Fintech and the economic capital of chinese commercial banks risk based on theory and evidence

In August 2019, the People's Bank of China issued a plan for the development of FinTech in the next 3 years, raising the application of Fintech in the field of financial risk management

R E S E A R C H A R T I C L E

Fintech and the economic capital of Chinese commercial bank's risk: Based on theory and evidence

1 Business School, University of Shanghai

for Science and Technology, Shanghai,

China

2 Guangxi University Xingjian College of

Science and Liberal Arts, Nanning, China

3 Business School, University of Shanghai

for Science and Technology, Shanghai,

China

Correspondence

Ting Yao, Business School, University of

Shanghai for Science and Technology,

516 Jungong Road, Shanghai 200093,

China.

Email: yaoting_99@163.com

Funding information

National Natural Science Foundation of

China, Grant/Award Number: 71871144

Abstract

differ-ent sizes of banks economic capital through the application of Fintech perspec-tive in China during the period January 2011 and September 2019, using a dynamic panel generalized method of moments (GMM) estimation technique The study found compared with small and medium-sized banks, large state-owned commercial banks have advantages in scale, capital and experience There is a negative correlation between the scale of assets of commercial banks and economic capital Further tests reveal the impact of Fintech on the profit-ability of different types of commercial banks shows significant heterogeneity

K E Y W O R D S economic capital, Fintech, Fintech application index, GMM, risk management

1 | I N T R O D U C T I O N

government work report In August 2019, the People's

Bank of China issued a plan for the development of

FinTech in the next 3 years, raising the application of

Fintech in the field of financial risk management to an

unprecedented level In December 2019, the People's

Bank of China announced the launch of a Fintech

innovation supervision pilot Beijing took the lead in

launching the Chinese version of the regulatory

model, which shows that the Chinese government

attaches great importance to the status of Fintech in

the field of financial risk control As the main body of

China's financial system, commercial banks need to

seize the opportunity of the development of Fintech to

achieve self-upgrading and transformation However,

few directly focus on how Fintech affects banks' risk

The changes in the economic capital of commercial

bank's risk under the impact of Fintech This study

aims to fill this gap in the literature We also

investi-gate whether Fintech will have different effects on

banks of different sizes to present new conclusions on these differences

Fintech (Milian, de Spinola, & de Carvalho, 2019; Puschmann, 2017; Thakor, 2020) combines the

Fintech ecosystem, using emerging technologies such as robotic investment advisors, distributed accounting, face recognition and brainwave payment, etc., innovated vari-ous new models, application scenarios, processes, new products and rich new visual experiences (Pi, Liu, &

Wu, 2018), breaking the existing industry structure and blurring the industry boundaries (Shim & Shin, 2016) Fintech can provide convenient access to services and cost reduction (Li, Spigt, & Swinkels, 2017; Vasiljeva & Lukanova, 2016) Fintech can also fill the original finan-cial industry gap (Truong, 2016), but it also poses

(Philippon, 2016) The essence of economic capital is funds to make up for unexpected losses It directly reflects the overall risk of the bank and is an effective tool for banks to conduct comprehensive risk management Economic capital is the amount of capital a bank needs

DOI: 10.1002/ijfe.2528

Int J Fin Econ 2021;1–15 wileyonlinelibrary.com/journal/ijfe © 2021 John Wiley & Sons, Ltd 1

to resist business risks It is not the bank's own capital or

actual capital, but just a kind of capital demand

Eco-nomic capital is a variable, which changes in the same

direction as the risks faced by the bank That is, when

the risk is high, the demand for economic capital is high;

