclusion is significant, indicating that their combined effect reduces
credit risk in banks. This is because when digital transformation is combined with financial inclusion, banks have the potential to meet
the diverse needs of their customers and develop loan risk management more effectively. For example, digital technology enables
banks to conduct more precise evaluations of their customers’ credit risk, while financial inclusion initiatives that expand the customer
base reduce the risk of non-performing loans. This dual effect significantly decreases the non-performing loan ratio in these banks. This
finding emphasizes the interdependence of digital transformation and financial inclusion in the banking sector, suggesting potential
synergies for enhancing loan quality and risk management. It also implies that other banks can benefit from aligned investments and
strategies in these domains. Hypothesis 2 is supported. Furthermore, without any moderating variables, a unit increase in DIG reduces
NPLs by 0.279 %.
In this model, the impact of DIG on NPLs is significantly negative, meaning that as DIG increases, NPLs decrease. This has generally
a positive impact as it reduces non-performing loans. However, due to the existence of interaction, the effect of DIG on non-performing
loans depends on the value of FINC. When FINC is at a high value, DIG has a more substantial impact on reducing non-performing loans
than when FINC is at a low value. The diminishing marginal effect means that as FINC gets higher and higher, the impact of DIG on NPL
will slow down. An explanation for this phenomenon is that the growth of financial inclusion leads to increased market saturation. In
saturated financial markets, loans may be extended to borrowers with poor credit or higher credit risk, leading to an overall increase in
credit risk.
Moreover, new customers may need more credit history or complete data as financial inclusion expands. In such cases, digital
transformation models and algorithms may struggle to assess credit risk accurately, exacerbating the risk. Therefore, the limited
effectiveness of digital transformation in reducing credit risk may be attributed to the rapid expansion of financial inclusion. Hy­
pothesis 1 is supported. This finding highlights the trade-offs between digital transformation and financial inclusion in banking,
emphasizing the need for banks to balance these factors carefully in their future efforts.
Let us delve into the regression analysis of the sub-indicator related to digital transformation. The findings indicate statistical
significance in the interaction between business and management digital transformation when considered alongside inclusive finance.
However, it is noteworthy that the interaction term between strategic digital transformation and inclusive finance does not exhibit
statistical significance. This observed phenomenon can be attributed to the distinctive nature of strategic digital transformation within
the context of the banking sector. It involves a comprehensive overhaul of the enterprise, encompassing a holistic transformation of the
bank’s corporate structure, organizational hierarchy, corporate culture, and more.
On the other hand, inclusive finance primarily focuses on driving innovation in business operations and management practices,
ultimately contributing to mitigating bank risks. For instance, business digital transformation initiatives, such as developing mobile

Table 2
List of descriptive statistics.
Variable Obs Mean Std.Dev. Min Max

NPL 928 1.65 0.70 0.05 7.70

DIG 928 74.56 37.05 0.00 197.10
DIGB 928 109.50 42.93 0.00 191.50
DIGM 928 42.46 29.93 0.00 137.40
DIGS 928 117.50 93.71 0.00 532.90
FINC 928 247.10 50.55 113.80 359.70
SIZE 928 26.49 1.54 23.26 31.19
BD 928 3.40 0.95 1.07 8.86
NA 928 0.07 0.02 0.04 0.33
DL 928 69.96 14.29 25.39 124.00
CAR 928 13.34 2.19 7.00 54.09
ROE 928 0.10 0.03 0.05 0.23
HHI 928 0.11 0.04 − 0.07 0.27

6
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

Table 3
Correlation Analysis.
NPL DIG DIGB DIGM DIGS FINC SIZE BD NA DL CAR HHI ROE

NPL 1.00
DIG − 0.02 1.00
DIGB 0.01 0.82 1.00
DIGM − 0.02 0.74 0.41 1.00
DIGS − 0.09 0.68 0.45 0.42 1.00
FINC 0.04 0.61 0.49 0.50 0.41 1.00
SIZE − 0.17 0.57 0.39 0.46 0.50 0.40 1.00
BD 0.54 − 0.06 − 0.05 − 0.01 − 0.12 0.06 − 0.23 1.00
NA 0.12 − 0.11 − 0.14 − 0.08 − 0.09 0.06 − 0.12 0.20 1.00
DL 0.20 0.35 0.27 0.30 0.21 0.52 0.39 0.01 0.22 1.00
CAR − 0.06 − 0.01 − 0.09 0.05 − 0.03 0.15 − 0.03 0.15 0.72 0.04 1.00
HHI − 0.02 − 0.04 − 0.02 − 0.08 0.01 − 0.29 − 0.10 − 0.01 0.02 − 0.17 − 0.02 1.00
ROE − 0.52 − 0.20 − 0.18 − 0.14 − 0.15 − 0.46 0.07 − 0.19 − 0.29 − 0.32 − 0.08 0.11 1.00

