This study uses quantitative research methods to answer the questions and analyze the impact of Fintech on financial stability of the commercial banking division in Vietnam.. This articl
INTRODUCTION TO THE RESEARCH TOPIC
Reason for choosing topic
Commercial banks system have a role as an intermediary financial institution in the economy, facilitating the efficient allocation and provision of capital to those in need, therefore fostering sustainable economic development In 2008, because of a series of structural shifts during the financial crisis, important banks and financial institutions had faced to enormous damage such as bankrupt, increasing a huge risk on in the long time for the financial market Consequently, this crisis has almost engulfed the entire world economy From the situation in the world, we can see the importance of the banking system to a country
Financial stability in a commercial banking system is a problem, although it has been studied a lot by authors around the world before the situation of the economy and the world and Vietnam has to face Facing many fluctuations such as pandemics and war,
It can be seen that the additional work on the stability of the commercial banking system is suitable for the current situation of the world and Vietnam
Moreover, in the increasingly integrated, quickly expanding, and highly competitive economic environment, banks must constantly deal with a variety of risks during their operations, including the speed of the fourth industrial revolution has been quickly growing in recent years, and we can witness how it is affecting a wide range of sectors and areas Therefore, the financial and banking industry was not an exception, it receives significant changes throughout the years As of 2023, Vietnam currently has 188 Fintech companies operating flexibly in various fields such as: The birth of big data (Big Data), the Internet of Things (IoT), and artificial intelligence (AI) have led to the emergence of new services to maximize the industry's potential But Fintech stands out above all and plays an important role in the bank system In Vietnam, research papers on the influence of Fintech on the banking system's strength have been quite long and not suitable for the current economic situation of the country Therefore, the author express to contribute more empirical evidence to help Viet Nam commercial banks have more grounds to maintain not only for the financial stability but also the established and operating of Fintech firms
In short, the author found that the studying on this fields is still in its early stage, the number of studies on financial stability is lacking in Viet Nam, especially the evolution of Fintech for the financial stability of Viet Nam commercial banks Therefore, identifying how the influence of Fintech may assist for Viet Nam commercial bank is necessary and meaning in order to strengthen the financial stability of commercial banks
As the result, the author decided to choose the topic “THE IMPACT OF FINTECH ON THE FINANCIAL STABILITY AT VIETNAMESE COMMERCIAL BANK” as the graduate thesis Using secondary data of Viet Nam commercial banks in the period 2010
- 2023, the study conducts the quantitative analysis to evaluate the impact of these factors on the financial stability of Viet Nam commercial banks.
Objective of study
The thesis analyzes the influence of Fintech affecting the financial stability of Vietnamese commercial banks The results of the study are used to propose a number of recommendations to improve financial stability for Vietnamese commercial banks
In order to achieve these goals, the thesis will analyze the factors affecting the stability of Vietnamese commercial banks, such as:
Firstly, identifying and evaluating the impact of Fintech on financial stabilities based on quantitative analysis models
Secondly, after determining the level of influence, the study will propose some solutions and recommendations to improve and enhance financial stability based on technology activities of commercial banks of Vietnam in the near future.
Research question
- What is the correlation between financial technology and financial stability of Vietnamese commercial banks?
- Which recommendation are proposed to improve the financial stability of Viet Nam commercial banks?
Subject and scope of the study
The research topic is aimed analyzing the impact of Fintech on the financial stability of commercial banks in Vietnam In which, the level of financial stability is represented by the bank’s bankruptcy risk through the z-score index based on the theory and related research
Scope of spatial research: According to data updated on March 31, 2023, the number of commercial banks currently is 31 banks (SBV, 2023) However, the collection of full public data of some Viet Nam commercial banks in the period is still limited Therefore, only 25joint stock commercial banks out of 31 commercial banks were selected because their data were sufficient for the study
Scope of time research: The research sample used is based on the data collected in the period from 2010 to 2023.
Contributions
The research aimed to analyze the link between the dramatic expansion of Fintech and the banking-finance sector In fact, the research provided is complete and domestic and foreign studies have been published to identify the research gaps related to selecting the most suitable model for the topic, results in this thesis can be used as a reference for scholars for traditional banks to improving the efficiency for Vietnam’s commercial banking system, thus contributing to giving a different perspective on the problem to easily identify risks and take appropriate measures.
The composition of the thesis
In addition to the table of contents, list of figures and tables, references and quotations using APA style (The American Psychological Association), the thesis is presented in 5 chapters:
Chapter 1: Introduction to the research topic
This chapter presents an overview of the research paper including the following contents: reasons for choosing the topic, research problem; objectives of the study; research questions; object and scope of the study; research significance; research paper structure
Chapter 2: Literature review and related empirical studies
In this chapter, the study will first present the concept of financial stability and theoretical basis related to Fintech; and relevant empirical studies from abroad and in Viet Nam to identify research gap, thereby presenting panel data regression models applied to assess the impact of factors affecting the financial stability of Viet Nam commercial banks
Chapter 3: Research model and methodology
Based on the content presented in chapter 2, chapter 3 will focus on presenting the content related to the research model, research variables, research data, research methods, and processes to achieve results for which the aim of the study is concerned
Chapter 4: Research results and discussion
Chapter 4 focuses on two topics: descriptive statistics of the research variables and testing of the research model, thereby obtaining research results and analyzing the correlation relationship, direction, and level of influence the impact of variables on the Fintech of Vietnamese commercial banks
After collecting the research results from chapter 4, chapter 5 will re-evaluate the research results, give comments on the limitations of the study (if any), and finally make recommendations to improve the efficiency of Fintech for Vietnamese commercial banks
In chapter 1, the author introduces an overview of the selected topic and presents issues surrounding the research topic, including the reason for choosing the subject, research objectives, research questions for the case The object and scope of the research, the structure of the topic and the contributions of research topic.