when the risk decreases, the demand for economic

capi-tal also decreases

change the original profit model of commercial banks

through the comprehensive three-dimensional

integra-tion between Fintech and commercial banks The data is

the core of operation Fintech can provide

diversified-dimensional risk operation innovative ideas for

commer-cial banks The impact of the application of Fintech on

the risks of commercial banks can be summarized in two

aspects: (a) The application of Fintech has reduced the

cost of information asymmetry and improved the bank's

profitability and risk-taking Fintech uses advanced

tech-nologies (predictive algorithms, distributed accounting,

visual recognition, natural language processing, etc.) to

provide commercial banks with new solutions to the

problems of information asymmetry, the need to expand

the original data capacity and data quality is low Fintech

cuts emerging technologies into the field of financial

ser-vices, connects the asset side of the financial industry

with other links, accurately matches the capital side and

the asset side, realizes efficient pricing, service and

prod-uct innovation and reduces operating costs and achieves

reasonable allocation of funds (Gong, Yang, & Qu, 2017;

Yang, 2018; Zhang & Jiang, 2018; Zhou & Li, 2016) The

financial sector is shifting towards smart finance,

conve-nience and efficiency Fintech can not only provide

con-venient financial services but also fill the gaps in the

retail customers and low-end customers service areas of

traditional banks Fintech analyzes massive data sets to

measure, track and describe consumer behaviour

pat-terns and can provide customers with personalized

solu-tions (Liao, 2016; Qi & Xiao, 2018; Thakor, 2020;

Yi, 2017) (b) Fintech makes various market risks

con-stantly superimposed, and commercial banks have to face

sudden changes Once a decision error is made, the risk

of commercial bank bankruptcy will increase Ba and

Bai (2016) found that with the diversification of

partici-pants in the financial sector (including traditional

undergone transformation and upgrading, etc.), the

ser-vice areas have also expanded to multiple areas,

includ-ing traditional credit, deposit and fundraisinclud-ing, bank

intermediary business, but also investment management

services and insurance, etc.(Thakor, 2020; Wang, 2018),

diversified participants and service areas make risks

com-plex and diverse The potential risks of Fintech

applica-tion are divided into two aspects: macro and micro The

micro includes financial institutions' process, network, legal risks, etc.; The macro includes risks such as

Jiang, 2017; Liu et al., 2017) The continuous develop-ment and evolution of risks has made it more difficult for commercial banks to predict, evaluate and measure risks Once commercial banks are not managed well, they are more likely to face bankruptcy risks Fintech is a product

of the continuous development and evolution of commer-cial banks, and it has a counterproductive effect on com-mercial banks Correctly grasping the law of risk evolution of commercial banks needs to be based on a deep understanding of Fintech Through the perspective

of Fintech application changes, clarify the internal logic and important issues of China's Fintech operation and commercial bank risks This will be of great significance for improving the commercial banks risk management and promoting the healthy development of China's finan-cial system

In response to the above issues, this article makes marginal contributions in the following points: First, the Fintech is incorporated into the analysis framework of the comprehensive risk management of commercial banks, and the understanding of the comprehensive risk management of commercial banks is expanded from the macro level, and the degree of Fintech application is used

to explain the internal logic and evolutionary mechanism

of the comprehensive risk of commercial banks and Fintech have made the mechanism of Fintech's impact

on the overall risk of commercial banks clearer, which will help to provide a theoretical basis for the compre-hensive risk management of commercial banks in the future; second, the application of Fintech in various fields

of the banking industry, combined with the data pro-vided by the regular reports published by commercial banks, and the use of text mining technology and princi-pal component analysis to innovatively construct a Fintech application index Since the index is constructed

by combining the data regularly announced by commer-cial banks, the index is more relevant to the actual situa-tion of commercial banks, and it is easier to obtain real data It is also easy to operate and implement flexibly according to the commercial banks' own conditions The quantitative analysis of economic capital required for comprehensive risk management of the bank industry is more accurate, enriching the relevant literature on the quantitative analysis of Fintech; third, studying the degree of Fintech application to commercial banks' risk tolerance is helpful and examines the changes in com-mercial bank risks under the background of Fintech from

a new perspective, which has special policy significance for the design of banking supervision and policy, and pro-vides micro-empirical evidence for commercial banks to