Table 4
Regression results of Main Analysis [DV: NPL].
GMM FE

(1) DIG = (2) DIG = (3) DIG = (4) DIG = DIGS (5) DIG = DIG (6) DIG = DIGB (7) DIG = (8) DIG =
DIG DIGB DIGM DIGM DIGS

L.NPL 0.169[2.08] 0.164[1.96] 0.161[2.03] 0.179[2.26]** – – – –

** ** **
lnDIG − 0.279 − 0.329 − 0.118 0.018[0.25] − 0.041[-0.75] 0.010[0.30] − 0.058[-2.50] 0.001[0.06]
[-1.69]* [-1.84]* [-1.54] **
lnDIG*FINC − 0.730 − 0.887 − 0.436 − 0.240[-1.25] − 0.334[-2.15] 0.037[0.32] − 0.280[-3.58] − 0.131[-1.70]
[-2.19]** [-1.77]* [-2.16]** ** *** *
FINC − 1.545 − 1.589 − 1.980 − 1.912[-2.79] − 0.662[-0.72] − 0.125[-0.13] − 0.610[-0.69] − 0.469[-0.52]
[-2.06]** [-1.99]** [-2.67]*** ***
SIZE 0.097[1.68]* 0.103[2.18] 0.111[1.99] 0.051[0.88] − 0.469[-3.19] − 0.472[-3.20] − 0.479[-3.28] − 0.454[-3.09]
** ** *** *** *** ***
BD 0.341[5.38] 0.323[5.18] 0.359[5.15] 0.346[5.31]*** 0.362[15.63] 0.362[15.53] 0.367[15.93] 0.360[15.53]
*** *** *** *** *** *** ***
NA − 3.330 − 2.773 − 3.891 − 3.929[-0.68] − 3.223[-1.47] − 4.272[-1.96] − 3.133[-1.46] − 3.513[-1.62]
[-0.60] [-0.44] [-0.73] *
DL 0.007[1.33] 0.006[0.94] 0.006[1.27] 0.007[1.32] 0.005[2.23]** 0.005[2.12]** 0.006[2.55]** 0.005[2.16]**
CAR − 0.013 − 0.031 − 0.010 − 0.010[-0.27] − 0.022[-1.73] − 0.019[-1.54] − 0.023[-1.82] − 0.022[-1.72]
[-0.39] [-0.63] [-0.29] * * *
ROE − 9.767 − 11.260 − 9.909 − 10.371 − 8.964 − 9.231 − 9.057 − 8.990
[-3.99]*** [-5.44]** [-4.76]*** [-4.65]*** [-12.60]*** [-12.99]*** [-12.87]*** [-12.57]***
HHI 2.926[1.04] 2.327[0.80] 2.146[0.84] 0.964[0.39] − 6.460[-1.10] − 6.881[-1.16] − 5.909[-0.99] − 7.066[-1.20]
Model Criteria
Constant 7.962[2.13] 8.943[2.36] 9.237[2.53] 10.231[3.19] 20.488[7.25] 21.927[7.66] 19.971[7.06] 20.607[7.19]
** ** ** *** *** *** *** ***
# ins 100 100 100 100 – – – –
AR (1) 0.011 0.008 0.009 0.008 – – – –
AR (2) 0.129 0.224 0.198 0.289 – – – –
Hansen 0.17 0.178 0.168 0.272 – – – –
# id 116 116 116 116 116 116 116 116
R2 – – – – 0.529 0.526 0.535 0.528
Marginal Effect
Min − 3.73 − 4.52 − 2.06 No-significant − 1.58 No-significant − 1.38 − 0.62
Mean − 4.30 − 5.22 − 2.40 No-significant − 1.84 No-significant − 1.60 − 0.72
Max − 4.58 − 5.55 − 2.57 No-significant − 1.97 No-significant − 1.71 − 0.77

Note: Asterisks *, **, and*** denote the 10%, 5%, and 1% significance levels, respectively. Figures in [ ] stand for t-statistics values. The values of the
Hansen and AR tests stand for the p-value. The model is estimated using the two-step model with robust estimation. #id refers to the number of
groups, and #inst denotes the number of instruments.

