Determinants of credit risk the case of 24 commercial banks in vietnam

The factors include credit risk inprevious year, loan growth, bank size, return on assets, capital adequacy ratio, unemployment rate, GDP growth rate, inflation, real estate prices index

ĐẠI HỌC KINH TẾ TP HÒ CHÍ MINH

ÁO CÁO TỐNG KẾT

ĐÈ TÀI NGHIÊN cứu KHOA HỌC THAM GIA XÉT GIẢI THƯỞNG

“NHÀ NGHIÊN CỨU TRẺ UEH” NÁM 2024

DETERMINANTS OF CREDIT RISK: THE CASE

OF 24 COMMERCIAL BANKS IN VIETNAM

Thuộc nhóm chuyên ngành: Tài chính - Ngân hàng

(thuộc nhóm 1 trong 7 nhóm trong quy định)

Thành phố Hồ Chí Minh, Tháng 1/2024

Topic: ''Determinants of Credit Risk: The Case of 24 Commercial Banks in VietNam "

The research aims to analyze the factors affecting credit risk in Vietnamese commercial

banks, comprising 24 banks from 2009 to 2021 The factors include credit risk inprevious year, loan growth, bank size, return on assets, capital adequacy ratio,

unemployment rate, GDP growth rate, inflation, real estate prices index, and new

institutional factors such as law compliance index and corruption control index

Variables with a positive impact on credit risk include credit risk in previous year,the inflation rate, and the unemployment rate, return on assets, and capital adequacy ratio Conversely, variables with a negative impact on credit risk are GDP growth rate,

loan growth, and real estate price index Two variables, bank size and legal compliance

exhibit directional impact on credit risk depending on the measurement approach.Ultimately, corruption control index shows no statistical significance in influencing

credit risk

Based on theoretical studies in developed and developing countries, along with some

studies in Vietnam, the authors have constructed a model and hypothesis to identify new

factors influencing credit risk in 24 Vietnamese commercial banks from 2009 to 2021

Using a sample of 312 observations, the authors employ econometric methods to analyze

and synthesize the impact of these factors on credit risk in the Vietnamese commercial

banking system This analysis will lead to conclusions and recommendations for commercial banks, the State Bank, the Government, and relevant agencies to minimize

credit risk and propose measures to enhance profitability and maintain stability in the

sustainable economic development of the Vietnamese commercial banking system

Keywords: Credit risk, Non-performing loan ratio, Institutional factors, Corruption

control, Commercial banks.

2.3.1.5 Capital Adequacy Ratio ( CAR) 23

2.3.2.1 Unemployment rate (UNEMPLOYED) 24

2.3.2.2 Gross Domestic Product growth (GDP) 24

2.3.2.4 Real estate price index (REALES) 25

2.3.3 Regulatory factors 26

2.3.3.1 Compliance with the law index (LAW) 26

2.3.3 2 Control of Corruption index (CORRUPTION) 27

3.1.1 Dynamic model with the non-performing loans ratio (NPL) as dependent

3.4.6 Unemployment ratio (UNEMPLOYED) 38

3.4.7 Gross Domestic Product growth (GDP) 39

3.4.11 Control of Corruption index ( CORRUPTION) 41

4.1 The credit risk situation at commercial banks in Vietnam during the period 2009

4.4.1 Regression results in POLS, FEM, REM, and GMM method, with the

dependent variable being the non-performing loan (NPL) ratio 51

4.4 1.1 Regression of dynamic table data 51

4.4 ì 2 Regression model goodness of fit test. 52

4.4.1.4 SGMM method in dynamic panel data 55

4.4.2 Regression results in POLS, FEM, REM, and GMM method, with the

dependent variable being loan loss provision ratio (LLP) 57

4.4.2.1 Regression of dynamic table data 57

4.4.2.2 Regression model goodness of fit test. 58

4.4.2.4 SGMM method in dynamic panel data with loan loss provision (LLP)

5.2.1.3 Enhancing capital adequacy ratio 71

5.2.2.4 Promoting real estate market development 73

LIST OF TABLES AND CHARTS

Table 2.1 The specific provisions for 5 debts group 15

Table 2.2 Previous foreign studies 17

Table 2.3 Previous domestic studies 18

Table 3.1 Selected variables in the model 30

Table 4.2a Descriptive analysis 44

Table 4.3a Pearson Correlation Matrix with the Non-Performing Loan Ratio (NPL) as Dependent Variable 47

Table 4.3b Pearson Correlation Matrix with the dependent variable as Loan Loss Provision ratio (LLP) of Commercial Banks 48

Table 4.3c V1F in model with Non-Performing Loan ratio (NPL) as dependent variables at commercial banks in Vietnam from 2009 to 2021 50

Table 4.3d VIF in model with loan loss provision (LLP) as dependent variables at commercial banks in Vietnam from 2009 to 2021 50

Table 4.4a Regression results with the dependent variable being NPL ratio 51

Table 4.4b Statistical diagnosis with the dependent variable being the non-performing loan (NPL) ratio of commercial banks 53

Table 4.4c Endogeneity test results in the model with the dependent variable being the non-performing loan ratio (NPL) 54

Table 4.4d Results in SGMM model with non-performing loan (NPL) ratio as dependent variables 56

Table 4.4e Regression results with the dependenet variable being LLP ratio (LLP) 57

Table 4.4f Statistical diagnosis with the dependent variable being loan loss provision ratio (LLP) of commercial banks 59

Table 4.4g Results in endogeneity test with the dependent variable being the Loan Loss Provision Ratio (LLP) 60

Table 4.4h Results in SGMM model with loan loss provision (LLP) ratio as dependent variables 60

LIST OF FIGURES Figure 3.1 Research analysis framework 29

Figure 4.1 The Non-Performing Loan (NPL) ratio and Loan Loss Provision (LLP) ratio al commercial banks in Vietnam from 2009 to 2021 43

ABBREVIATIONS - TERMS

GDP Gross Domestic Product

GSO General Statistics Office

IMF International Monetary Fund

OECD Organization for Economic Cooperation and Development

POLS Pooled Ordinary Least Squares

SGMM System Generalized Method of Moments

CHAPTER 1: INTRODUCTION

1.1 Introduction

Lending operations is the primary business activity of banks to carry out credit,

generating about 60-70% of the bank's total income The bank serves as an intermediaryorganization in the circulation of capital, transferring funds from surplus areas to those

in need of utilization

According to the provisions of Clause 16, Article 4 of the Law on Credit Institutions

in 2010, credit granting activity means that the lender will deliver or commit to deliver

to the borrower a certain amount of money within a certain period of time to be used for

a specified purpose The agreement between the two parties will be based on theprinciple of repayment of both principal and interest

In Vietnam, in recent years, with great social changes and the establishment of many commercial banks, credit activities have become more competitive, along with potential

risks Once credit risk is high, which means bad debts increase, banks will likely face

capital shortages or low liquidity This can reduce the bank’s business profits, and the

worst scenario is bankruptcy Therefore, the issue of credit risk in Vietnamese banks has

been attracting the attention of the society, especially after the economic recovery period

and the Van Thinh Phat incident caused SCB to restructure and seize assets to ensure

debt recovery According to data from the State Bank of Vietnam (SBV), the bad debt ratio of commercial banks in Vietnam in 2009, 2010 and 2011 was 2.5%, 3% and 3.3%,

respectively, peaking in 2012 to 4.08% and the bad debt ratio of the whole system by

the end of February 2023 has reached 2.91%, up quite sharply from 2% at the end of