REVIEW OF THE LITERATURE
Overview of basic concept
The concept of Fintech in the economic literature has been defined in many different ways:
Dorfleitner et al (2017) argued the term “Fintech” that combines the phrases Financial and Technology, aim to attract customers with products and services that are more user-friendly, efficient, transparent, and automated Nowadays, with the expansion of technology by the 4.0 revolution, the definition of Fintech is expanded to perform other functions The finance sector includes a Fintech segment that makes financing available for both private individuals and for businesses such as crowdfunding, credit and factory, alternative payment methods, personal financial management, investment and banking, insurance, etc
It is frequency referred to as the “new marriage” of financial services and information technology in today’s world (Buckley et al., 2016) The concept of Fintech has been formed for a long time and has gone through many different stage and periods (Armer et.al., 2015) These stages are:
Fintech 1.0 (1866 - 1967): from analogue to digital
Fintech 2.0 (1967 - 2008): Development of Traditional Digital Financial Services Fintech 3.0 (2008 - present): Democrating Digital Financial Services
Fintech 3.5: The Examples of Asia and Africa
But critically, Fintech became more known after 2008 financial crisis, when at that time, certainly and transparency were essential factors for the operation of banks in particular and the financial market in general
Nowadays, Fintech spans a variety of sectors and businesses, including education, retail banking, nonprofit fundraising, investment management, and many more, Fintech also encompasses the creation and usage of digital currencies like bitcoin (Farahani et al., 2022)
Fintech’s operating segment is divided into 5 main segments including: (i) payments, clearing, and settlement; (ii) deposits, lending and capital raising; (iii) insurance; (iv) investment management; and (v) market support (International Monetary Fund, 2019)
Fintech could impact market structure through different channels Among them the most emphasized are Mobile payments, Mobile Peer-to-peer, Lending Platform, Personal financial management, Mobile insurance, Online Stock Trading and Buy Now Pay Later
Mobile payments: Popular mobile wallet apps include Apple Pay, Google Pay,
Samsung Pay, Momo, ZaloPay, AirPay, which is linked to a customer’s credit card, debit card, or bank account
Mobile peer-to-peer: This type of transaction, which allows individuals to transfer money to other individuals via a mobile apps such as PayPal, TransferWise, Remitly, WorldRemit, In Vietnam, Momo, ZaloPay, ViettelPay and AirPay are the most commonly used
Peer-to-peer Lending Platform: Lending Club, Prosper, Zopa, Funding Circle, Kiva and Mintos are the most reliable peer-to-peer lending platforms in the world There are several significant financial technology firms located in Vietnam that have expanded this type of financing loans like Tima, Vaymuon, VaytienOnline, EasyCredit,etc
Personal financial management: Personal finance apps such as Mint, Personal
Capital, PocketGuard, You Need a Budget (YNAB), Spendee, Money Lover are used widely around the world including Vietnam
Mobile insurance: Lemonade, PolicyBazaar, Metromile, Oscar and Clover Health are a standard examples of insurance companies In Vietnam, insurance apps including
BIDV Insurance, FWD Insurance, PVI Insurance, Liberty Insurance and Manulife Insurance are expanded
Online Stock Trading: E-Trade, TD Ameritrade, Robinhood, Fidelity and Charles
Schwab are the best online broker for online trading In Vietnam, some stock brokerage companies including VNDIRECT, HSC, SSI and MBS are starting to improve on online stock trading
Buy Now Pay Later: Klarna, Afterpay, ZipPay and Affirm are the most popular Buy
Now Pay Later apps in the world In Vietnam, there are some start-up companies in this field like Finhome, Fundiin, Bizzi
Table 2.1 Top 12 Fintech countries in the world as of 25th August, 2023
(Source: Global Fintech Index, American Banker, 2023)
As a consequence, many Fintech startups in the mentioned countries have been established and some of them have now become Fintech giants in the world possessing extremely high values such as:
United States: Fintech has been growing rapidly in the USA since the 21 st century The top Fintech companies based in the USA such as Stripe, PayPal and Robinhood have enhanced the quality of financial stability in the world In there, Stripe is known as the largest Fintech company in the world at the moment when it is valued at about $95 billion
United Kingdom: For UK, it is crucial for the legal framework to enable rather than impede innovation so that the potential of being the top countries about Fintech Money Box, Monzo, Transfer Wise, Payment Sence and Starling Bank are known as the key leaders of finance technologies in the England market
Sweden: Stockholm-based payments company Klarna which currently has nearly four million monthly active users in the US and nearly 90 million users globally
Brazil: With one of the largest Latin American financial markets, Brazil’s number of Fintechs has soared over the past years, approximately 750 businesses At the moment, Brazil has exhibited rapid growth in Fintech activity within the overall financial system Nubank is one of the leading Fintech businesses in Latin America with over 40 million customers
When it comes Asia, Fintech is one of the fastest expanding technology sectors Therefore, the number of Fintech startups also grows In particular, India and Singapore are home to many startups with the value of 1 billion USD or more, also known as
India: India currently boasts the third-largest Fintech ecosystem in the world, behind the US and the UK The number of startups has grown significantly over the past few years with the top businesses, some of the prominent Fintech startups in Indonesia are Paytm, Pine labs, PayU and Faircent
Singapore: Thanks to its robust regulatory framework and advanced infrastructure,
Singapore has recently emerged as a frontrunner in Asia of Fintech With outstanding government, infrastructure and digital capabilities, Singapore has spawned many potential Fintech startups Some of them are Distributed Ledger Technologies, Endowus, ShopeePay, TranSwap,,
Overview of financial stability of commercial banks
2.2.1 The concept of Commercial banks
The country’s system of commercial banks always acts an important role Therefore, a Commercial bank has been defined in many different ways
A commercial bank is an organization that deals in money, receives deposits from actors in the economy, then makes loans and provides a variety of financial services to actors in the economy, according to Nguyen Viet Hung (2008)
Hoa et al (2016) mentioned that a commercial bank is a financial intermediary that raises capital from entities with excess capital and makes loans to entities lacking capital According to The Law on Credit Institutions of Vietnam, Commercial bank is a type of bank that is allowed to conduct all banking operations and other business activities under the law for profit Banking operations such as: receiving money from an organization or individual as demand or term deposit, savings deposit; credit extension to individuals and organizations; providing via-account payment services
In short, as a financial intermediary, commercial banks convert idle capital into loans For individuals and organizations to conduct business, with the goal of profit Simultaneously, commercial banks are an important factor in the market for government- issued bills and bonds to finance community programs
2.2.2 The concept of financial stability
Defining “financial stability” plays an important role in developing appropriate analytical tools as well as macroprudential operating policies Currently in the world, there are many definitions of “financial stability” According to the Bank of England