banking applications and online loan platforms, facilitate the extension of banking services to a broader spectrum of customer groups.
This is particularly significant for reaching low-income and marginalized segments of the population who may have yet to access
traditional banking services. Consequently, the bank’s loan portfolio becomes less reliant on a few large customers, diminishing credit
concentration and diversifying credit risks. Moreover, leveraging big data and artificial intelligence technologies allows banks to
conduct more precise assessments of the creditworthiness of these new customer segments, further reducing non-performing loan
ratios. Management digital transformation efforts, including adopting advanced technologies like big data analytics and artificial
intelligence, equip banks with enhanced tools for credit risk management. These technologies enable real-time monitoring of loan

7
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

repayments, early detection of potential defaults, and the implementation of timely corrective measures. Additionally, by analyzing
extensive customer data, banks can identify specific risk factors that may lead to default and adjust their credit policies accordingly.
These strategies empower banks to control credit risks efficiently while broadening their service horizons.
In summary, the interaction effects observed in the regression analysis underscore the interplay between different facets of digital
transformation and inclusive finance in the banking industry. This synergy not only aids in expanding service coverage but also
contributes to effectively managing credit risks. It also highlights the potential benefits for banks in aligning their investments and
strategies in these domains.

4.1. Heterogeneity analysis

Even though the GMM model is utilized in this investigation to investigate the primary connections, we could not conduct a
thorough heterogeneity analysis due to the relatively low number of participants in our sample and the relatively high number of
instrumental variables. To better understand the factors contributing to the observed heterogeneity, we carried out some exploratory
research using the fixed effects model. Even though it is not the optimal option, the fixed effects model may provide helpful insights.
This is especially true when the sample size is restricted, contributing to our early comprehension of possible variances across the
various subgroups. Nevertheless, considering these constraints, when we conduct research in the future, we will increase the sample
size to permit more precise and reliable evaluations of heterogeneity. Following this, we will analyze the heterogeneity of the total
sample concerning the size classification of the total asset and the digital transformation index. This will allow us to determine whether
or not the interaction effects differ between samples if they do exist at all. The categorization is determined by taking the mean of the
total assets and the mean of the digital transformation index.
Table 5 illustrates that within the sub-sample, banks with greater total assets exhibit non-significant base effects of digital trans­
formation on credit risk, although the interaction term is significant. Conversely, smaller banks display significantly adverse base
effects and interaction terms of digital transformation. Due to their increased resources, such as human, technological, and financial
assets, larger banks are often better equipped to handle credit risk. However, internal resistance and complexity can hinder the
implementation of digital transformation, which may affect its effectiveness. As a result, even if larger banks undergo digital trans­
formation, their impact on credit risk may be insignificant. Smaller banks, in contrast, exhibit greater flexibility, are more prone to
implementing and adjusting to digital transformation, and rely more heavily on it to improve their credit risk management.

Table 5
Regression results of Subsample Analysis [DV: NPL].
SIZE-HIGH SIZE-LOW

(1) DIG = DIG (2) DIG = DIGB (3) DIG = (4) DIG = DIGS (5) DIG = DIG (6) DIG = (7) DIG = (8) DIG =
DIGM DIGB DIGM DIGS

lnDIG − 0.049[-0.53] − 0.073[-0.87] − 0.187 0.031[0.67] − 0.112 − 0.036 − 0.044 − 0.022