formulate reasonable response measures for

comprehen-sive risk management

The remaining part is arranged as follows: The second

part explains the internal logic and evolution of the

Fintech application level and commercial bank risk; The

third part draws on the text mining method and the

prin-cipal component analysis method to construct the

Fintech application index and empirically analyzes the

impact of the Fintech application on the business bank

risk-taking and the heterogeneity of different types of

corresponding improvement policy suggestions

2 | L I T E R A T U R E R E V I E W ,

T H E O R Y A N D H Y P O T H E S I S

There is an intensive debate about the essence of Fintech

and the economic capital of commercial bank risk (Ma &

Zhang, 2019) The internal logic and evolution

mecha-nism of Fintech application and commercial bank risk

can be summarized as follows:

(1) The application of Fintech improves commercial

banks' risk-taking and changes profit models

Fintech can optimize banking business processes and

reduces overall risk levels (Xue & Hu, 2020) In terms of

credit granting and rating to customers: Fintech can use

data mining, machine learning, neural network and

other advanced technologies to provide data sources

references for bank credit and solve non-linear problems

that cannot be handled by traditional rating models and

improve commercial banks' credit granting capabilities

In terms of market risk management process: Fintech

can improve the accuracy of capital position forecasts,

strengthen capital operation efficiency, analyse customer

risks in a timely manner and track reports In terms of

operational risk process management, Fintech can use

advanced technologies such as biometrics, voice

recogni-tion and intelligent robots to reduce manpower, capital

and time costs, improve data accuracy and reduce

inter-nal personnel fraud risks and systemic risks (Fuster,

Plosser, Schnabl, & Vickery, 2019) In addition, in terms

of finance inclusive: Fintech's blockchain technology,

digital intelligence and supply chain finance can provide

precise risk control and supply chain financing services

in the loan process of small and micro enterprises (SMEs)

and achieve professional and real-time monitoring and

risk-tracking service (Blythin & Cooten, 2017; Basel

Com-mittee on Banking Supervision, 2018; TSAI, 2017;

Loukoianova & Yang, 2018; Shen, Hueng, & Hu, 2020;

Jin, Li, & Liu, 2020; Hasan, Lu, & Mahmud, 2020)

Fintech overthrows the original profit model of

banks The original profitability of commercial banks

mainly relied on the spread between loans and deposits Under the reform of interest rate marketization, the prof-itability of commercial banks has declined, and commer-cial banks urgently need to change the original single profit model In 2018, global investment in Fintech reached US$55.3 billion, of which China's participation

in transactions reached US$25.5 billion Commercial banks have invested heavily in Fintech for the purpose of Fintech to bring huge profits to commercial banks The essence of Fintech lies in innovation While the Fintech innovation can not only bring more innovative products and services to customers but also bring more new ways for commercial banks to obtain funds For example, com-mercial banks can use technologies such as blockchain and distributed accounting to process the transaction data of SMEs and individuals Fintech can solve the prob-lem of the high cost of acquiring information from long-tail customers by the original financial institutions and the long and cumbersome transaction chain Fintech improves the efficiency of commercial bank loans and increase bank revenue (Gu & Zhang, 2018) In addition, Fintech can use emerging technologies to provide one-to-one differentiated services to high-quality customers of commercial banks and use the open banking model to

retail customers with a better experience, improve the operating efficiency of retail business

Fintech can solve the dilemma of poor original data quality Although commercial banks have a large amount

of national consumption and fund storage data, a large amount of data (such as customer information, product information, credit behavior, third-party platforms, etc.) cannot be used rationally Because the original technol-ogy and data are not perfect, the quality of the data and the way of generating and obtaining are also different Commercial banks are unable to make good comprehen-sive evaluations and judgements of customers in transac-tions It is extremely difficult to segment customer groups and to judge fraud in transactions Nowadays, Chinese credit investigation system is not perfect, and there are uncertainties in the source and reliability of customer data Fintech can use text mining, cluster analysis, iris recognition and other technologies to compare and ana-lyse customers' online behaviours, user preferences and public data and through learning and correction methods, the original data can be efficiently analyzed to improve the quality of data