(2001), financial stability implies identifying risks in the financial system and acting to minimize them Wellink (2002) argued that a stable financial systems is capable of efficiently allocating resources and absorbing shocks, preventing adverse effects on the economy and financial system The Banks of Australia (2005) defined financial stability as a state in which financial intermediaries, markets and financial infrastructures allocate capital flows between saving and investment, thereby promoting economic growth Buiter (2008) found that financial stability will be ensured if the following factors are not present: (1) Asset price bubble; (2) Lack of liquidity; (3) The default of financial institutions threatens the stability of the system The European Central Bank Expressed
(2014) found that financial stability is the state in which financial institutions are able to absorb shocks, minimizing serious failures that can cause significant harm to the economy The State Bank of Viet Nam (2021) defined that financial stability of a system includes many components such as the stability of financial intermediaries, financial infrastructure and financial markets
In particular, the stability of the operation of financial intermediaries is one of the most important factors of the stability of the financial system Commercial banks are one of the most important financial intermediaries in the financial system The monetary business of commercial banks is also affected by many direct and indirect impacts from the internal difficulties of the financial system and the economy as well as external influences Therefore, the financial stability of commercial banks is considered as an important content in the stability of the financial system
Shortly, from a macro perspective, financial stability is a state in which the financial system performs its functions smoothly, contributing to the efficient allocation of resources of the economy At the level of commercial banks, financial stability can be understood as the state in which the organization operates stably, effectively, and is able to withstand shocks from the external environment and itself does not cause a shock affecting the background economy.
The importance of Fintech for the financial stability of commercial bank
As Pham Tien Dung, Director of Payment at State Bank, has said, “Fintech used to support or enable Mobile Banking and thanks to the growth of Fintech, traditional banks have the entire digital ecosystem” In the past, banks used to build up a Mobile Banking project; however, because of the high investment cost, these projects have to be stopped When cooperation between banks and Fintech occurred, the rise of Internet Banking became more popular
Fintech contributes to the country’s economic development Therefore, the increased of Fintech at commercial banks leads to positive impacts on the development of the national economy, especially in the process of commercial banks:
Improving customer services: Fintech that use technology in financial institutions, such as lending and investment, card payments, crowdfunding and foreign exchange With the aim of providing customer-oriented solutions in more personalized, transparent and accessible for via digital channels
Increasing financial and payment capacity: Fintech includes new models, such as mobile wallet, mobile payment, online transfer and bitcoin, which can rapidly be defined as a term of solution to provide financial services in the most efficient way and at the lowest cost possible
Enhancing internal processes and optimizing operations: Fintech has a serious effect on the internal process at commercial banks For example, the RPA technology (Robotic Process Automation) automates repetitive tasks, it helps different systems talk to each other better and reduce manual work Otherwise, Fintech leads to efficient banking performance when handling data, analyzing data and including AI software In other words, it leads to a reduction in costs, offers more scalability and flexibility for businesses
Handling data: Fintech influences commercial banks from their techniques and data analysis ability Banks more accurately describe customers' trading habits, investment and financing needs; evidently, by collecting and analyzing data, banks can increase process efficiency and minimize risk management
Exploring and investment in new technology: Fintech refers to the promotion of financial innovation through techniques like blockchain, AI, Internet of Things (IoT), etc Blockchain technologies have the potential in improving security and transparency in financial transactions; customer information can be analyzed by AI; IoT provides information by devices linking with payment processes so that banks identify customer behavior and need to enhance their service
Risk management: To minimize loan loss as well as credit risk, banks need to have an effective Fintech system Banks must have systems with high technology such as AI, learning machines and blockchain to ensure that they can analyze and assess default risk without their knowledge
The financial stability of the commercial bank industry in Vietnam benefits greatly from Fintech, which creates an environment that is more stable, transparent and efficient However, it is crucial to recognize the wider economic difficulties that exist, information security and risk management create an intricate setting for the expansion of Fintech
Chart 2.1 Cooperation between Banks and Fintech
Source: Cooperation model between Banks and Fintech of Bửmer & Maxin (2018)
The model show that there are various ways for cooperating between banks and Fintech:
Banks enable new Fintech products: Because banks have expertise in the national economy, especially in the network, knowledge and product while Fintech have innovative ideas for technology With the collaboration, Fintech can expand their product and improves the customer service such as lending, payments, insurance, investment, etc
Banks enable Fintech’s market entry: By sharing banking performance, Fintech companies can easily understand the finance market and customers Moreover, through cooperation with Fintech, banks can react more quickly to the emerging market changes and be more effective in product and service development
Banks increase Fintech’s profit: Fintech companies can acquire clients, network, funds and reputation from banks and then customize to represent their brand As a result, cooperation with banks allows Fintech to build trust among customers and prove their reliability
Beside the traditional banking model, the opportunities and benefits of Fintech for the financial industry from different perspectives are diverse Firstly, Fintech offers lower transaction cost and faster banking services by applying technical information and AI; then reducing many steps; operating cost and optimization risk management Secondly, customer standards can be expanded rapidly, not only for business but also for individuals This appears in the P2P, mobile payment, automatic transfer and personal financial management Secondly, Fintech allows customer access easily with financial services without traditional banks Loan registration is improved in a more flexible way than in the past; as a result, startup, boarding house owners, small and medium-sized businesses who do not have access to bank loans can open new possibilities of access to capital
In conclusion, the cooperation between Fintech and banks provides many benefits to both parties Fintech companies are more directly connected to their target customers, having exploitation of opportunities arising from the bank’s capital and brand The partnership may also provide banks with new innovative solutions to expand their product and service portfolio.