[-4.22]*** [-2.29]** [-1.25] [-1.95]* [-1.31]
lnDIG*FINC − 0.520[-2.57] − 0.351[-1.54] − 0.639 − 0.049[-0.45] − 0.522 − 0.201 − 0.276 − 0.162
** [-5.92]*** [-4.78]*** [-2.90]*** [-4.68]*** [-2.52]**
FINC − 0.434[-1.34] − 0.432[-1.35] − 0.262 − 0.538[-1.63] 0.035[0.14] − 0.028 0.037[0.15] − 0.060
[-0.86] [-0.11] [-0.24]
SIZE − 0.592[-3.11] − 0.594[-3.09] − 0.601 − 0.594[-3.05] − 0.676 − 0.723 − 0.675 − 0.668
*** *** [-3.27]*** *** [-5.16]*** [-5.45]*** [-5.14]*** [-4.98]***
BD 0.549[14.81] 0.542[14.47] 0.555[15.57] 0.548[14.47] 0.336[13.55] 0.338[13.47] 0.340[13.73] 0.339[13.47]
*** *** ** *** *** *** *** ***
NA − 11.106 − 11.941 − 9.161 − 13.026 − 3.522 − 4.515 − 3.679 − 5.038
[-2.76]*** [-2.94]*** [-2.38]** [-3.23]*** [-1.52] [-1.94]* [-1.59] [-2.17]**
DL 0.004[1.42] 0.003[1.26] 0.005[2.13] 0.003[1.03] 0.007[2.80] 0.007[2.53] 0.007[2.62] 0.006[2.35]
** *** ** *** **
CAR − 0.100[-4.29] − 0.099[-4.25] − 0.099 − 0.099[-4.22] − 0.007 − 0.005 − 0.006 − 0.003
*** *** [-4.45]*** *** [-0.53] [-0.34] [-0.44] [-0.23]
ROE − 11.239 − 11.538 − 11.289 − 11.826 − 8.056 − 8.435 − 8.002 − 8.590
[-11.18]*** [-11.46]*** [-11.59]** [-11.62]*** [-11.08]*** [-11.58]*** [-10.85]*** [-11.74]**
HHI − 9.502[-1.12] − 11.981 − 8.114 − 13.977 − 17.121 − 18.081 − 15.398 − 17.348
[-1.41] [-1.00] [-1.64] [-3.51]*** [-3.63]*** [-3.08]*** [-3.48]***
Model Criteria
Constant 23.107[4.87] 23.655[4.95] 22.510[4.93] 24.051[4.94] 20.488[7.25] 21.927[7.66] 19.971[7.06] 20.607[7.19]
*** *** *** *** *** *** *** ***
# id 42 42 42 42 74 74 74 74
R2 0.737 0.732 0.758 0.729 0.509 0.496 0.507 0.493
Marginal Effect
Min − 2.46 No-significant − 3.21 No-significant − 2.58 − 0.95 − 1.35 − 0.77
Mean − 2.87 No-significant − 3.71 No-significant − 2.99 − 1.11 − 1.56 − 0.89
Max − 3.06 No-significant − 3.95 No-significant − 3.19 − 1.18 − 1.67 − 0.95

Note: Asterisks *, **, and*** denote the 10%, 5%, and 1% significance levels, respectively. Figures in [ ] stand for t-statistics values. The values of the
Hansen and AR tests stand for the p-value. The model is estimated using the two-step model with robust estimation.

8
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

For the sub-indicators of the overall digital transformation index (DIGS, DIGB, DIGM), banks with a higher total asset size exhibit an
interaction with financial inclusion in management digitalization but do not show such interaction in business and strategic digitali­
zation. On the other hand, banks with smaller total assets exhibit interactions with financial inclusion across all sub-indicators. Banks
with considerable total assets often possess more resources and capabilities to engage in digital transformation, especially at the man­
agement level. This may include investments in advanced technologies, artificial intelligence, data analytics, and more to enhance their
operational efficiency and decision-making capabilities. Such transformation enables banks to better serve financial inclusion by offering
more personalized and convenient services or making data-driven decisions to mitigate risks. Business and strategic digitalization,
however, do not interact with financial inclusion in larger banks. These banks may be more inclined to serve large enterprises or high-net-
worth individuals rather than the target financial inclusion groups, such as small and micro-enterprises or low-income individuals.
Financial inclusion may not directly influence their digitalization efforts at the business and strategic levels. In contrast, smaller
banks with smaller total assets may rely more on financial inclusion to expand their market share. Therefore, their digitalization efforts
at the business, management, and strategic levels are likely to be directly impacted by financial inclusion.
Table 6 analyses how the interaction of financial inclusion affects banks with high and low digital transformation indexes. The
analysis shows that for banks with a high digital transformation index, the credit risk reduction is not significant, the credit risk
reduction by financial inclusion is significant, and its interaction term is significant. However, for banks with a low digital trans­
formation index, the reduction of credit risk is significant, the reduction of credit risk by financial inclusion is not significant, and its
interaction term is significant. This is because banks with a high digital transformation index are already doing an excellent job in
digital transformation. There may be a marginal diminishing effect as financial inclusion increases.
On the other hand, financial inclusion as an additional strategy may still significantly reduce credit risk because it targets different
factors or customer groups. Thus, there may be a substitution effect between digital transformation and financial inclusion in banks
with high digital transformation indices, especially where digital transformation is already relatively complete. Financial inclusion
may be a critical factor in further improving credit quality. Banks with a low digital transformation index may have more room for
improvement due to the low level of digital transformation. Therefore, the digital transformation itself reduces credit risk significantly.
Digital transformation may have improved lending processes, customer data analytics, and risk assessment, thus improving loan
quality. The impact of financial inclusion on credit risk is insignificant because the low level of digital transformation limits the
implementation of financial inclusion measures. However, the interaction term between the two is significant, suggesting that the
synergy between digital transformation and financial inclusion can still significantly reduce credit risk among banks with low digital
transformation indices.