2022 and nearly 2 times more than at the end of 2021 It can be seen that the bad debt ratio increased in most banks Among 28 commercial banks, only 2 units recorded a

decrease in bad debts compared to the beginning of 2023 The rest all recorded high bad

debts, of which one unit had an increase in bad debts of 188%, nearly 3 times higher than the figure at the beginning of the year As noted, VPBank is currently the bank withthe largest bad debt with VND 31.864 billion, up 27% compared to the end of 2022, theratio of bad loans to outstanding loans is more than 3% The cause of bad debts of banks

can come from many related factors such as macroeconomics, micro and specific factors

of banks Therefore, the study of factors affecting credit risk is very meaningful in thecontext that credit risk is the biggest risk that banks face and also the main cause of the

global economic crisis in the recent period The above research plan will help bankadministrators identify the impact of factors on credit risk, thereby better controllingbanking activities during the economic crisis, recovering after 2 years of the pandemic

Up to now, many countries have conducted research on credit risk at commercial

banks, including overseas studies by Salas & Saurina (2002), Rajan & Dhal (2003),

Louzis & Associates (2012), Ahlem Selma Messai (2013), Castro (2013), Marijana Curak & Associates (2013), Bucur & Associates (2014), Tehulu & Associates (2014)

and domestic studies such as: Đào Thị Thanh Bỉnh & Đồ Vân Anh (2013), Đồ QuỳnhAnh & Nguyền Đức Hùng (2013), Võ Thị Quý & Bùi Ngọc Toàn (2014), Nguyền ThịNgọc Diộp & Nguyen Minh Kiều (2015), Nguyễn Văn Thép & Nguyền Thị Bích Phượng (2016), Nguyền Văn Thuận & Dương Hồng Ngọc (2015), Bùi Duy Tùng &Đặng Thị Bạch Vân (2015), Phạm Dương Phương Thảo & Nguyền Linh Đan (2018), Nguyền Thị Như Quỳnh et al (2018) However, all the above studies have not

comprehensively delved into the topic of factors affecting credit risk of commercial banks in the context of Vietnam Most focus on the influence of two groups of factors: bank-specific groups and macroeconomic factors In addition, all of the above­

mentioned studies are: (i) Excluding a comprehensive consideration of all factors

influencing credit risk through quantitative methods, including internal bank factors, macroeconomic factors and institutional factors; (ii) Not accounting for the impacts of

real estate price factors, corruption control factors, and law compliance factors.; (iii) The two-step System General Method of Moments (SGMM) regression method has not been

used in estimating factors affecting credit risk at commercial banks

Credit risk always exists in the business activities of Vietnamese commercial banks

Therefore, in order to ensure safety in credit activities, factors affecting credit risk at Vietnamese commercial banks need to be clearly identified and carefully measured As

a result, it is possible to suggest solutions and tools to prevent and limit maximum credit risk, improve efficiency for Vietnamese commercial banks, towards a highly liquid economy, and improve credit quality

Stemming from the above-mentioned scientific and practical urgency, the authorschose to study the topic "Determinants of Credit Risk: The Case of 24 Commercial

Banks in Viet Nam" with new contributions on approaches, providing more empirical

evidence into the previous theoretical framework

1.2 Research Objective

V General research objective: Identify factors affecting credit risk of commercial banks in Vietnam

/ Specific research objectives:

• Objective 1: Measure and estimate the impact of factors affecting credit risk of commercial banks

• Objective 2: Based on the above level of impact, the authors proposerecommendations to prevent and minimize credit risk for Vietnamese commercial banks

1.3 Research Materials and Methods

1.3.1 Research Materials

The research object of the topic is the factors influencing credit risk in the commercial

banking system of Vietnam However, there are various methods of assessment and

different evaluation criteria for credit risk, such as bad debt ratio, overdue debt ratio, loan loss provision In this paper, the author chose the Non-Performing Loan ratio and

the Loan Loss Provision ratio as the representative variables of the research subject, credit risk, according to the studies of authors Sukrishnalall Pasha and Tarron Khemraj(2009); Louzis et al (2010); Nabila Zribi and Younes Boujelbene (2011), because these two indicators most clearly show the credit quality situation of banks and they alsoreflect the ability of credit management in credit granting and debt recovery of the Vietnamese commercial banking system during the study

1.3.2 Data Sample

Sample observation: The data was collected based on data from 24 commercial banks between 2009 and 2021 The reason for the choice of this observation pattern:

• Spatial scope: The authors only chose to study 24 commercial banks form 2009

to 2021 because in this period, some banks restructured, others did not provide enoughdata, so the authors could not collect all data for the entire commercial bank in Vietnamduring this period These 24 banks accounted for about 90% of charter capital and 80%

of the total number of commercial banks at the time of the study

• Time scope: First, non-performing loan data before 2009 was incomplete

Moreover, during the 12-year period 2009 to 2021, the bad debt situation of Vietnam's commercial banking system had many fluctuations Therefore, the author chose thisscope to carry out research on the topic

Data source:

• Independent variables: data is extracted from the annual financial statements of

the State Bank of Vietnam (SBV), General Statistics Office of Vietnam (GSO), International Financial Statistics (IFS), The Asian Development Bank (ADB), International Monetary Fund (IMF), World Bank (WB)

• Dependent variables: the authors got information from the sources of financial

statements and annual reports of appraisal statistical agencies

Sample size: 24 commercial banks in the territory of Vietnam, collected from 2009

-2021 with a total of 312 observations Therefore, the sample is sufficiently statistically

representative

1.4 Methodology

To solve the problem that the topic poses, the authors used a combination of secondary data synthesis and analysis methods, correlation analysis and common panel data regression methods including: Pooled OLS (POLS, Fixed Effect Model (FEM) and

Random Effect Model (REM) to analyze causal relationships In addition, to overcomethe potential endogenous issue, the authors used the System Generalized Method ofMoments (SGMM), as suggested by Arellano and Bover (1995), Blundell and Bond (1998)

1.5 Research significance

The research topic complemented and completed the theoretical basis of research

methods on determinants of credit risk in Vietnam's commercial banking system,

providing more empirical evidence on the relationship between credit risk in commercial

banking activities with macro factors in the economy and specific factors at banks

Finding out the institutional factors, macroeconomic factors and internal banking factors

affecting credit risk at Vietnamese commercial banks In addition, the team analyzed the different influences of these factors on credit risk al Vietnamese commercial banks

during the study period The research results are the basis for the authors to propose

solutions to limit credit risk at Vietnamese commercial banks

Therefore, it can be said that the topic "Determinants of Credit Risk: The Case of 24 Commercial Banks in Vietnam" is a new and suitable topic to the development needs aswell as deep integration of our country with the current complex macro situation

Chapter 2 - Theoretical basis

Presenting and summarizing the theoretical foundations used in research papers

worldwide and in Vietnam, in addition to a preliminary introduction of factors affecting

loan loss provision

Chapter 3 - Model and Data

Specifically presenting the meaning of the variables used in the model, explaining

hypotheses, research methods performed

Chapter 4 - Research results

Presenting empirical evidence used for research objectives, including: descriptive

statistics, correlation analysis between variables in the model, testing model violationsand explaining the suitability of the model to come up with the most optimal model,

then testing the research hypothesis with regression models and analyzing the impact of

factors to credit risk

Chapter 5 - Conclusions

Concluding the research issue, including the content and results of the study, outlining

the remaining limitations in the research paper, suggesting directions for developing the topic and providing some policy suggestions to help banking agencies andadministrators assess the impact of factors on credit risk at Vietnamese commercial

banks, thereby proposing specific and sustainable solutions to prevent and limit bad debt risk