RESEARCH MODEL AND METHODOLOGY
Data processing methods
The impact of Fintech on the financial stability of Viet Nam commercial banks is examined through four models separately which are popular models for pannel data, including: (1) The Pooled OLS model; (2) The Fixed Effect model (FEM); (3) The Random Effect model (REM); (4) Generalized Least Squares (GLS) The author uses tests such as F-test, Breusch and Pagan test and Hausman test to determine the best model based on table data regression To address issues such as changes in variable errors and correlation analogies, after that applying the GMM strategy by Blundell and Bond (1998) to solve the endogenous problems
Based on this formal model, in order to analyze how Fintech affecting the financial stability of commercial banks, the research conducted to test the reliability of the research results as follows:
Multicollinearity occurs when two or more independent variables in a data frame have a high correlation with one another in a regression model In this research, the method to detect multicollinearity is to calculate the variance inflation factor (VIF) for each independent variable According to Muda and et al (2013) a VIF value greater than
10 shows that there is a multicollinearity, therefore, variables which are greater than 10 will be removed
• Test of OLS and FEM:
The F test is used to determine whether to use OLS or FEM model with with 2 hypotheses posed as:
H 0 : The Pooled-OLS model is more suitable for the research variables
H 1 : The FEM model is more suitable for the research variables
If the results of both models show a p-value of ≤5%, so we reject H0 to infer the appropriate FEM model
• Test of OLS or REM:
The Breusch and Pagan test is used to determine whether to use OLS or REM model with 2 hypotheses posed as:
H 0 : The Pooled-OLS model is more suitable for the research variables
H 1 : The REM model is more suitable for the research variables
If the p-value ≤ 5% you can reject the H0, choosing REM model, however, if the p- value ≥ 5%, you can choose the OLS model
• Test of FEM and REM:
The Hausman test is used to determine whether to use FEM or REM model with 2 hypotheses posed as:
H 0 : The REM model is more suitable for the research variables
H 1 : The FEM model is more suitable for the research variables
If the p-value ≤ 5%, you can reject the H0, choosing FEM model, however, if the p- value ≥ 5%, you can choose the REM model
• Test of variance and autocorrelation of the model:
To check the model has the phenomenon of variance or not by performing the Wald test with 2 hypotheses posed as:
H 0 : The model has no variance phenomenon
H 1 : The model has a variance phenomenon
If p-value ≤ 5%, you can reject H0, accept H1, therefore, the model has a variable variance phenomenon However, if p-value ≥ 5%, you can reject H1, accept H0, therefore there is no variable variance phenomenon is this model
To test whether model has autocorrelation by performing the Wooldridge test, the research test hypothesis is set as:
H 0 : Research model has no autocorrelation
H 1 : The research model has autocorrelation
If p-value ≤ 5%, you can reject H0, accept H1, therefore, the model has an autocorrelation phenomenon Moreover, if the model appears variance phenomenon or autocorrelation phenomenon, the writer will use the FGLS (Feasible Generalized Least Squares) to overcome these phenomenon
The GMM model is a model used to eliminate endogenous variables in the research model, and at the same time, GMM also solves the phenomena occurring in the model such as methodological phenomena the difference, autocorrelation, etc To prove that the GMM model is appropriate, the research model needs to ensure the following 4 conditions are met: i The number of tools in the research model does not exceed the number of research groups ii The P-value in the Arellano-Bond test must be greater than 10%, in which the research hypothesis is set as:
H 0 : The research model does not occur sequence autocorrelation
H 1 : The research model occurs the phenomenon of series autocorrelation iii The P-value in the Sargan test must be greater than 10%, and the hypothesis is:
H 0 : Instrumental variables are appropriate and endogeneity does not occur
H 1 : Instrumental variables are not suitable and endogenous phenomenon does not occur iv The P-value in the Hansen test must be greater than 10%, the hypothesis posed in this test is:
H 1 : Instrumental variables do not match
After collecting enough secondary data from 25 commercial banks in Vietnam from
2010 to 2023, the author put panel data into STATA data analysis software, then used four methods as mentioned above The specific steps in data collection analysis are as follows:
Table 3.1 The order of execution
Source: Synthesized by the author
Step 1: The author collected data from the published annual financial statements of
Viet Nam commercial banks for the period 2010 - 2023 After that, analyzing descriptive statistics to get an overview of the independent and dependent variables in the research model
Step 2: On the basis of theory and empirical studies, the author reviews the overview of previous studies and builds a suitable research model to analyze the impact of income diversification on financial performance of commercial banks in Vietnam
Step 3: The author analyzes the impact of Fintech on the financial stability of the bank From the proposed research model, applying quantitative methods through Pooled- OLS model, FEM model, REM model, FGLS model and GMM model, the author will estimate the impact of Fintech as well as independent variable on financial stability of Vietnamese commercial banks
Step 4: To ensure transparent research results, the author conducts related tests such as testing for multicollinearity, autocorrelation, variable variance and endogeneity
Step 5: The author presents research results on the impact of Fintech on financial stability of Vietnamese commercial banks, and discusses and compares the results with related empirical studies
Step 6: Based on the research results, the author makes conclusions and recommendations to improve the level of Fintech and financial stability of Vietnamese commercial banks.