Table 6
Regression results of Subsample Analysis [DV: NPL].
DIG-HIGH DIG-LOW

(1) DIG = (2) DIG = (3) DIG = (4) DIG = DIGS (5) DIG = DIG (6) DIG = (7) DIG = (8) DIG = DIGS
DIG DIGB DIGM DIGB DIGM

lnDIG − 0.038 0.040[0.40] − 0.112 − 0.058[-1.96]* − 0.103 − 0.036 − 0.050 − 0.012[-0.61]

[-0.45] [-2.94]*** [-1.97]** [-1.30] [-2.04]**
lnDIG*FINC − 0.636 − 0.526[-2.10] − 0.450 − 0.127[-1.30] − 0.503 − 0.185 − 0.26[-3.80] − 0.109[-1.38]
[-3.51]*** ** [-4.85]*** [-3.83]*** [-2.50]** 3***
FINC − 0.760 − 0.778[-2.86] − 0.710 − 0.647[-2.32] 0.125[0.42] 0.053[0.18] 0.135[0.47] − 0.006[-0.02]
[-2.74]*** *** [-2.65]*** **
SIZE − 0.502 − 0.562[-3.53] − 0.451 − 0.509[-3.20] − 0.689 − 0.711 − 0.690 − 0.663[-4.18]
[-3.18]*** *** [-2.88]*** *** [-4.45]*** [-4.54]*** [-4.46]*** ***
BD 0.462[15.03] 0.456[14.80] 0.460[15.13] 0.455[14.62] 0.357[12.61] 0.361[12.63] 0.364[12.92] 0.362[12.57]
** *** *** *** *** *** *** ***
NA − 9.300 − 10.200 − 9.063 − 11.397 − 2.915 − 3.817 − 3.106 − 4.478[-1.61]
[-3.22]*** [-3.55]*** [-3.20]*** [-3.96]*** [-1.06] [-1.38] [-1.13]
DL 0.004[2.01] 0.004[1.83]* 0.005[2.37]** 0.004[1.71]* 0.010[2.96] 0.009[2.75] 0.010[2.96] 0.009[2.56]**
** *** *** ***
CAR − 0.068 − 0.070[-3.69] − 0.066 − 0.065[-3.44] − 0.011 − 0.007 − 0.009 − 0.004[-0.26]
[-3.61]*** *** [-3.53]*** *** [-0.65] [-0.44] [-0.56]
ROE − 10.137 − 10.334 − 10.256 − 10.657 − 8.234 − 8.339 − 8.156 − 8.544[-9.37]
[-2.58]** [-12.94]*** [-12.97]** [-13.40]*** [-9.21]*** [-9.22]*** [-9.04]*** ***
HHI − 21.399 − 24.163 − 20.701 –22.399[-3.45] − 16.129 − 17.544 − 15.333 − 16.807
[-3.33]** [-3.73]*** [-3.25]*** *** [-2.87]*** [-3.08]*** [-2.65]*** [-2.94]***
Model Criteria
Constant 22.617[6.08] 24.423[6.54] 21.062[5.70] 22.626[6.02] 19.898[5.94] 20.782[6.12] 19.485[5.84] 19.761[5.81]
*** *** *** *** *** *** *** ***
# id 63 63 63 63 53 53 53 53
R2 0.607 0.605 0.616 0.599 0.576 0.566 0.576 0.561
Marginal Effect
Min − 3.00 − 2.49 − 2.24 No-significant − 2.48 − 0.88 − 1.28 No-significant
Mean − 3.50 − 2.90 − 2.59 No-significant − 2.87 − 1.02 − 1.48 No-significant
Max − 3.75 − 3.10 − 2.76 No-significant − 3.07 − 1.09 − 1.58 No-significant

Note: Asterisks *, **, and*** denote the 10%, 5%, and 1% significance levels, respectively. Figures in [ ] stand for t-statistic values. The values of the
Hansen and AR tests stand for the p-value. The model is estimated using the two-step model with robust estimation

9
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

After heterogeneity analysis, it is found that the interaction of digital transformation and financial inclusion has been playing a
significant role. However, there is some heterogeneity in the underlying effects of digital transformation on credit risk. Hypothesis 3
holds.