CHAPTER 2: LITERATURE REVIEW

2.1 Theoretical basis

2.1.1 Definition

According to the State Bank of Vietnam (2005), Credit risk in banking is the potential

loss to a bank when borrowers fail to fulfill or are unable to fulfill their repayment obligations as committed

"Bad debt," also known as "non-performing loan" (NPL) or "doubtful debt," as

defined by the World Bank, refers to loans that fall below standard, may be overdue,

and are doubted for both repayment and recovery capability by the creditor, typically

occurring when debtors declare bankruptcy or disperse their assets

Credit risk in banks is sometimes assessed through the ratio of Non-Performing Loans (NPLs), calculated as the ratio of total NPLs to total outstanding loans (Fadzlan Sufian

& Roy-faizal R Chong (2008); Nguyền Thị Thái Hưng (2012); Rasidah M Said & Mohd H Tumin (2011); Somanadevi Thiagarajan & others, (2011); Tobias Olweny &Themba M Shipho, (2011)) Some other studies measure credit risk through the ratio of

credit risk provisions to total bank assets (Luc Laeven & Giovanni Majnoni (2002), Nabila Zribi & Younes Boujelbene (2011)), as loans constitute a significant portion oftotal assets, allowing the direct use of total asset value to calculate credit risk

2.1.2 Credit risk measurement criteria in commercial banks

2 1 2.1 Non-performing loans

Non-performing loans (NPLs) are debts that become difficult to collect when borrowers are unable to repay them according to the terms specified in the credit

agreement Specifically, if the repayment is overdue for more than 3 months (90 days),

it is considered a non-performing loan

According to Article 8, Clause 3 of Circular 11/2021/Directive-State Bank of Vietnam (SBV), non-performing loans (NPLs) are classified as internal NPLs Loans falling under groups 3, 4, and 5 are considered difficult-to-collect debts, as regulated in Article 10 of Circular 11/2021/Directive-SBV

Non-performing loans are most prominently reflected through various indicators:

Bad debt =

Total outstanding balance

Bad debt to equity ratio = Bad debt

Total Bank's Equity

Bad debt to contingent fund Bad debt

Contingent fund

Non-performing loan ratio (NPL) is a critical indicator for measuring credit risk at

financial institutions The NPL ratio moves in the same direction as credit risk As theNPL ratio increases, so does credit risk, leading to customers' inability to repay their

debts to the bank, resulting in reduced revenue and profits for the bank In the worst­

case scenario, this situation can lead to bankruptcy

Current Loan Classification in Vietnam

According to Decision No 493/2005/Decision-SBV, 18/2007/Decision-SB V, issued

by the Governor of the State Bank of Vietnam is as follows:

Group 1: Standard debt - Loans that are within the due date and are assessed by credit institutions to have the full ability to recover both principal and interest on time

Overdue loans of less than 10 days are assessed to have the full ability to recover both

principal and interest within the remaining time frame Assessed to have the full ability

to recover both principal and interest on time

Group 2: Watch-list debt Overdue loans ranging from 10 to 90 days; loans with the

first-time adjusted repayment period (for corporate customers, credit institutions must have customer assessment records regarding their ability to fully repay the principal and

interest on the first adjusted repayment period) Assessed to have the ability to fully

recover both principal and interest, with signs of reduced repayment capability from

Group 4: Doubtful debt - Overdue loans ranging from 181 to 360 days; loans withthe first-time adjusted repayment period overdue for less than 90 days according to the

first adjusted repayment period; loans with the second adjusted repayment period

Assessed to have a high risk of loss

Group 5: Loss debt - Overdue loans exceeding 360 days; loans with the first-timeadjusted repayment period overdue for more than 91 days according to the first adjustedrepayment period; loans with the second adjusted repayment period overdue according

to the second adjusted repayment period; loans with the third or subsequent adjusted

repayment period, including those not yet overdue or already overdue; non-performing

loans awaiting resolution Assessed to have no recovery ability, with a high likelihood

of capital loss

2.1.2.2 Loan Loss Provision

Provisions for potential losses associated with specific debts accurately reflect Credit

Risk As it is considered a cost for impaired assets Some studies on Credit Risk have utilized the annual Loan Loss Provision ratio (LLP) compared to the annual total

outstanding loans as an independent variable In terms of risk management, the LLP

ratio is a policy set by banks to address potential Credit Risk occurrences, thus serving

as a tool to control Credit Risk

In Vietnam, according to Article 2, Decision No 493/2005/Decìsion - SBV on the

Classification of Debts, Provisioning, and Use of Reserves for Dealing with Credit Risk

in Banking Activities of Credit Institutions, the Provision for Credit Risk Reserve is

understood as follows: The Credit Risk Reserve is an amount set aside to reserve forlosses that may occur when customers fail to fulfill their contractual obligations

The level of provision for credit risk in banking activities is specified in Article 12 of

Circular No 02/201 3/Directive - SBV, which is outlined in the table below:

Table 2.1 The specific provisions for 5 debts group

Source: Article 12 of Circular No 02/2013/Directive - SBV

According to Article 10 ofCircular No 11/2021/Directive - SBV the non-performing

loans ratio (NPLs) include debts from Group 3 to Group 5 However, provisioning forCredit Risk Reserve (Credit risk reserve) must be set aside starting from Group 2 Therefore, when comparing NPLs with total outstanding loans or total assets, it does not accurately reflect the nature of Credit Risk Hence, the provisioning for Credit riskreserve is based on the classification of different types of debts at each bank Creditinstitutions rely on qualitative and quantitative criteria to assess the risk level of loans

and off-balance sheet commitments, thereby categorizing debts into appropriate groups

The principle of using Credit risk reserve is to allocate specific provisions for each

debt first, utilize collateral assets for debt recovery, and only resort to general provisions

if asset sales fail to fully recover the debt Each bank needs to have an appropriate

provisioning calculation method that is sufficient to offset risks while avoiding excessive costs that could impact net income The indices that reflect Credit RiskReserve include:

Loan loss reserve ratio= -—————xioo%

Gross loan portfolio

The coefficient of the ability to offset lost loans= ——7—- —X 100%

Total debt write-off

Loan loss reserve

The credit risk mitigation coefficient= - —————

Irrecoverable debts

2.2 Empirical studies

2.2 Ỉ Foreign studies

In Spain, Salas & Saurina (2002) pioneered a study on how micro and

macroeconomic variables impact non-performing loans during the period of 1985-1997

In this study, the authors found that bank scale has an inverse effect on the non­performing loan ratio, loan growth has a positive impact on credit risk, and GDP growthrate has an inverse impact on the loan ratio During the credit boom, high-risk borrowers

may obtain loans even with low-quality collateral In Rajan & Dhal's study (2003), analyzing macroeconomic factors affecting non-performing bank loans separately, it was found that GDP growth rate and credit risk have a positive correlation, depending

on specific business environments This suggests the need to expand other macro

variables for a broader scope of research Moving towards a broader research scope,