Research model
Based on previous studies of many authors on bank stability, along with the impact of Fintech on the stability of the domestic and foreign commercial banking system in each period Different periods from which to propose a separate research model for the financial stability of Vietnam's commercial banking system in the period from 2010 to
2023, the following is the author's research model:
LnZScore it = β 0 + β 1 FT it + β 2 ROE it + β 3 SIZE it + β 4 GROW it + β 5 NPL it + β 6 CAP it + β 7 CIR it + β 8 LDR it + β 9 GDP t + β 10 INF t + ԑ it
LnZScore it represents the financial stability of the commercial banks (i) at time (t)
FT it represents the number of Fintech company at time (t)
ROE it represents the return on equity of the commercial bank (i) at time (t)
SIZE it represents the size of commercial bank (i) at a time (t)
GROW it represents the asset growth ratio of the commercial bank (i) at a time (t)
NPL it represents the Non-Performing loan ratio of the commercial banks (i) at time (t)
CAP it represents the capital adequacy ratio of the commercial banks (i) at time (t)
CIR it represented the banks performance of the commercial banks (i) at time (t)
LDR it represents the loan to deposit of the commercial banks (i) at time (t)
GDP t represented the economic growth at the time (t)
INF t represented the inflation rate of the year of observation at time (t)
Measurement of research variables and research hypothesis
The financial stability will facilitate the development of commercial banks and vice versa The study applies the z-score to measure the bank's financial stability Accordingly, the z-score is the dependent variable in the research model The calculation of z-score was proposed by Beck et al (2013); Houston et al.(2010); Laeven & Levine
Fintech is considered based on the operating revenue of fintech company compared to commercial banks or fintech company’s total current liabilities However, a full public data in Viet Nam is still limited so the author using logarithm of fintech company to estimate the Fintech variable according to the research of Phan et al (2019) Theories related to fintech often analyzed that fintech helps to increase financial performance, especially when the size and scope of the bank’s operations increase Many studies at home and abroad show a positive impact between fintech and financial stability such as Dorfleitner et al (2017), Yen et al (2022), In contract, some other studies show a negative impact between fintech and financial stability of banks In conclusion, this study proposes the hypothesis 1 that the Fintech has a positive impact on financial stability of commercial banks
FT = Logarithm(Fintech company at time t) Hypothesis 1: The Fintech (FT) has a positive impact on financial stability of commercial banks
The rate of return on total equity is an indicator showing the correlation between a company's profit and its equity to evaluate the efficiency of the enterprise's equity use and the bank's efficiency in generating revenue profits by exploiting that bank's equity The higher the ROE, the higher the profit of banks generates related to its equity According to previous study, Allen and Santomero (1997); Laeven and Levine (2009) believed that there is a negative relationship between the Return on equity and money stability However, Adusei (2015) and Chand et al (2021) showed that ROE has a positive impact on banking stability Therefore, this study proposes the hypothesis 2 that the ratio of return on equity has a positive impact on financial stability of commercial banks
Hypothesis 2: The return on equity (ROE) has a positive impact on financial stability of commercial banks
The bank size is measured by taking the logarithm of the bank's total assets According to a previous study of Ullah (2021) found that bank size (SIZE) has a negative relationship with financial stability of commercial banks However, Quoc (2020), Mkadmi et al (2022) showed that there is a positive relationship between SIZE and future earnings and profitability Therefore, this study proposes the hypothesis 3 that the bank size has a positive impact on financial stability of commercial banks
SIZE = Logarithm(total asset) Hypothesis 3: The bank size (SIZE) has a positive impact on financial stability of commercial banks
Asset growth is determined by calculating the difference between the total value of bank asset provided in the calculation period, so with the comparison period, the growth rate is more or less reflected in absolute value Moreover, asset growth rate can be in an extended or closed state Lan (2021) found that the revenue growth of the current year compared to the previous year contributes to the cash flow of commercial banks, which is one of the factors that significantly impact the bank's stability Therefore, this study proposes the hypothesis 4 that the asset growth rate has a positive impact on financial stability of commercial banks
Hypothesis 4: The credit growth (GROW) has a positive impact on financial stability of commercial banks
Non-Performing loan ratio (NPL)