5. Conclusion and policy implication

Commercial banks have been given a boost in their efforts to promote digital transformation and broaden the long-tail market due
to the integration and penetration of digital technology and inclusive financing, which have offered effective methods to address pain
problems. To analyze the effect of digital transformation on Chinese commercial banks’ credit risk, we gathered data from 116 in­
stitutions throughout China between 2014 and 2021. First, the digital transformation of commercial banks has a significant negative
impact on bank credit risk in the interaction of financial inclusion. It reduces bank credit risk. Second, the interaction of digital
transformation and financial inclusion always plays a role, but there is heterogeneity in the underlying impact of digital transformation
on credit risk. To be clear, our study is based on observed correlations, not causation. There are other factors not considered that may
affect both digital transformation and credit risk, so we need more research to verify whether this relationship is causal. We recom­
mend that future research conduct more in-depth experimental studies to verify a causal relationship between digital transformation
and credit risk to better guide policy and practice.
The following policy suggestions are made to aid in expanding financial institutions: Even though most central banks are digital,
financial inclusion may still provide them with credit quality benefits. Utilizing their resource advantages, these banks could provide
more specialized and accessible services to low-income and excluded communities. Small banks can increase competitiveness and
improve credit risk management through digital transformation. The government and other organizations should provide financial and
technical assistance to these banks to facilitate their digital transformation. The management of credit risk should place a greater
emphasis on financial inclusion and digital transformation. Policymakers should encourage banks, technological companies, and social
organizations to collaborate. In light of the rapid expansion of digital transformation and financial inclusion, governments and reg­
ulators should establish feedback systems to evaluate the effectiveness of pertinent policies and make any necessary adjustments.

CRediT authorship contribution statement

Fan Yang: . Tajul Ariffin Masron: Conceptualization, Methodology, Software, Investigation, Formal analysis, Writing – original
draft.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data availability

Data will be made available on request.

Acknowledgement

We wish to thank all reviewers involved in review process that help to improve the quality of this paper.

References

Abdulquadri, A., Mogaji, E., Kieu, T.A., Nguyen, N.P., 2021. Digital transformation in financial services provision: A Nigerian perspective to the adoption of chatbot.
Journal of Enterprising Communities: People and Places in the Global Economy. 15 (2), 258–281.
Ahamed, M.M., Mallick, S.K., 2019. Is financial inclusion good for bank stability? International evidence. J. Econ. Behav. Organ. 157 (January), 403–427.
Akhter, N., 2023. Determinants of commercial bank’s non-performing loans in Bangladesh: An empirical evidence. Cogent Economics & Finance. 11 (1), 2194128.
An, J., Jiang, M., Xu, J., 2021. Professional norms and risk-taking of bank employees: Do expectations of peers’ risk preferences matter? J. Financ. Stab. 56 (October),
100938.
Anderson, C. (2004). The long tail. Wired. Outubro. Available: www.wired.com/wired/archive/12.10/tail.html.
Baltagi, B.H., 2008. Forecasting with panel data. J. Forecast. 27 (2), 153–173.
BANK OF CHINA. (2022). 2023 Economic and Financial Outlook Report.
Berger, A.N., Molyneux, P., Wilson, J.O., 2020. Banks and the real economy: An assessment of the research. Finance 62 (June), 101513.
Bernanke, B., 2007. Regulation and financial innovation. The Speech of Board of Governors of the Federal Reserve System (US). Retrieved in January 2022 from https://
www.federalreserve.gov/newsevents/speech/bernanke20070515a.htm.
Betz, J., Krüger, S., Kellner, R., Rösch, D., 2020. Macroeconomic effects and frailties in the resolution of non-performing loans. J. Bank. Financ. 112 (March), 105212.
Bourreau, M., Valletti, T., 2015. Enabling digital financial inclusion through improvements in competition and interoperability: What works and what doesn’t. CGD
Policy Paper 065, Washington DC: Center for Global Development.
Cao, X., Han, B., Huang, Y., Xie, X., 2022. Digital Transformation and Risk Differentiation in the Banking Industry: Evidence from Chinese Commercial Banks. Asian
Economic Papers. 21 (3), 1–21.
Cappa, F., Oriani, R., Peruffo, E., McCarthy, I., 2021. Big data for creating and capturing value in the digitalized environment: Unpacking the effects of volume,
variety, and veracity on firm performance. J. Prod. Innov. Manag. 38 (1), 49–67.
Chen, Z., Li, H., Wang, T., Wu, J., 2023. How digital transformation affects bank risk: Evidence from listed Chinese banks. Finance Research Letters. 58 (Part A,
December), 104319.