Festic & colleagues (2011) studied credit risk and macro variables in 5 EU member

countries The results also supported the hypothesis that loan growth is positively

correlated with the non-performing loan ratio Additionally, they mentioned the decrease in economic activity and lack of supervision Furthermore, Louzis &colleagues (2012), using GMM estimation tool, discovered that in Greek banks, besides internal bank variables, macroeconomic variables also influence bank non-performingloans Expanding the scope to 85 banks in 3 countries during the period of 2004-2008(Messai, 2013), the relationship between ROA, GDP growth rale, and credit riskassessment criteria was analyzed Findings showed that increasing unemployment ratesand interest rates have negative impacts on non-performing loans due to lending

environments and borrowers losing repayment capacity due to objective factors

Moreover, there are other studies using panel data and revolving around macroeconomic

factors and bank internal variables, summarized in the table below

Table 2.2 Previous foreign studies

Salas &

Saurina

(2002)

Using FEM, REM to study micro

and macroeconomic variables’

impacting on bank non-performing

loans in Spain during 1985-1997

Inverse relationship: Bank scale, GDP growth

Positive relationship: Loan growth

Rajan &

Dhal

(2003)

Using FEM, REM to study factors

influencing non-performing loans of Indian banks during 2003-2008

Inverse impact of bank scale,

positive impact of GDP growth

on non-performing loans, when

business environment is good,

non-performing loans decrease

Berge &

Boye

(2007)

GMM method researched factors

affecting loan problems of Nordicbanking system during 1993-2005

Troubled loans affect real

interest rates and domestic unemployment rate

Economic activity decline, lack

of supervision, and financial

loan growth establish a positive relationship with non­

performing loans

Zribi

Boujelben

e(2011)

Using FEM, REM to study micro

and macroeconomic variablescontrolling credit risk of 10 Tunisian banks during 1995-2008

Regulations ensuring capitalsafety, ownership structure,

profit, and rapid macroeconomic indicators like

GDP growth, exchange rate,

Source: Synthesized by the author from reference materials

inflation, and interest rate

influence credit risk

Louzis &

colleagues

(2012)

Using GMM method with Dynamic

Panel Data to study macroeconomicfactors and bank variables affecting non-performing loans in Greek

banks during 2003-2009

Troubled loans in Greek

banking system are explained

by macroeconomic variables (real GDP growth, interest rates, public debt, and

Using FEM, REM method to study

factors impacting non-performingloans of 85 banks in Italy, Greece,

and Spain during 2004-2008

ROA, GDP growth rate

inversely affect non-performingloans, while unemployment and interest rates positively impact

credit risk of 5 European banks

(Greece, Ireland, Portugal, Spain,and Italy)

GDP growth rate, house price

index, interest rate,

unemployment rate, realexchange rate, and loan growthall positively correlate with

loans during 2003-2010

Inverse relationship between non-performing loan ratio and bank size

2005-2011

Except for inflation rate, all macroeconomic variables studied impact non-performing

loan ratio

2.2.2 Domestic studies

Table 2 3 Previous domestic studies

Đào Thị

Thanh Bình &

Đỗ Vân Anh

(2013)

Using FEM, REM method and

datasets of 14 Vietnamese banks

to investigate factors influencing

credit risk of Vietnamese banks

Bank scale positively affects

credit risk Conversely, ROE ratio negatively affects

credit risk

Using REM, FEM, GMM models

to study factors affecting credit risk of Vietnamese banks during

2005-2011

GDP growth, inflation

impact credit risk Past

credit risk affects the subsequent year, bank scale correlates positively with

banks examine factors influencing

credit risk of banks during 2009­2012

GDP growth rate, loan

growth with a lag of 1 year

significantly affect credit

Using Dynamic Panel Data

Methods and GMM regression

method to investigate the

influence of bank internal factors

on credit risk of Vietnamese banks during 2004-2014

Equity capital on assets and

after-tax profit have aninverse relationship with

credit risk Shareholding

ratio correlates positively

with credit risk

credit risk at Vietnamese banks

(including 29 banks during 2007­

2014)

Operating expenses, economic growth rate, andbank scale correlate

positively with credit risk,

while the minimum capitaladequacy ratio (CAR)

inversely correlates with

Using FEM to analyzes factors

affecting credit risk provisions of

on assets, loan growth rate,and customer loan-to-

deposit ratio credit risk

provisions ratio of banks.Phạm Dương

Phương Tháơ

& Nguyễn

Using GMM method and analysis

of data from 27 commercial banks

to examine the impact of

Operating expenses, provisioning for risk, after­

tax profit on equity, and

Linh Đan

(2018)

macroeconomic and bank

characteristics on the credit risk

ratio of Vietnamese banks during

2005-2016

economic growth rate all influence the credit risk ratio

of banks

Source: Synthesized by the author from reference materials

Đào Thi Thanh Bình & Đô Vân Anh (2013) laid the foundation for researching factors

influencing credit risk in the context of Vietnam The scale of banks at that time had a positive impact on the non-performing loan ratio This is because larger banks tend to

have tighter credit management, leading to more reputable lending and therefore lower risk Following this, Đồ Quỳnh Anh & Nguyền Đức Hùng (2013) and Võ Thị Quý &

Bùi Ngọc Toán (2014) added variables such as non-performing loan value and loan

growth with a one-year lag, identifying them as significant factors worth considering.Expanding the scope of research, Nguyen Văn Thcp & Nguyen Thị Bích Phượng (2016) examined 29 commercial banks during the economic crisis in 2008, revealing that

operating expenses and economic growth rate positively correlate with credit risk Asthe market becomes more volatile, credit risk increases due to loss-making businessactivities, affecting the income of borrowers

Starting from the urgent need both scientifically and practically, the authors anticipate

providing additional evidence on new factors impacting the credit risk of Vietnamese commercial banks These factors include political stability, real estate prices index, andlaw compliance index They propose to employ the System GMM (SGMM) two-step

method for the first time to analyze the influence of internal bank factors, macroeconomic factors, and regulator}' factors on credit risk

This approach aims to enhance the understanding of the dynamics of credit risk in

Vietnamese commercial banks by incorporating a comprehensive set of variables andutilizing an advanced econometric technique By examining both internal and external factors alongside regulatory compliance, the study seeks to provide valuable insights for

policy makers, bank managers, and researchers in assessing and managing credit riskeffectively in the Vietnamese banking sector

2.3 Determinants

Through the process of reviewing studies worldwide in general and in Vietnam in

particular, it is evident that there are two main groups of factors influencing Credit risk

at commercial banks, including: bank-specific factors and macroeconomic factors

2.3.1 Bank-specific variables

2 3 1.1 Credit risk of the previous year (NPLi t_ lt LLPịt_f)

According to the study conducted by Somanadevi Thiagarajan and colleagues (2011),

the factors influencing Credit risk at banks in India were investigated by collecting data from 22 state-owned banks and 15 private-owned banks over the period of 2001 - 2010

The project showed that Credit risk in the past has a positive impact on the Credit risk

of banks in the current year with a lag of one year This result was also found by DanielFoos et al (2010), Abhiman Das & Saibal Ghosh (2007), and Gabriel Jimenez & Jesus

Saurina (2006) Furthermore, anotherstudy suggests that poor management of bad debts

in the previous year, as indicated by the bad debt ratio of the previous year according to Marki & Associates (2014), is a reason for the increase in bad debts in the current year,leading to difficulties in recovering difficult debts Moreover, if bad debts from theprevious year remain unresolved, it could lead to a high increase in bad debts in the current year (Nguyền Thị Như Quỳnh et al (2018)) This indicates that the credit quality

of the previous year declines, leading to an increase in bad debts in the following year

and vice versa Because a high level of bad debts in the past indicates poor risk

management in lending, leading to an increase in bad debts in the present Additionally,

a shock related to bad debts has long-term impacts on the banking system if not

addressed promptly Võ Thị Quý & Bùi Ngọc Toàn (2014) evaluated Credit risk in thepast with a lag of 1 year, which has a similar impact to Credit risk It can be seen that

Credit risk from the past has not been completely eliminated but will be transferred and

have a significant impact until the following year

2 3.1.2 Loan Growth (LG)

Loan growth is also one of the important factors that need to be considered

According to the research of Gabriel Jimenez and Jesus Saurina (2006), loan growth rate has a positive impact on bad debts The explanation for this correlation is that during

credit booms, high-risk borrowers may obtain loans even if the collateral quality is poor.