(2) Fintech reduces banks' risk-taking and increases the probability of commercial bank bankruptcy

The superposition and aggregation effects of various risks generated by the application of Fintech are unprece-dented, and the probability of bank bankruptcy risks has

increased Various products and services of Fintech

inno-vation are accompanied by various risks, which are

dis-persed in various markets Fintech combines the

boundaries of multiple markets, such as currency,

deriva-tives, foreign exchange, etc., resulting in the continuous

superposition and integration of various risks, and the

form of change is different from the past, various capital

chains have become complicated and various financial

dimen-sion of transmisdimen-sion are also constantly changing The

uncertainty and relevance of global financial risks due to

the development of Fintech applications is

unprece-dented In the face of the continuous development of

Fintech, the cross-risks generated by national, industry

and regional risks have become complicated, and the

cus-tomers' own situation and expectations of the economic

situation have also become complicated Fintech is

con-stantly innovating, and risks are concon-stantly

that will have a huge destructive effect on commercial

banks

The application of Fintech has increased the risks of

commercial banks In terms of credit risk management:

while Fintech continues to innovate products and

ser-vices, commercial banks' credit risks are also changing

The original credit business process and risk mitigation

technology need to be revised according to actual

condi-tions Whether credit ratings, industry standards,

guaran-tees, authorizations, reviews and asset disposal policies

meet the new development requirements, and how to

modify them, are also major issues that the bank's senior

management must face Commercial banks have a credit

preference Their investments in credit rating, small and

micro enterprises(SME) and movable property financing

are relatively weak An undeveloped or even neglected

value network will cause commercial banks to lose some

high-quality customers(Anagnostopoulos, 2018) At the

same time, with the development of finance inclusive

and increasing coverage of low-end customers, the

default probability of credit risk may increase In terms of

market risk management, financial innovation and tools

are becoming increasingly diversified, various market

risk components are complex, the total portfolio

invest-ment risk is increasing and the identification, evaluation,

measurement and verification of market risks are

exchange and commodity market price fluctuations affect

commercial banks' market risk will also become

compli-cated during investment and asset-liability mismatches

period Various collections, payments, settlements,

prod-uct transactions and even investment decision making,

wealth management and interest rate pricing are very

dif-ferent In terms of operational risk management,

traditional profit methods have led to a decline in com-mercial banks' profits Comcom-mercial banks need to invest

in riskier assets with higher profits to obtain more profits

to make up for losses Risk preference to the asset side has increased, leading to increased risks on the asset side (Qi & Xiao, 2018) In addition, the fraud of internal per-sonnel of commercial banks, the management of internal control and operating procedures, operational risk infor-mation management system failure, collapse, loopholes, error risks, etc., will bring investors' capital loss or infor-mation leakage and other operational risks

Fintech produces technical risks (Gu & Shi, 2020; Ma,

Q L, & Dai, 2020) As Fintech has created a new financial ecosystem, all-day service methods and diversification of participants, massive amounts of data are constantly increasing at the power level, a large amount of informa-tion data has been generated Improper handling and storage of this information will lead to information and data leakage, increase the difficulty of risk identification and increase the risk of technology out of control Due to the lack of both rich financial knowledge and excellent computer technology professionals, the processing of unprecedented massive data generated by Fintech has encountered a bottleneck Whether the original technol-ogy can carry the capacity and processing speed of these data can be followed up? The emerging technology will also have risks in the storage, reading and processing of the data during the running-in period with the original technology Incomplete and distorted data processing improperly will also bring huge risks, leading to a signifi-cant impact on the information management system of commercial banks

Through the analysis of the internal logic mechanism

of the above Fintech applied to commercial bank risks, Hypothesis 1a and Hypothesis 1b are obtained:

Hypothesis 1a Fintech can reduce the economic capi-tal of commercial banks

Hypothesis 1b Fintech increase the economic capital

of commercial banks

A large number of scholars are studying whether bank risks are heterogeneous Most of them believe that large state-owned commercial banks have scale effects compared to small and medium banks in risk manage-ment (Gu & Shi, 2020; Gu & Zhang, 2018; Ma et al., 2020; Sheng, 2020) Therefore, from the perspective of the application of Fintech to study whether Fintech is hetero-geneous to commercial bank risks of different asset scales, it can provide a new research perspective and enrich the relevant literature On the one hand, large state-owned commercial banks have the advantages of

scale, talent, technology and more practical experience in

upgrading self-risk management technologies and

pro-cesses In 2017, with the help of the internet and other

technological developments, the five state-owned

com-mercial banks of China, Agriculture Bank of China,

Industry and Commercial Bank of China, China

Con-struction Bank and Bank of Communications established

rapidly expand their Fintech field Among them, Alibaba

and Ant Financial and China Construction Bank

cooper-ates in the areas of electronic payment and QR code

mutual recognition and scanning; JD and Industrial and

Commercial Bank of China cooperate in financing,

cor-porate credit and asset management; Baidu and

Agricul-tural Bank of China cooperate in intelligent banking and

finance inclusive; Tencent and Bank of China established

a unified financial big data platform; Bank of

Communi-cations and Suning cooperate in smart finance By

cooperating with Fintech companies, large state-owned

banks can quickly adjust their asset structure to keep up

with the pace of economic development and market

com-petition On the other hand, due to the limitations of

cap-ital and talents, small- and medium-sized banks have

difficulties in the layout of the Fintech field Therefore, it

needs to be treated differently According to their own

characteristics, joint-stock banks have different layouts in

the field of Fintech For example, the Bank of Nanjing is

mainly deployed in the small- and medium-sized bank

ecosystem and connected e-commerce, express delivery

and other industry platforms Shanghai Pudong

approach, with technology companies, scientific research

institutions, upstream and downstream suppliers and

consumers participating in cooperation China

Mer-chants Bank and Industrial Bank have incorporated

“long-tail customers” into their Fintech development

strategies Urban commercial banks and rural

commer-cial banks have limited investment in Fintech due to

cap-ital, business scope, geographical and customer groups,

and it is difficult to keep up with the pace of Fintech

Therefore, it is necessary to study the heterogeneity of

the impact of Fintech on different types of commercial

banks(Chunbing et al., 2014; Fang, Lau, Lu, Tan, &

Zhang, 2019; Lorenc & Zhang, 2020; Sleimi, 2020; Xie &

Ling, 2018) Hypothesis 2a and Hypothesis 2b are

proposed:

Hypothesis 2a The impact of Fintech on commercial

banks of different sizes is heterogeneous

Hypothesis 2b The impact of Fintech on commercial

banks of different sizes is not heterogeneous

3 | E M P I R I C A L D E S I G N Based on the theoretical hypothesis described above, this article adopts the model of Phan, Narayan, Rahman, and Hutabarat (2020) The major interest of this article is the coefficient of Fintech in the Equations (1) and (2) The study attempts to testify whether the Fintech has signifi-cant effects on the economic capital of commercial bank risk Based on this, it is modified according to specific conditions to obtain the following model:

ð1Þ

ð2Þ

The explained variables in Equations (1) and (2) are the economic capital of commercial bank risk (EC), the core explanatory variables are the Fintech Application Index (Fintech) and the control variables are the bank's profitability (ROE), asset size (SIZE), the growth rate of gross domestic product (GGDP), the growth rate of currency supply (GM2), CAR (Capital adequacy ratio), NI (Non-interest income), COST (Cost to revenue ratio) and Liquidity ratio Among

The explained variable of economic capital of commercial bank risk (EC), this article uses the VaR model to measure