10
F. Yang and T.A. Masron Journal of International Financial Markets, Institutions & Money 91 (2024) 101934

Chiaramonte, L., Poli, F., Oriani, M.E., 2015. Are cooperative banks a lever for promoting bank stability? Evidence from the recent financial crisis in OECD countries.
Eur. Financ. Manag. 21 (3), 491–523.
Correa, R., Paligorova, T., Sapriza, H., Zlate, A., 2022. Cross-border bank flows and monetary policy. Rev. Financ. Stud. 35 (1), 438–481.
Cuesta, C., Ruesta, M., Tuesta, D., Urbiola, P., 2015. The digital transformation of the banking industry. BBVA Research Digital Watch.
Degryse, H., Kim, M., & Ongena, S. (2009). Microeconometrics of banking: Methods, applications, and results. Oxford University Press, USA.
Delis, M.D., Kouretas, G.P., 2011. Interest rates and bank risk-taking. J. Bank. Financ. 35 (4), 840–855.
Dell’Ariccia, G., Marquez, R., 2010. Risk and the corporate structure of banks. J. Financ. 65 (3), 1075–1096.
Demirgüç-Kunt, A., Klapper, L., Singer, D., Ansar, S., Hess, J., 2020. The Global Findex Database 2017: Measuring financial inclusion and opportunities to expand
access to and use of financial services. World Bank Econ. Rev. 34 (Supplement_1), S2–S8.
Dewatripont, M., Freixas, X., 2012. The crisis aftermath: New regulatory paradigms. Center for Economic Policy Research: London.
Ding, J., 2015. Paradox of Theory and Practice of Internet Finance and Inclusive Finance. Finance & Economics. 6, 1–10.
Du, L., Liu, Z., 2022. The Impact of Digital Finance on Commercial Banks’ Credit Risk and Operating Efficiency. Studies of International Finance. 6, 75–85.
Ebong, J., Babu, G., 2020. Demand for credit in high-density markets in Kampala: Application of digital lending and implication for product innovation. Journal of
International Studies. 13 (4), 295–313.
Fecht, F., Grüner, H.P., Hartmann, P., 2012. Financial integration, specialization, and systemic risk. J. Int. Econ. 88 (1), 150–161.
Feng, X., Lu, B., Song, X., Ma, S., 2019. Financial literacy and household finances: A Bayesian two-part latent variable modeling approach. J. Empir. Financ. 51
(March), 119–137.
Gomber, P., Kauffman, R.J., Parker, C., Weber, B.W., 2018. On the fintech revolution: Interpreting the forces of innovation, disruption, and transformation in financial
services. J. Manag. Inf. Syst. 35 (1), 220–265.
Guo, F., Kong, T., Wang, Q., 2020. Analysis of Internet Finance Spatial Agglomeration Effect: Evidence from Internet Finance Development Index. Q. J. Econ. 19 (4),
1401–1418.
Guo, L., Zhu, K., 2022. Fintech, Bank Risks, and Business Performance: From the Perspective of Inclusive Finance. China Economic Transition= Dangdai Zhongguo
Jingji Zhuanxing Yanjiu. 5 (2), 242–261.
Guo, F., Zhuang, X., Wang, R., 2023. Bank Digital Transformation, Exogenous Fintech and Credit Risk Governance: Empirical Test Based on Text Mining and Machine
Learning. Secur. Mark. Her. 4, 15–23.
Hakimi, A., Boussaada, R., Karmani, M., 2022. Are financial inclusion and bank stability friends or enemies? Evidence from MENA Banks. Applied Economics. 54 (21),
2473–2489.
Hanelt, A., Bohnsack, R., Marz, D., Antunes Marante, C., 2021. A Systematic Review of the Literature on Digital Transformation: Insights and Implications for Strategy
and Organizational Change. J. Manag. Stud. 58 (5), 1159–1197.
Hasan, M.F., Hadi, M.R., Flayyih, H.H., Al-Obaidi, W.K.O., 2023. A Comprehensive Review of Financial Frictions: Causes, Consequences, and Policy Implications.
Ishtar Journal of Economics and Business Studies. 4 (2), 1–13.
Hassan, T.A., Hollander, S., Van Lent, L., Tahoun, A., 2019. Firm-level political risk: Measurement and effects. Q. J. Econ. 134 (4), 2135–2202.
Hawes, A., Chitra, T., 2016. Keeping up with the FinTechs: A banker’s strategy for digital transformation. Journal of Digital Banking. 1 (2), 102–110.
Huang, Y., 2021. Several views on China’s digital financial innovation and development. Financial Forum. 26 (11), 3–5.
Kirchler, M., Lindner, F., Weitzel, U., 2018. Rankings and risk-taking in the finance industry. J. Financ. 73 (5), 2271–2302.
Laeven, L., Levine, R., 2009. Bank governance, regulation and risk taking. Journal of Financial Economics. 93 (2), 259–275.