Bad debts will significantly increase in the future due to the inability to address

low-quality credit loans Additionally, studies by Sukrishnalall Pasha and TaiTon Khemraj

(2009), as well as Louzis et al (2010), suggest that loan growth rate has an inverseimpact on bad debts When financial institutions experience rapid loan growth,

businesses have easier access to loans, expand their business operations, and have a higher ability to repay debts Therefore, the likelihood of banks bearing bad debts is low Furthermore, Ahlem s M and colleagues (2013) found evidence demonstrating that loan growth rate does not affect bad debts in the banking systems of three countries:Spain, Italy, and Greece

2 3.1.3 Bank Size (SIZE)

Rajiv Ranjan and Sarat Chandra Dhal (2003) found that bank size is inversely related

to the bad debt ratio, contrary to s Pasha and T Khemraj (2009); Nabila Zribi andYounes Boujelbene (2011) Bank size is the market value of the bank, often measured

by the logarithm of the total loan portfolio of the bank to adjust this variable to a

comparable value with the remaining variables in the model Bank size affects bad debts

in both positive and negative directions Larger banks often manage bad debts more

effectively due to their ability to diversify loan portfolios along with superior RRTD management capabilities (Das & Saibal, 2007) However, larger banks are also willing

to accept higher risks because of expectations of government protection, so the bad debt

ratio may be higher (Nguyền Thùy Dương & Tran Thị Thu Hương, 2017) Therefore, the research group believes that larger banks do not always screen loan customers more

effectively than smaller banks

2.3.1.4 Return on Assets (ROA)

The profitability of a bank is usually measured primarily by two variables: return on

equity (ROE) and return on assets (ROA) However, many studies indicate that ROA is

the key indicator of a bank's profitability ROA measures the after-tax profit on total

assets of the bank and is used to assess the efficiency in organizing and managing business operations of the enterprise Furthermore, ROA indicates the profit per unit of

assets, reflecting the bank's ability to manage resources efficiently According to the

research by Misman & Ahmad (2011) and Mustafa & colleagues (2012), there exists an

inverse relationship between the LLP ratio and ROA This can be explained as follows:

during economic downturns, banks may experience recession, with the risk ofpoor loans

not being recovered, leading to a higher provision for LLP because the quality of loan

portfolios deteriorates, thereby reducing the return on assets With the study by Hu et al

(2004), banks with high profits are less likely to engage in risky activities because these

banks do not feel pressure to generate profits Additionally, banks with high profits tend

to select customers with good financial capabilities and low risks Therefore, as profits increase, the likelihood of bank managers participating in risky investment projectsdecreases, leading to a decrease in loans transferred to bad debts Conversely, when

banks perform poorly, they tend to engage in risky lending activities because managers

are pressured to generate profits in the short term, increasing the likelihood of loans turning into bad debts and thus increasing the bad debt ratio at the bank:

Net Income

Return on Assets = - —-—

-Average Total Assets

2.3.1.5 Capital Adecptacy Ratio (CAR)

Some studies suggest that the minimum capital adequacy ratio (CAR) affects the bad debt ratio CAR is expressed as equity capital/total risk-adjusted assets and is a regulatory indicator in many countries to ensure stability in the credit system Kim &Santomero (1988) argue that regulatory indicators cause banks to change the structure

of their asset portfolios towards less risk, thus minimizing bad debts Through the

research of Zribi & Boujelbene (2011) on the credit risk control capacity of 10 Tunisian

banks during the period 1995-2008, it was found that larger banks with higher CAR ratios than prescribed had lower bad debt ratios than the rest of the group In contrast, studies by Nguyền Văn Thép & Nguyền Thị Bích Phượng (2016) and Vò Hồng Đức et

ai (2014) suggest that CAR has a positive relationship with credit risk because whenCAR is high, banks assert their ability to compensate for risks beyond the safetythreshold (CAR regulated by the State Bank of Vietnam - 9%) As a result, banks may become complacent and lack customer screening, leading to an increase in credit risk

2.3.2 Macroeconomic variables

2.3.2 J Unemployment rate (UNEMPLOYED)

The unemployment rate is one of the factors directly related to the ability of individuals and businesses to generate cash flow to repay debt- Therefore, researchers

are interested in the relationship between the unemployment rate and credit risk In Vietnam, there are several measures of the unemployment rate, such as the labor force

participation rate, the unemployment rate, and the underemployment rate The most commonly used indicator to measure the unemployment rate is the percentage of people

in the labor force who arc unemployed When unemployment occurs, the income of

borrowers decreases, leading to a reduced ability to repay both principal and interest on loans, thereby increasing the bank's bad debt ratio (Filip 2015) Studies by Bucur et al

(2014) and Castro (2013) also show a positive relationship between the credit risk ratio

and the unemployment rate However, in Vietnam, research by Nguyen Thị Như Quỳnh

et al (2018) on the factors affecting bad debts at Vietnamese commercial banks,

including 25 commercial banks during the period 2006 - 2016, yielded opposite results:

the unemployment rate had an inverse relationship with credit risk This can be

explained as follows: The proportion of consumer loans to total credit outstanding is

still relatively low, and the unemployment rate in Vietnam is also low Therefore,

theoretically, when unemployment increases and the number of unemployed individuals with little credit transactions with banks is low the risk of bad debt decreases

2.3.2.2 Gross Domestic Product growth (GDP)

When it comes to economic growth factors, most previous studies have used the Gross Domestic Product growth rate (GDP) as a measure to assess bank credit risk The relationship between economic growth and credit risk has been extensively studied indocuments related to economic cycles These empirical studies have found that the GDPgrowth rale has an inverse relationship with the non-performing loan (NPL) ratio In a developing economy, non-performing loans only account for a small proportion because individuals and businesses have sufficient income and revenue to repay debts on time

Therefore, banks continue to expand credit without considering the repayment ability of

borrowers In an economic recession, the ability of customers to repay debts decreases, leading to an increase in non-performing loans, which has adverse consequences for

banks (Ahlem s M et al., 2013) According to the study by Louzis et al (2010), during

an economic recession, a significant decrease in the GDP growth rate has a substantial

impact on the ability of individuals and businesses to repay debts, resulting in a high non-performing loan ratio Furthermore, Pasha s and Khemraj T (2009) reached a

similar conclusion when studying non-performing loans in the banking system in

Guyana When the economy experiences strong growth, the non-performing loan ratio

decreases Abhiman Das & Saibal Ghosh (2007) conducted a study on non-performing

loans in India and also found that the GDP growth rate, as a macroeconomic factor,

significantly affects credit risk

2 3.2.3 Inflation rate (INFL)

In Vietnam, to measure inflation, we often use the Consumer Price Index (CPI) The