As mentioned above, the economic capital of com-mercial bank risk is the capital required by the bank's board of directors to compensate for unexpected losses within a certain period according to its asset risk status and preferences It is a measure of risk, and the essence

is to cover the risk with capital Value at risk (VaR) as a standard measure of risks has been widely implemented

by financial institutions Therefore, this article will adopt

the VaR model to measure the economic capital of

com-mercial banks' risk

In July 1993, G30 Group proposed a standard method

for measuring market value at risk (VaR) VaR has been

widely implemented by financial institutions VaR is called

the value at risk, which depends on the absolute level of

risk, the manager's tolerance to risk and the length of the

risk period VaR refers to the measurement of the

maxi-mum possible loss faced by a particular investment

combi-nation within a certain confidence level and holding

period It is a forward-looking risk measurement method

(Drenovak et al., 2017) This method is a standard method

designated by the Basel Agreement to measure the market

risk of financial institutions, and its application has become

the international banking risk management standard

In order to measure risk more accurately, this article

uses the ThaiVaR (Conditional Tail Expectation, CTE)

method proposed by Artzner, Delbaen, Jean-Marc, and

Heath (1999) to estimate the economic capital of 16 listed

commercial banks from China Statistical formula to

express the definition, namely

That is:

That is:

variable X, and VaR is the value at risk under the confidence

so that more risks can be accommodated The larger the

value, the poorer risk-taking capacity of commercial banks

This paper uses a simplified method of normal

distri-bution assumption to calculate ThaiVaR When the loss

of the bank (that is, the opposite of profit) X follows a

normal distribution, namely

1−F xð Þq ;f() is the density function of X;F() is

quantile of cumulative probability is the q So the TailVaR of the general normal distribution X is equal the mean plus the result of multiplying the standard devia-tion of X and the TailVaR value of the standard normal distribution Y, namely

This paper uses the return on assets (ROA) instead of profit, namely

ð8Þ

The bank's net loss rate is the negative value of the ROA We use SPSS software to do the Kolmogorov

− Smirnov (K − S) test on ROA If the p value is less than 1, the bank net loss rate does not follow the normal dis-tribution, otherwise, it follows the normal distribution The results show that ROA obeys a normal distribu-tion So we can calculate the TailVaR value according to

The value of the standard normal distribution TailVaR is shown in Table 2 The confidence level in this paper selected is 99.99%

Finally, we multiply the total bank assets of a certain

bank's economic capital value

Fintech application index The construction of Fintech application index is a very critical step in this article Existing research is mainly constructed through text mining and index synthesis methods This article refers to Guo (2015) Internet finan-cial index construction method to construct the Fintech application index The index includes four secondary indicators: Fintech composite index, personal deposits of commercial banks, personal loans of commercial banks,

composite index comes from Baidu Search Index and Baidu Consulting The construction steps are as follows: First, select keywords This article selects keywords based

on the application of Fintech to commercial banks' third-party payment, wealth management, open banking and other fields, combined with the data provided by finan-cial reports of commerfinan-cial bank, and finally selects Fintech, electronic payment, internet finance, third-party

T A B L E 1 Descriptive statistics for ROA

T A B L E 2 The value of VaR and

TailVaR under different confidence

levels under the standard normal

distribution

0

1000

2000

3000

4000

5000

6000

13.10 2014

16.10 2017

F I G U R E 1 Trends in the various component of Fintech indexes [Colour figure can be viewed at wileyonlinelibrary.com]

construct a basic dimension Second, according to the

rel-evant data and factor analysis, the monthly average of

the keyword attention is obtained and the trend chart of

each component of the Fintech composite index from

January 2011 to September 2019 is drawn (Figure 1)