Non-Performing loan ratio is one of the problems that commercial banks face because Non-Performing loan ratio can affect the performance and profitability of the bank According to Ozilli (2018), Defund (2022) Non-Performing loan ratio is defined as an amount for which a borrower has failed to make scheduled payments within at least
90 days An increase in Non-Performing loan ratio will increase banks' credit risk, negatively impacting banks' financial stability Therefore, this study proposes the hypothesis 5 that the Non-Performing loan ratio has a negative impact on financial stability of commercial banks
Hypothesis 5: The Non-Performing loan ratio (NPL) has a negative impact on financial stability of commercial banks
Capitalization ratio is measured by the total amount of debt in a company’s capital structure relative to its two capital sources, equity or debt Bank capital is considered an important source that can bring financial strength to banks According to research by Diamond and Rajan (2000), Quoc (2020), Ozilin (2018), Aroghene and Ikeora (2022), bank capital harms credit risk so higher capital requirements ensure that banks have sufficient capital to absorb unexpected losses when losses materialise Therefore, this study proposes the hypothesis 6 about the relationship between the bank's capital ratio and banks’ financial stability is positive
Hypothesis 6: The bank capital (CAP) has a positive impact on financial stability of commercial banks
The cost to income ratio (CIR)
The banks performance (CIR) is a measured by operating expenses divided by operating income of commercial banks i at time t The cost to income ratio shows the relationship between the bank’s expenses and its income This ratio gives investors a better view of the organization’s performance So the smaller the ratio, the more efficient the bank is Thanh (2021), Quynh (2020) and Ullah et al (2021) shows that the higher to cost to income ratio, the lower the financial stabilities of banks Therefore, this study proposes the hypothesis 7 that the cost to income ratio (CIR) has a negative impact on financial stability of commercial banks
Hypothesis 7: The cost to income ratio (CIR) has a negative impact on financial stability of commercial banks
The loan to deposit (LDR)
The ratio of loan to deposit is measured by the total loans divided by total customer deposits This variable can reflect a bank's liquidity position according to Odeduntan
(2016), Nurfalah et al (2018) The higher the value of the LDR, the higher the bank's liquidity risk, so the lower the bank's stability Therefore, this study proposes the hypothesis 8 that the loan to deposit has a negative impact on financial stability of commercial banks
Hypothesis 8: The loan to deposit (LDR) has a negative impact on financial stability of commercial banks
During periods of economic growth, individual and business borrowers need adequate capital to pay their debts, but during a recession, the ability to repay the loan decreases According to Lan (2021), Almahadin and et al (2020), Mkadmi and et al
(2022) research shows that economic growth increases bank profitability, thus increasing bank stability Moreover, the study of Alfiyan et al (2023) shows that the growth rate is kept stable during the Covid 19 pandemic Therefore, this study proposes the hypothesis
9 about the relationship between the growth rate and banks’ financial stability is positive
Hypothesis 9: The growth rate (GDP) has a positive impact on financial stability of commercial banks
The inflation rate is calculated as the inflation rate of the year of observation According to Lan (2021), Defung and Yudarudding (2022), there is a negative relationship between bank stability and the current year's inflation rate The lower financial stability in the banks Moreover, the study of Alfiyan et al (2023) shows that the inflation rate is kept stable during the Covid 19 pandemic Therefore, this study proposes the hypothesis 10 that inflation rate has a negative impact on financial stability of commercial banks
Hypothesis 10: The inflation rate (INF) has a negative impact on financial stability of commercial banks
Table 3.2 Variables in the research model
Code Measure Expected relationship Studies
FTit Logarithm (Fintech company at time t) +
Phan et al (2019) Dorfleitner et al (2017); Yen et al (2022)
Allen and Santomero (1997); Laeven and Levine (2009); Adusei (2015);
NPLit Total Non − Performing loan ratio
GDPt The growth rate at time (t) +
INFt The inflation rate at time (t) -
Defung and Yudarudding (2022); Alfiyan et al (2023)
Source: Synthesized by the author.
In chapter 3, the thesis presents the research method of the topic in the direction of quantitative analysis Accordingly, the quantitative analysis method is performed on an unbalanced panel dataset within the range of Vietnamese commercial banks in the period 2010 – 2023 Accordingly, the author proposes to build a financial stability measurement model suitable to practice in Viet Nam to draw conclusion about the impact of Fintech affecting the financial stability of Viet Nam commercial banks In addition, the author present quantitative analytical research methods to answer the research questions thereby achieving the research objectives of the thesis.