CPI is a percentage-based index reflecting the relative change in prices of goods over time, so inflation is also a factor worth considering The positive correlation between inflation and credit risk has been found by Gunsel (2011) when studying 24 banks in

Cyprus during the period 1984 - 2008 The author theorized that when the economy

experiences high inflation rates, banks face difficulties in assessing credit risk inbusinesses, leading to a decrease in the quality of borrowing customers and an increase

in non-performing loans Similarly, Nkusu's study (2011) also found that an increase inthe inflation index leads to an increase in non-performing loans However, contrary tothis viewpoint, Nabila Zribi and Younes Boujelbene (2011) argued that the credit risk

of the banking system in Tunisia depends on inflation and asserted that inflation has anadverse effect on credit risk

2.3.2.4 Real estate price index (REALES)

The volatility in the real estate market contributes to the credit risk of banks, as real

estate serves both as a lending target and as collateral The boom in the real estate market

leads to an expansion in lending activities, with banks providing more loans On the other hand, when banks hold more collateral assets than the value of the loans, they tend

to lower their lending standards to increase credit value

In summary, a review of previous studies shows that the factors influencing credit risk at commercial banks include two main groups:

(i) Bank-specific factors: Past credit risk, loan growth, bank scale, profitability, minimum capital adequacy ratio (Salas & Saurina, 2002; Rajan & Dhal, 2003; Berge &Boye, 2007; Festic et al., 2011; Boujelbene, 2011; Louzis et al., 2012; Messai, 2013;

Castro, 2013; Curak et al., 2013; Bucur el al., 2014; Tehulu et al., 2014; Chaibi & Ftiti,

2015)

(ii) Macroeconomic factors: Economic growth, inflation, unemployment (Berge &Boye, 2007; Castro, 2013; Tehulu et al., 2014) Additionally, studies by Grossman (2001), Eichler (2016), Ashraf (2017), and Ozili (2018) indicate that improving

regulatory environments has a positive impact on the stability of the banking system Therefore, it can be seen that credit risk al commercial banks, in addition to being influenced by bank-specific and macroeconomic factors, is also influenced by certain regulatory factors

2.3.3 Regulatory factors

2.3.3.1 Compliance with the law index (LAW)

Compliance with the law plays a crucial role in reducing credit risk and creating a clear and stable business environment Firstly, when financial institutions adhere to legal

regulations and laws related to business operations, they become more trustworthy inthe eyes of investors and the market This can reduce credit risk, thereby helping

financial institutions and individuals access loans more favorably

Moreover, non-compliance with the law can lead to serious legal consequences,

including penalties from regulatory authorities and reputational damage This can result

in credit risk as financial institutions incur significant legal costs and the risk of

reputation loss, affecting their ability to borrow and their cost of capital

Furthermore, implementing risk management standards and procedures can help reduce credit risk by creating strong control mechanisms and improving risk prediction

and assessment

Lastly, investors typically prefer markets and financial institutions where there is a

stable and transparent legal environment Compliance with the law can create trust and attractiveness for investors, thereby reducing credit risk by attracting investment capitaland reducing the cost of capital

In summary, law compliance plays a crucial role in reducing credit risk by fostering

a stable, specific, and highly predictable financial business environment

2 3.3.2 Control of Corruption index (CORRUPTION)

When government agencies enforce anti-corruption measures vigorously, they oftenenhance responsibility and transparency in financial management and businessoperations This can help reduce credit risk by creating a fair competitive businessenvironment Additionally, corruption often accompanies the conduct of illegal businessactivities Pushing enterprises to comply with laws and regulations helps control

corruption and reduces the financial legal risks that financial institutions must face when

collaborating Corruption can increase the risk of financial instability by weakening the

financial and economic system Controlling corruption also helps enhance the stability

of the financial system, reducing the risk of credit loss and economic recession

Therefore, controlling corruption can help reduce credit risk by creating a stable,

transparent, and law-abiding financial business environment

2.4 Chapter Summary

In this chapter, the authors have presented theoretical foundations regarding credit

risk at commercial banks (CBs), along with factors influencing credit risk at CBs.Additionally, we have reviewed previous studies, both domestic and international, related to this topic Based on the synthesis of theoretical foundations, we, the authors,

have proposed an analytical framework of factors influencing credit risk at CBs This

framework serves as the foundation to construct the research model in Chapter 3

CHAPTER 3: MODEL AND DATA HYPOTHESES

3.1 Panel data regression models

Based on studies by Salas & Saurina (2002), Rajan & Dhal (2003), Berge & Boye

(2007), Boujelbene (2011), Louzis & colleagues (2012), Messai (2013), Curak &colleagues (2013), Đào Thị Thanh Bỉnh & Đồ Vân Anh (2013), Đồ Quỳnh Anh &

Nguyền Đức Hùng (2013), Bùi Duy Tùng và Đặng Thị Bạch Vân (2015), Nguyền Vãn

Thép & Nguyền Thị Bích Phượng (2016), Nguyền Thị Như Quỳnh et al (2018), Phạm

Dương Phương Thảo & Nguyên Linh Đan (2018), the above researches suggest that the non-performing loan (NPL) ratio can be used to assess the level of credit risk at Vietnamese commercial banks

According to Zribi Boujelbene (2011), Tehulu et al (2014), Hasna Chaibi & Zied

Ftiti (2015), Võ Thị Quý & Bùi Ngọc Toản (2014) Nguyền Thị Ngọc Diệp & Nguyền

Minh Kiêu (2015), Nguyên Văn Thuận & Dương Hỏng Ngọc (2015), the results show

that credit risk in commercial banks can be assessed through banks' loan loss provision ratio

Moreover, the authors also found that the credit risk of the previous year can affect

the credit risk of this year, hence the decision to use the lagged variable as an independent variable in the regression model Therefore, the authors used two dependentvariables in two dynamic table models, the Non-Performing Loan ratio and the Loan

Loss Provision ratio (LLP)

Dependent variables Independent variables

Credit risk in previous year (NPLi.t-1 or LLPij-1)

Loan growth (LGtO

Bank size (SIZEi.t)

Return on assets (ROA,t)

Capital adequacy ratio (CARụ)

GDP growth (GDPt) Inflation rate (INFLt) Unemployed rate (UNEMPLOYED:) Real estates prices (REALES:)

Compliance with the law (LAW.)

Control of corruption (CORRUPTION?)

Figure 3 I Research analysis framework

Source: The author's synthesis

3.1.1 Dynamic model with the non-performing loans ratio (NPL) as dependent variable

NPLit = /?0 4- (ỊỵN PLi'1-ỵ 4- faLGi't 4- P^SlZEi t 4- [ỉ 4 R0Aịt 4- PsCARịt

4- p 6UNEMPL0YED t 4- f7 GDP L 4- Ps INFt + faREALESt 4- P^LAWt

4- /^CORRUPTION! 4- 8 iit

3.1.2 Dynamic model with the loan loss provision ratio (LLP) as dependent variable

LLPit = a0 4- ctỵLLPịt.ỵ 4- a 2 LGit 4- a 3 SIZE it 4- a4 R0A it 4- as CAR it

4- a6UNEMPLOYED t 4- a 7GDP t 4- a Q INFt 4- a 9 REALESt 4- a10LAWt + a^CORRUPTION t+ 3 iit

With:

Dependent variables

NPL it: The non-performing loans ratio of zlh bank in z'11 year

LLPit : The loan loss provision ratio of fb bank in zlbyear

- P q ,a Q : The slope coefficient of the model

- Independent variables

NPLịt_ỵ The non-performing loans ratio of /th bank last year of year t

LLP^^'. The loan loss provision ratio of /th bank last year of year t

LG it : Loan growth of zih bank in /“‘year

SIZE it: Bank size of fh bank in rlh year

R0Aiit : Return on assets of zlh bank in /,hyear

CARit : Capital adequacy ratio of z,h bank in rIhyear

UNEMPLOYED^ The unemployed rate in r,hyear

GDP t: GDP growth in r,hyear

INFt: Inflation rate in rthyear

REALES t : Real estates price index in ^year

LAW l : Compliance with the law index in ?hyear

CORRUPTION^. Control of corruption index in rthyear

E í)t, ỉi,t : Errors of the model

3.2 Variables

Based on previous studies, both in Vietnam and internationally, regarding the relationship between potential factors and the non-performing loan ratio of commercial

banks, the author group has chosen to use a range of variables and expected effect in the

research model as follows:

Table 3 1 Selected variables in the model

Source

Dependent variables

LLP U

The non­performing

in rth year

The loan loss provisionratio ofzlh bank in lhyear

Võ Thị Quý

& Bùi Ngọc Toan(2014), Nguyền Thị

Như Quỳnh etal.(2018),

Nguyễn Văn

Thuận &

Dương HồngNgọc (2015)

(2013),Nguyền Thị Như Quỳnh

et al (2018)

Financial statement

LLP it -i The loan loss

Author's calculati­on

LGi,t Loan growth

SIZEiit Bank size of

zth bank in /"’year

Vân Anh

(2013),

Đồ Quỳnh Anh &

Nguyền Đức

Hùng(2013),

Nguyền Văn

Thuận &

DươngHồng Ngọc(2015)

Net income Average total assets

(%)

Misman &

Ahmad

(2011),Nguyền

Tuấn Kiệt &

Đinh Hùng

Phú (2016)

Financial statement

CARi't Capital

adequacy ratio

of i,h bank in r'hyear

Thép &

Nguyền Thị Bích

Phượng(2016)

Nabila Zribi

& YounesBoujelbene(2011)

Income statemenet

UNEMPLOYED; The

unemployedrate in rlh year

The unemployed rate

Messai (2013), Filip

(2015)

Nguyền Thị Như Quỳnh

et al.(2018)

ADB

Source: Synthesized by the author from reference materials

et al (2018)

Nabila Zribi

& YounesBoujelbene(2011)

1FS -IMF

REALESt Real estates

GSO

with the law index in i"1 year

Compliance with the

law index

No existing studies

WB

CORRUPTION Control of

corruption index in lh

year

Control of corruption

index

No existing studies

WB

3.3 Data

The study uses data on factors affecting credit risk in 24 commercial banks including:

Joint Stock Commercial Bank for Investment and Development of Vietnam(BIDV), Joint Stock Commercial Bank for Foreign Trade of Vietnam (VCB), VietnamInternational Commercial Joint Stock Bank (VIB), Vietnam Prosperity Commercial Joint Stock Bank (VPB), Vietnam Joint Stock Commercial Bank for Industry and Trade (CTG), Asia Commercial Joint Stock Bank (ACB), Vietnam Export-Impoil Commercial

Joint Stock Bank (EIB), An Binh Commercial Joint Stock Bank (ABB), Orient

Commercial Joint Stock Bank (OCB), Petrolimex Petroleum Joint Stock Commercial Bank (PGB), Viet Capital Commercial Joint Stock Bank (BVB), Southeast Asia

Commercial Joint Stock Bank (SEABANK), National Commercial Joint Stock Bank

(NCB), Military Commercial Joint Stock Bank (MBB), Viet A Commercial Joint Stock

Bank (VIETABANK), Ho Chi Minh City Development Commercial Joint Stock Bank

(HDB), Vietnam Maritime Commercial Joint Stock Bank (MSB), Nam A Commercial Joint Stock Bank (NAB), Saigon Thuong Tin Commercial Joint Stock Bank (STB),

Saigon Joint Stock Commercial Bank for Industry and Trade (SGB), Saigon-Hanoi Commercial Joint Stock Bank (SHB), Vietnam Technological and Commercial Joint

Stock Bank (TCB), Lien Viet Post Joint Stock Commercial Bank (LPB), Tien Phong Commercial Joint Stock Bank (TPB) by collecting secondary data of independent anddependent variables from annual financial statements of banks and statistical agencies,

in the period of 2009 - 2021 for review and analysis The above 24 commercial banks

meet the criteria to exist and operate until the end of 2021

The basis to choose 24 commercial banks as the study sample: (i) with the

announcement of capital adequacy ratio; (ii) banks with charter capital of over 3000

billion VND; (iii) These 24 banks accounted for about 90% of charter capital and 80%

of the number of commercial banks at the time of the study Therefore, it can be

concluded that the selected study sample is representative of commercial banks

3.3.1 Sample

The data was collected based on 24 commercial banks in Vietnam, with continuous

statistics from 2009 to 2021 The objective is to avoid the impact of the global economic crisis in 2008 The total assets of these 24 commercial banks account for more than 90%

of the total assets of the entire commercial banking system in Vietnam Therefore, thissample is statistically significant and representative

3.3.2 Data Source and Extraction

The variables in the research are used in the form of secondary data Depending on

the characteristics and measurement of each variable, data may be available or need to

be processed and calculated The determination of variables in the model was done

manually by the authors

• Independent data variables: State Bank of Vietnam (SBV), General Statistics

Office of Vietnam (GSO), International Financial Statistics (IFS), The Asian Development Bank (ADB), International Monetary Fund (IMF), World Bank (WB)

• Dependent data variables: Financial statements, Income statements from 24 commercial banks

3.4 Hypotheses

3.4 L Credit risk of the previous year (NPLit_ lf LLPịỊ^)

In Vietnam, it is commonly observed that credit from banks often results in non­performing loans after one year This means that a decrease in credit quality in theprevious year leads to a higher non-performing loan ratio in the following year, and vice versa According to the study by Nguyễn Thị Như Quỳnh et al (2018), the credit risk ofthe previous year correlates positively with the non-performing loan ratio of the current

year This aligns with the findings of Salas and Saurina (2002), Đào Thị Thanh Bình &

Đồ Vân Anh (2013), Võ Thị Quý & Bùi Ngọc Toàn (2014), and Makri, Tsagkanos, and Bellas (2014) on the trend of the impact of credit risk from the previous year A high credit risk from the previous year is equivalent to the bank's inability to recover loansand poor credit management The increase in past non-performing loans requiresfinancial institutions to set aside credit risk reserves In the study by Suluck and Supat

(2012), it was indicated that the lag of the dependent variable correlates positively withthe dependent variable for most banks within the OECD bloc This suggests that credit risk reserves have a tendency to prolong, and a higher ratio of setting aside credit risk

reserves in the past (LLP t_f) indicates the increasing trend ofcredit risk in the followingyear, directly impacting the bank's business activities If risk is not addressed in a timely

manner, it may lead to prolonged credit shocks in the future

Hypothesis Hr. The past credit risk with a one-year lag is positively coiTelated withthe current credit risk of credit institutions

3.4.2 Loan Growth (LG)

The Loan Growth rate has both positive and negative impacts on the credit risk On one hand, a high Loan Growth rale indicates that the bank is operating well, funds arecirculating, businesses have easy access to loans, leading to high operational

profitability This increases the ability to repay loans, enhances liquidity, and reducesthe risk of non-performing loans for the bank (Sukrishnalall Pasha and Tarron Khemraj