Finally, through the use of SPSS software for principal

component analysis and factor analysis, to reduce the

dimensionality of keywords, get the common factors of

keywords and then calculate the Fintech composite

index Then use the same method to integrate the four

secondary indicators and finally establish a Fintech

appli-cation index

When using SPSS for data processing, the KMO test

value is 0.599 In the Bartlett test, the Sig value is less

than 0.001, indicating that the selection method in this

article is reasonable

We choose several bank and country level characteristics

as control variables In terms of the country

characteris-tics, the control variables include (a) The growth rate of

GDP (the logarithm of Gross national product) Studies

suggest when the economy is on an upward trend, banks

are willing to invest more funds to obtain higher returns

Once the economic situation reverses, the market risk of

commercial banks also increases In order to draw

lessons from past experiences, Basel III proposed

macro-prudential principles and management of reverse

commercial banks (b) The growth rate of M2 (the

loga-rithm of currency supply) The monetary authority

imple-ments monetary policy mainly through commercial

banks, and the change in the money supply represents

the intention of the monetary authority to adjust and

control the market With the continuous development of

Fintech, the speed and cost of transactions have

decreased Recently, the monetary authorities have begun

to try to issue digital currencies, and the money (paper

currency) supply may decrease in the future

In terms of the bank-specific characteristics, the

con-trols include (a) ROE (net assets per share) On the one

hand, Fintech can reduce the asymmetric cost of

transac-tion informatransac-tion and increase profitability On the other

hand, in accordance with the principle of high risk and

high return, commercial banks will be encouraged to

par-ticipate in high-risk behaviours in order to pursue higher

returns (b) COST (the cost to revenue ratio, operating

expenses and depreciation account for the proportion of

operating income), it reflects bank's ability to obtain

income The higher the proportion, the higher the cost

and expense and the poorer profitability of the bank In the environment of Fintech innovation, commercial banks can develop and provide more new products and services In the process of innovation, they can expand their business scope and increase the channels for obtaining funds, which will reduce operating costs and increase economic benefits, and increase profitability of commercial banks (c) NI(the Non-interest income) is also an important indicator for measuring bank profit-ability The original single profit model of commercial banks' deposit-loan spreads limited the profitability of commercial banks under interest rate marketization Fintech can help banks vigorously develop capital trans-actions and clearing, asset management and robot con-sulting services These non-interest income businesses can improve the profitability of banks However, some scholars have found that the expansion of non-interest income business will lead to increased instability of bank income, which will increase bank market risks And these risks are heterogeneous for different types of banks (Jin, 2018; Xu & Zheng, 2018) For example, large state-owned commercial banks have advantages in terms of asset scale and customer sources, which can partially reduce instability Joint-stock banks are subject to geo-graphical and scale restrictions, but due to their relatively small business volume, the overall risk is lower than that

of state-owned banks (d) Asset size (the logarithm of total assets) The relationship between the size of bank assets and the level of risk management is temporarily

larger the scale of bank assets, the diversified investment can diversify risks and improve risk management and profitability On the other hand, due to the huge divi-dends generated at the beginning of the Fintech era, com-mercial banks tend to invest in high-risk areas, and the probability of commercial bankruptcy has increased (e) CAR (Capital adequacy ratio) It reflects the capital adequacy of bank and is a comprehensive index that measures the ability of bank to bear risks The higher the capital adequacy ratio, the higher the capital required for market risks and the reduction in bank funds available, but the more robustness (f) Liquidity ratio, it is due to the mismatch in the maturity of bank assets and liabili-ties, and the high liquidity ratio indicates that the bank is safe, but it also indicates that the bank's capital utiliza-tion is insufficient

Since the units of different variables are different, if they are used directly, it will cause errors in the empirical test results Therefore, this article uses the mean-SD method to process the original data of the above variables

to avoid the influence of incorrect results due to different original data units in the comprehensive evaluation process

n

i = 1

i= 1

ð9Þ

To identify the impact of Fintech on the market risk of

commercial banks, this article used a database that

con-tains 16 listed commercial banks in China from January

2011 to September 2019 It is the reflection of Chinese

sit-uation but not suitable for other countries

Among the 16 listed banks, 5 large state-owned

com-mercial banks: Bank of China, Agricultural Bank of

China, Industrial and Commercial Bank of China, China

Construction Bank 8 joint-stock commercial banks: Ping

An Bank, Hua Xia Bank, Industrial Bank, China Minsheng Banking Corp., Ltd., Shanghai Pudong Devel-opment Bank, China Everbright Bank, China Merchants Bank, China CITIC Bank Three city commercial banks: Bank of Beijing, Bank of Ningbo, Bank of Nanjing The data comes from the Wind database and the financial reports of commercial banks, the control vari-able data comes from the EPS database, and the missing data comes from online collection