EMPIRICAL RESULT ANALYSIS
Descriptive statistical analysis
The study was carried out on a sample of 25 Vietnamese commercial banks in the period 2010 - 2023 The data to calculate the models are collected from audited financial statements of Vietnamese commercial banks from 2010 - 2023 and macroeconomic data from General Statistics Office Vietnam
Table 4.1 Statistical results of variables used in the research method
Variables Obs Mean Std Dev Min Max
(Source: Analysis result from STATA software)
According to Table 4.1, the average value of the bankruptcy risk coefficient (Z-Score) of 25 Vietnamese commercial banks from 2010 to 2023 is 1.9019, a standard deviation of 0.5585, the minimum and the maximum values are 0.9489 and 3.9671, respectively Saigon Banks for Industry & Trade (SGB) in 2010 and Viet A Commercial Joint Stock Banks (VAB) in 2017 As a result, it is clear that Vietnamese commercial banks have a large divergence when the distance between the two extremes is relatively long and the standard deviation of this factor is very large
The Fintech (FT) has an average value is 1.8962, standard deviation is 0.3199, the maximum value is 2.274 and the smallest value is 1.2788
The mean value of the return on Equity (ROE) is 0.1158, with a standard deviation of 0.0662 In which, the minimum value is 0.0020 and the maximum value is 0.2682, respectively National Citizen Commercial Joint Stock Banks (NVB) in 2015 and Asia Commercial Joint Stock Bank (ACB) in 2011
The mean value of the bank size (SIZE) is 18.8128, with a standard deviation of 1.1293 In which, the minimum value is 15.9227 and the maximum value is 21.5565, respectively Viet Capital Commercial Joint Stock Banks (BVB) in 2010 and Joint Stock Commercial Banks for Investment and Development of Viet Nam (BID) in 2023
It can be seen that the mean value of the non-performing loan ratio (NPL) is 0.0205, with the standard deviation of the non-performing loan ratio ratio of 0.0114 The largest value of the period from 2010 to 2023 is 0.0881 and the smallest is under 0.0001, respectively Saigon - Hanoi Commercial Joint Stock Banks (SHB) in 2012 and Tien Phong Commercial Joint Stock Banks (TPB) in 2010
The mean value of the capital adequacy ratio (CAP) is 0.0910, with a standard deviation of 0.0368 In which, the minimum value is 0.0406 and the maximum value is 0.2564, respectively Joint Stock Commercial Banks for Investment and Development of Viet Nam (BID) 2017 and Joint Stock Commercial Banks for Investment and Development of Viet Nam (KLB) in 2010
The mean value of the loan to deposit (LDR) is 0.8998, with a standard deviation of 0.2005 In which, the minimum value is 0.1257 and the maximum value is 1.4282, respectively Joint Stock Commercial Banks for Investment and Development of Viet Nam (BID) in 2014 and Viet Nam Commercial Joint Stock Banks for Private Enterprise (VPB) in 2021
Based on 271 observations during the period from 2010 to 2023, it can be seen that the average value of the asset growth rate (GROW) during this period is 0.2272, the standard deviation is 0.2094 The minimum value of the growth rate is 0.004 and the maximum value is 1.4701, respectively Tien Phong Commercial Joint Stock bank (SSB) in 2014 and Viet Capital Commercial Joint Stock Banks (BVB) in 2010
The mean value of the cost to income ratio (CIR) is 0.2926, with a standard deviation of 0.1434 In which, the minimum value is 0.0476 and the maximum value is 0.8384, respectively Joint Stock Commercial Banks for Foreign Trade of Viet Nam (CTG) in
2015 and Petrolimex Group Commercial Joint Stock Banks (PGB) in 2010
The mean value of the economic growth rate (GDP) is 0.0601 with a standard deviation of 0.0161 In which, the minimum value is 0.0256 belonging to 25 commercial banks in the observational scope of the study in 2021 and the maximum value is 0.0802 in 2018
The mean value of the inflation rate (INF) is 0.0498 with a standard deviation of 0.0433 In which, the minimum value is 0.0063 belonging to 27 joint stock commercial banks in the observation range of the study in 2015 and the maximum value is 0.1868 in
Correlation matrix analysis
Table 4.2 Correlation coefficients between research variables
Variables LnZscore FT ROE SIZE NPL CAP LDR GROW CIR GDP INF
(Source: Analysis result from STATA software)
According to Gujarati (2004), if the correlation between the independent variables exceeds 0.8, there is a high possibility of multicollinearity appearing in the model Then the sign of the regression coefficient in the model may be changed, leading to skewed research results Table 4.2 describes the correlation coefficient matrix between pairs of independent variables in the model, showing that the correlation coefficients of all pairs of independent variables are less than 0.8 Therefore, there is no multicollinearity between the variables In which FT, ROE, SIZE, NPL, CAP, LDR, GROW, CIR and INF are variables that have a negative correlation with the dependent variable LnZscore; while the remaining variables are positively correlated with LnZscore.
Multicollinearity testing
Multicollinearity is a phenomenon in which the independent variables in the model are linearly correlated with each other The research model ensures that no high multicollinearity occurs Multicollinearity can cause the standard error or confidence interval of the estimate to be large, and may even cause the estimate to be erroneous Multicollinearity occurs when the VIF value >10 according to Muda and et al
(2013) Therefore, the study has tested the hypothesis that there is no multicollinearity phenomenon by using VIF (Variance Inflation Factor) criteria
Table 4.3: Results of multicollinearity test
(Source: Analysis result from STATA software)
Table 4.3 shows the model results, the VIF of the independent variables is less than
10 so multicollinearity in the model is assessed as not serious Therefore, when the variables included in the model are considered suitable, these variables can be used to conduct research regression analysis.
The analysis of regression results
Table 4.4 Results of Pooled-OLS, FEM and REM
Model Pooled-OLS FEM REM
Variables Coef P-value Coef P-value Coef P-value
(Source: Analysis result from STATA software)
From the results of table 4.4, there are 2 variables that are not statistically significant, namely CAP and GDP in the research model The variables that are significant at the 1% level are FT, ROE, SIZE, NPL, LDR, GROW, CIR, and INF In which, the variables FT, ROE, NPL, CAP, GROW, CIR, GDP, INF have a negative correlation to the LnZscore variable, the other two variables are SIZE and LDR have a positive correlation to the independent variable LnZscore
According to R-squared = 0.5134 (51.34%) the independent variables interpreting the variation of the dependent variable is 51.34%
The FEM model results in 3 variable that is not statistically significant, which are SIZE, CAP and GDP, and the other variables are statistically significant In which, at the 1% significance level, the variables FT, ROE, NPL, LDR, GROW, CIR are statistically significant The remaining INF variable is statistically significant at the 5% level However, only the variable LDR have a positive correlation with the independent variable LnZscore, the rest of the variables FT, ROE, NPL, GROW, CIR and INF have a negative correlation with the independent variable LnZscore
From table 4.4, the R-squared of the FEM model is 0.5004, which means the independent variables can clarify 50.04% of the variation in LnZscore
In the REM model, the CAP and GDP variables are not statistically significant All the remaining variables are statistically significant in the REM model, in which the FT, ROE, NPL, GROW, CIR variables have a negative correlation to the LnZscore variable at the 1% statistical significance level, the INF macroeconomic variable also has a negative correlation to the LnZscore variable but at the 5% statistical significance level The other two variables, SIZE and LDR, have the statistical significance of 5% and 1% respectively, they also have a positive correlation with the independent variable LnZscore
The R-squared coefficient of the REM model is 0.5123 which means that the variation of the LnZscore variable can be explained by 51.23% by the independent variables.