Le, T.-H., Chuc, A.T., Taghizadeh-Hesary, F., 2019. Financial inclusion and its impact on financial efficiency and sustainability: Empirical evidence from Asia. Borsa
Istanbul Rev. 19 (4), 310–322.
Levine, R., 1997. Financial development and economic growth: Views and agenda. J. Econ. Lit. 35 (2), 688–726.
Levine, R., 2005. Finance and growth: Theory and evidence. Handbook of Economic Growth. 1, 865–934.
Lucas Jr, R.E., 1988. On the mechanics of economic development. J. Monet. Econ. 22 (1), 3–42.
Malladi, C.M., Soni, R.K., Srinivasan, S., 2021. Digital financial inclusion: Next frontiers—Challenges and opportunities. CSI Transactions on ICT. 9 (2), 127–134.
Marcelin, I., Sun, W., Teclezion, M., Junarsin, E., 2022. Financial inclusion and bank risk-taking: The effect of information sharing. Financ. Res. Lett. 50 (December),
103182.
Metawa, N., Itani, R., Metawa, S., Elgayar, A., 2023. The impact of digitalization on credit risk: The mediating role of financial inclusion (National Bank of Egypt
(NBE) case study). Economic Research-Ekonomska Istraživanja. 36 (2), 2178018.
Mishchenko, S., Naumenkova, S., Mishchenko, V., Dorofeiev, D., 2021. Innovation risk management in financial institutions. Investment Management and Financial
Innovations. 18 (1), 191–203.
Musau, S., Muathe, S., Mwangi, L., 2018. Financial inclusion, GDP and credit risk of commercial banks in Kenya. Int. J. Econ. Financ. 10 (3), 181–195.
Neaime, S., Gaysset, I., 2018. Financial inclusion and stability in MENA: Evidence from poverty and inequality. Financ. Res. Lett. 24 (March), 230–237.
Olson Jr, M., 1971. The Logic of Collective Action: Public Goods and the Theory of Groups, with a New Preface and Appendix, Vol. 124. Harvard University Press.
Ozili, P.K., 2021. Financial inclusion: globally important determinants. Financial Internet Quarterly. 17 (4), 1–11.
Panos, G.A., Wilson, J.O., 2020. Financial literacy and responsible finance in the FinTech era: Capabilities and challenges. Eur. J. Financ. 26 (4–5), 297–301.
Papadopoulos, G., 2019. Income inequality, consumption, credit and credit risk in a data-driven agent-based model. J. Econ. Dyn. Control 104 (July), 39–73.
Phan, D.H.B., Narayan, P.K., Rahman, R.E., Hutabarat, A.R., 2020. Do financial technology firms influence bank performance? Pac. Basin Financ. J. 62 September),
101210.
Pierri, M. N., Timmer, M. Y., 2020. Tech in Fin before FinTech: Blessing or Curse for Financial Stability? International Monetary Fund Working Paper Series WP/20/
14.
Raza, S.A., Hanif, N., 2013. Factors affecting internet banking adoption among internal and external customers: A case of Pakistan. International Journal of Electronic
Finance. 7 (1), 82–96.
Rong, Y., 2022. Systemic Risk in Complex Banking Networks [PhD thesis, Nankai University].
Rumbogo, T., McCann, P., Hermes, N., Venhorst, V., 2021. Financial Inclusion and Inclusive Development in Indonesia. In Holzhacker, R. L., Tan, W. G. Z., (Eds.),
Challenges of Governance: Development and Regional Integration in Southeast Asia and ASEAN, 161–181. Springer International Publishing.
Schumpeter, J.A., Opie, R., 1934. The theory of economic development: An inquiry into profits, capital, credit, interest, and the business cycle. Harvard University
Press.
Stiglitz, J.E., Weiss, A., 1981. Credit rationing in markets with imperfect information. Am. Econ. Rev. 71 (3), 393–410.
Subramaniam, Y., Masron, T.A., 2021. Food security and environmental degradation: Evidence from developing countries. GeoJournal 86 (3), 1141–1153.
United Nations, 2023. World Economic Situation and Prospects: February 2023 Briefing, No. 169 | Department of Economic and Social Affairs.
Wang, H., Mao, K., Wu, W., Luo, H., 2023. Fintech inputs, non-performing loans risk reduction and bank performance improvement. Int. Rev. Financ. Anal. 90
(November), 102849.
Xie, X., Wang, S., 2023. Digital transformation of commercial banks in China: Measurement, progress and impact. China Economic Quarterly International. 3 (1),
35–45.
Yang, F., Masron, T.A., 2023. Does financial inclusion moderate the effect of digital transformation on banks’ performance in China? Cogent Economics & Finance. 11
(2), 2267270.
Yue, P., Korkmaz, A. G., Yin, Z., Zhou, H., 2022. The rise of digital finance: Financial inclusion or debt trap? Finance Research Letters. 47 (Part A, June), 102604.

11