Table 3 lists the descriptive statistical results of the main variables in the model From this, we can notice several important statistics The average value of Bank asset scale

of bank market risk is 1.902, and the SD is 0.917, which indicates that the range of change during the sample period is large and the stability is weak The Fintech application index is 3.19, and the SD is 0.46 The EC vari-ation range is also large The SD ROE, COST and GGDP are also large, indicating that their stability is also weak Moreover, there is a big difference between the

T A B L E 3 Descriptive statistics for model variables

level

Core explanatory

variable

index

Financial technology index

JRKJ2

Third-party payment index

JRKJ3

Electronic payment index JRKJ6

GDP

maximum and minimum values of all variables, which

indicates that there may be heterogeneity in the sample

General econometric methods such as ordinary least

squares, fixed effects, random effects and generalized

least squares may not meet our estimation requirements,

resulting in biased results To solve this problem, we use

the generalized method of moments (GMM) proposed by

Arellano and Bond (1991) This method has less strict

requirements on the assumptions than the least square

method The loose assumptions of the GMM method

make it widely used in Econometrics The bias of the

finite sample in the GMM estimation is negligible, and

the variance is much smaller

4 | E M P I R I C A L T E S T

Before presenting and interpreting our test results, we

first check the possible multicollinearity between the

model variables Multicollinearity can distort the

accu-racy of regression coefficients and make their estimated

fluctuate to the data The results of the multicollinearity

test are shown in Table 4

Table 4 shows that state is highly correlated with

sev-eral variables: GGDP, GM2 and NI Except for the three

cases, the relationship between the other variables is

weakly, so there is a problem of multicollinearity

We conducted empirical tests according to

Equa-tions (1) and (2) and found that the empirical results

were not ideal, and the coefficients of some indicators

were contrary to economic significance Therefore, we

decided to modify the original model to eliminate

variables that are both highly correlated and contrary to economic significance The final results are shown in Tables 5 and 6

Table 5 presents the results of Fintech on the eco-nomic capital of bank risks This article uses panel mixed regression, fixed effects and random effects, the differen-tial GMM method of the GMM method and the system GMM (Sys-GMM) to test Because economic capital is a continuously adjusted variable, the first-order lag period

of economic capital is introduced as an explanatory vari-able in the model At the same time, other varivari-ables will influence each other, which will cause the explanatory variable to be related to the disturbance term, and there will be endogenous problems In order to solve the corresponding endogeneity problem and avoid the loss of sample information caused by the differential GMM, we use the dynamic panel system generalized estimation method to estimate the model, namely SyS-GMM This article will focus on the experimental results of the SyS-GMM In order to ensure the applicability of the estimation method, AR(1) statistics and AR(2) statistics are used to test the autocorrelation of the disturbance items, and he Sargan test is used to analyse the exogeneity of the instrumental variables The p value is greater than 5% The validity of the instrumental variable was tested by Sargan test, which confirmed the validity of the instrumental variable because its p value was greater than 5% That is the choice of this model is reasonable Empirical results are displayed in Table 5, the Fintech has a negative and significance effect at the 1% level of significance on the performance of China banks, indicat-ing that the higher the development of Fintech, the lower the risk economic capital (measured by EC) required by

Hypothesis 1a

The coefficient of GGDP has a significant negative correlation with the economic capital of commercial bank risk It shows that Fintech can provide commercial

T A B L E 4 Correlation matrix