Selection of estimation method
Chibar2(8) = 25.70 Prob>chibar2 = 0.0000 Prob>F=0.0000Chi2=0.9955>0.05 Prob>Chi2=0.0000 F = 0.0061 < 0.05, so the hypothesis H0 is rejected and the appropriate research model is the Fixed-effect model (FEM)
After determining that the FEM model is more suitable than the Pooled-OLS model, the author continues to perform Hausman test to determine the appropriate model between FEM and REM in the next step Hypothesis of Hausman test:
H 0 : The REM model is more suitable for the research variables
H 1 : The FEM model is more suitable for the research variables
Based on the analysis results from STATA software, it shows that P-value = 0.9955
> 0.05, so the hypothesis H0 is accepted, which means that the appropriate model to analyze the impact of independent variables on financial stability is the Random Effects Model (REM)
Finally, the author performs Breusch-Pagan test to determine the model fit between Pooled-OLS and REM Hypothesis of the test:
H 0 : The Pooled-OLS model is more suitable for the research variables
H 1 : The REM model is more suitable for the research variables
Based on the analysis results from STATA software, it shows that P-value = 0.000
< 0.05, so the hypothesis H1 is accepted, which means that the appropriate model to analyze the impact of independent variables on financial stability is the Random Effects Model (REM)
In conclusion, after using F-test, Breusch-Pagan and Hausman to determine which model is better The author agree that the appropriate model to analyze the impact on independence variables on financial stability is the Random Effects Model (REM).
Test of heteroskedasticity and model autocorrelation
After concluding that the FEM model is the most suitable model for the independent variable ROE, it is essential to check for heteroscedasticity Hypothesis put forward:
H 0 : The model does not have heteroscedasticity
Table 4.6 Modified Wald test Research model Heteroscedasticity
LnZscore chi2( 10) = 290.79 Prob > chi2 = 0.0000 The research model has heteroscedasticity
(Source: Analysis result from STATA software)
The results of the Modified Wald test show that Prob>chi2 = 0.0000 < 0.05, so the author rejects the hypothesis H0, which means that the research model has heteroscedasticity
To check whether the research model has autocorrelation phenomenon or not, the author applying Wooldridge Test to verify, the research test hypothesis is set as:
H0: Research model has no autocorrelation
H1: The research model has autocorrelation
Table 4.7 Wooldridge test result Research model Autocorrelation
F(1, 24)= 9.038 Prob > F = 0.0061 The research model has autocorrelation
(Source: Analysis result from STATA software)
After applying Wooldridge Test, the author’s research model obtained P-value 0.0061 < 0.05, the research model rejected the hypothesis H0, that is, the results of the research model are presented auto-correlation.
Overcoming the research model by FGLS method
Table 4.8 FGLS model troubleshooting result
Cross-sectional time-series FGLS regression
Coef St.Err t-value p-value [95%
(Source: Analysis result from STATA software)
From the table of FGLS test result, the author finds that the variables of Fintech company (FT), Return on Equity (ROE), Bank size (SIZE), Non-Performing Loan (NPL), Loan to deposit (LDR), Asset growth (GROW), Cost to income (CIR) and Inflation (INF) has a statistically significant, in which the variables FT, ROE, GROW, NPL, CIR and INF have a negative impact on LnZscore, the remaining variable SIZE have a positive relationship to LnZscore CAP and GDP are variables that are not statistically significant From the resulting table there is a model as follows:
LnZscore it = 2.4389**- 0.7753***FT it - 4.7672***ROE it + 0.1026***SIZE it - 7.9643***NPL it - 0.2767CAP it + 0.5737***LDR it - 0.7740***GROW it -
1.2817***CIR it - 1.8236GDP t - 2.2553***INF t + ԑ it
Endogenous variables testing
The author tests the phenomenon of endogenous variables with the hypothesis:
Table 4.9 Endogenous and exogenous variables in the research model
Variables P-value Endogenous variables Exogenous variables
(Source: Analysis result from STATA software)
From the data in table 4.9, it can be seen that endogenous variables include Return on Equity (ROE), Capital adequacy (CAP), Economic growth (GDP) Exogenous variables include the variables Fintech (FT), Bank size (SIZE), Non-performing loan (NPL), Loan to deposit (LDR), Asset growth (GROW), Cost to income (CIR).
GMM regression model method
After overcoming the research model through FGLS regression, the author continues to conduct endogenous control by GMM regression, the results from the GMM regression equation are presented in the following table:
Arellano-Bond test for AR(2) in first differences
Sargan test of overid restrictions 0.007
Hansen test of overid restrictions 0.730
LnZscore Coef St.Err t-value p-value [95%
(Source: Analysis result from STATA software)
Based on the results of GMM regression model, it shows that the number of tools in the regression model does not exceed the number of research groups (23