(2009)) The reason for this inverse coiTelation during the research period in Vietnam

from 2007 to 2013 (Nguyen Tuấn Kiệt and Đinh Hùng Phú (2016)) is that credit from banks often results in non-performing loans after about a year This means that if loan

growth is low this year due to a high non-performing loan ratio from the previous year,the bank will have to focus on addressing non-performing loans while restricting loan

growth due to the imposition of the state bank

On the other hand, excessive credit expansion can make it more challenging to control

loan management activities, leading to higher credit risk Weinberg (1995) hypothesized

that during economic development, banks tend to loosen guarantee standards in

anticipation of high profits from lending However, lending activities need to adhere to

strict standards, so an increase in credit will lead to an increase in non-performing loans(Louzis el al., 2010) The impact of the Loan Growth rate on the non-performing loan

ratio has been proven in previous studies during credit booms, although there is still

mixed evidence in different research

Hypothesis Hi: Loan growth is positively correlated with credit risk in the banking system of Vietnam

3.4.3 Bank size (SIZE)

According to Đào Thị Thanh Binh & Đồ Vân Anh (2013), large banks often take

more risks in their lending activities because they believe there will be support from the

government The expansion of the bank's size is usually accompanied by an expansion

of credit operations, as it is their core business area With an increase in the number of

borrowing customers, there is also an increase in potential risks, leading to a higher non­performing loan ratio Conversely, according lo Nguyền Vãn Thuận & Dương Hong

Ngọc (2015), large banks often attract the attention of large investment organizations

Therefore, they tend to apply stable and long-term investment strategies, reflecting

confidence in government support when needed This results in an increase in the credit risk reserve ratio to protect the repayment ability of customers of large banks Theseresults are also consistent with the research of Dỗ Quỳnh Anh & Nguyền Đức Hùng

(2013) as well as Nguyền Văn Thép & Nguyền Thị Bích Phượng (2016)

However, some international studies have presented a different perspective

According to Salas & Saurina (2002), the scale of large banks is often associated withmore careful credit management Large banks tend to have better risk management capabilities through portfolio diversification and better management compared tosmaller banks Therefore, the credit risk level is often lower The reason for thisinconsistency may depend on the bank's management ability, capital management, assetstructure choices, as well as national management factors Based on the actual situation,the research group found that the model of Commercial Banks in Vietnam from 2009 to

2021 is consistent with the following hypothesis:

Hypothesis Hr Bank size is positively correlated with credit risk in the banking system of Vietnam

3.4.4 Return on Assets (ROA)

ROA is one of the important indicators to assess the profit-generating ability of a

bank It represents the net profit ratio that a bank obtains from each unit of investment

in its assets and business services ROA is calculated by dividing after-tax profit by the

total assets of the bank An increasing profit rate on total assets often indicates theefficiency of the business operations, effective use of assets, and good resourcemanagement According to Misman and Ahmad (2011), Mustafa et al (2012), Louzis

et al (2012), the provision for credit risk reserve and the profit rate on total assets of the

bank often have an inverse relationship An increase in the provision for credit risk

reserve, predicting to continue rising in the future, will reduce the bank's profit and increase the loan loss provision ratio, thereby decreasing the profit rate on total assets

According to Hu et al (2004), it is hypothesized that banks with high ROA are less

motivated to engage in risk activities Based on these theories, the author group presents

the following hypothesis

Hypothesis Hr Return on Assets is negatively correlated with credit risk in the

banking system of Viet Nam

3.4.5 Capital Adequacy Ratio (CAR)

CAR is a crucial indicator for assessing the risk level of banks for investors In

Vietnam, the minimum capital ratio is regulated at 9% according to Circular

36/2014/Directive - SBV According to Nabila Zribi and Younes Boujelbene (2011),

there is negative correlation between the capital adequacy ratio (CAR) and the credit risk level.This result indicates that banks with surplus capital tend to accept less risk

than those with insufficient capital

The study by Vồ Hông Đức et al (2014) suggests that the loan loss provision ratio(LLP) has a positive correlation with the capital adequacy ratio Similarly, Nguyền Văn

Thép & Nguyền Thị Bích Phượng (2016) also agreed that the CAR index has a positivelycorrelated relationship with the credit risk level When the capital adequacy ratio is high,

banks can confirm their risk compensation capability beyond the safety threshold,

leading to negligence in managing loans, thereby increasing RRTD Therefore, the

group proposes the following hypothesis:

Hypothesis H5: Capital Adequacy Ratio is positively correlated with credit risk in thebanking system of Vietnam

3.4.6 Unemployment ratio (UNEMPLOYED)

The unemployment rate plays a crucial role in determining the provision of personal consumer loans by banks Supporting consumer loans for unemployed individuals is one

of the credit policies encouraged by the Government of Vietnam In a recent study by Nguyền Thị Như Ọuỳnh et al (2018), an inverse relationship between the

unemployment rate and the non-performing loan (NPL) ratio was discovered This

finding is in stark contrast to previous studies by Messai and Jouini (2013), Klein (2013), Filip (2015), Ghosh (2015), Makri et al (2014) This difference can be explained by the

specific context of the study, where the ratio of personal consumer loans to total credit outstanding remains low (only 18%), and the unemployment rate has maintained a low level with little significant variation from 2009 to 2016 This is an aspect that needs to

be further examined and verified during the period from 2009 to 2021 Therefore, the

group proposes the following hypothesis:

Hypothesis H ô : Unemployment rate is negatively correlated with credit risk in the

banking system of Vietnam

3.4.7 Gross Domestic Product growth (GDP)

GDP, a crucial macroeconomic indicator, is often associated with economic

development and improvement in the living standards of the people in a country A rapid increase in GDP is usually a positive sign, indicating strong national development, growing and stable businesses, thereby enhancing profit-making capabilities This can

contribute to strengthening the debt repayment capacity for Commercial Banks, creating favorable conditions for debt repayment and reducing the NPL ratio Conversely, a

decline in GDP often leads to a downturn in business activities, potentially causing

difficulties in debt repayment and increasing the NPL ratio Based on theories studied

by Salas & Saurina (2002), Rajan & Dhal (2003), Mcssai (2013), Đồ Quỳnh Anh &Nguyền Đức Hùng (2013), the group proposes

Hypothesis H7: GDP is negatively correlated with credit risk in the banking system

of Vietnam

3.4.8 Inflation (INFL)

Inflation is the continuous increase in the general price level, with both positive and

negative impacts on the credit risk Nabila Zribi and Younes Boujelbene (2011) found

that the credit risk in Tunisia banking system depends on inflation and asserted that

inflation has an inverse effect on credit risk However, the theory supporting this

hypothesis is not clear, as the authors combine independent variables into a group of

macroeconomic factors and only examine the most pronounced effects, neglecting a detailed explanation of the Inflation variable

On the other hand, the inflation rate is positively correlated with the non-performing

loan (NPL) ratio (Nguyền Thị Như Quỳnh et al (2018), Gunsel (2011)) The authors

theorize that a high inflation rate forces the central bank to implement tight monetary

policies to combat inflation, leading to an increase in lending interest rates, therebyaffecting the input costs of businesses and reducing their operational efficiency The rise

in the inflation rate can impact the ability of businesses to repay loans to banks as high borrowing costs push the burden of non-performing loans onto the banks Wheninflation increases, market demand and purchasing power decrease, causing inventorybuildup, reducing business profitability, and even leading to businesses operating at a