This literature contends that GDP growth rate, exchange rate, inflation, unemployment, bank size, bank diversification and bank performance are the primary factors of credit risk in Viet
Literature review
Research on commercial bank’s credit risk
In the 2022 study "Completing the Legal Framework on Credit Risk Management for Off-Balance Sheet Activities at Vietnamese Commercial Banks," Pham Anh examines the Bank for International Settlements (BIS) regulations concerning risk management in commercial banks The research highlights the current state of these regulations as applied by the State Bank and assesses how Vietnamese commercial banks manage off-balance sheet credit risks Additionally, the author proposes enhancements to the legal framework governing credit risk management for these activities, aiming to strengthen compliance and oversight within the sector.
In her 2023 study, Nhung Nguyen examines the effectiveness of credit risk assessment tools in Vietnamese commercial banks, utilizing primary data and descriptive statistical methods alongside comparative analysis The findings reveal that many individuals do not fully leverage the 5C model's components for credit analysis, leading to insufficient input in the analytical framework and diminished reliability of credit risk assessments for borrowers To address this issue, the author suggests several solutions aimed at enhancing the credit assessment process for commercial banks.
Research on factors affecting commercial bank’s credit risk
Recent research by Amit Ghosh (2015) highlights the factors influencing credit risk in U.S commercial banks, focusing on the interplay between bank characteristics, regional economic indicators, and macroeconomic factors Utilizing Fixed Effects Models (FEM) and Dynamic Generalized Method of Moments (GMM), the study addresses issues of autocorrelation and endogeneity, ensuring robust results The findings reveal that increased credit growth and loan loss provisions elevate non-performing loans (NPL), while higher bank profitability, indicated by Return on Assets (ROA), has a negative impact on credit risk Contrary to expectations, greater diversification through non-interest income introduces additional risks for banks Moreover, the systems-GMM estimation offers more reliable evidence than FEM results, particularly emphasizing the significant influence of equity-to-total assets on credit risk deterioration in banks.
The "too big to fail" hypothesis suggests that the size of the banking industry positively correlates with key economic variables Both fixed effects and GMM estimations reveal that increases in GDP and personal income growth rates reduce banks' credit risk Additionally, the unemployment rate negatively impacts non-performing loans (NPL) However, inflation exhibits a contrasting relationship, showing a positive significance in GMM analysis To enhance the analysis, the author includes comparative results for savings institutions derived from both fixed effects and GMM estimations.
A study by Khan (2017) explored the relationship between bank-specific factors such as income diversification, profitability, capitalization, and operating efficiency, and the loan loss rate of commercial banks in Pakistan Analyzing data from 2005 to 2017, the research found that an increase in interest income or bank profits negatively affects the capital loss rate Notably, the study highlighted that higher bank capitalization correlates with a reduced likelihood of risk.
In their study, Ahmed et al (2021) employed GMM models to analyze the influence of bank-specific and macroeconomic factors on non-performing loans (NPL) from 2008 to 2018 The findings indicated that credit growth, loan loss provisions, and bank diversification significantly increased NPL, while higher return on assets (ROA) correlated with reduced loan losses, aligning with Amit Ghosh's research (2015) Conversely, the study revealed a conflicting outcome regarding bank size, as larger banks exhibited a lower risk ratio.
Research by Naili et al (2022) examined 53 commercial banks across five emerging markets in the Middle East and North Africa from 2000 to 2019, utilizing Fixed Effect, Random Effect, and GMM models The findings indicate that as bank size increases, credit risk significantly decreases Additionally, variables such as profitability, economic growth rate, and unemployment rate showed consistent results with previous studies.
In their 2016 study, Anastasiou Dimitrios, Louri Helen, and Tsionas Mike utilized the System GMM model to analyze data from 15 Eurozone countries spanning from 1990Q1 to 2015Q2 They developed three static and three dynamic models, examining macroeconomic variables, bank-specific variables, and a combination of both The findings indicated that rising unemployment negatively impacts credit quality, while improvements in GDP growth rates and return on assets (ROA) help mitigate credit risk, as measured by non-performing loans (NPL) Additionally, the research highlighted that the NPL ratio from the previous year significantly influences the current year's NPL ratio.
A study by Lobna Abid et al (2013) utilized GMM models to examine the impact of inflation on non-performing loans (NPLs) in consumer loan portfolios, revealing that inflation is a significant indicator of NPLs This finding aligns with Amit Ghosh's (2015) GMM model results, indicating that a decrease in the inflation rate positively influences household financial conditions in Tunisia, ultimately enhancing loan repayment rates Thus, there exists a positive relationship between inflation and NPLs.
A study by Quy, Vo et al (2014) analyzed 26 commercial banks from 2009 to 2012, employing GMM methods due to violations of OLS linear regression assumptions The findings highlighted three key variables impacting credit risk: past bank credit risk with a one-year lag, credit growth rate with a one-year lag, and GDP growth rate with a one-year lag The results indicated that a declining GDP growth rate, coupled with previous low-quality loans, has heightened the credit risk for commercial banks The authors concluded that increasing high-quality credit could foster GDP growth and reduce future credit risks for these banks.
Duong et al (2016) conducted a study titled "The Analysis of Major Credit Risk Factors - The Case of Vietnam Commercial Banks," utilizing an estimated GMM model Their findings revealed a statistically significant positive relationship between GDP growth and credit risk, which contrasts with historical research that typically identified a negative correlation between these two factors The authors suggest that this unexpected result aligns with Vietnam's economic context, where the economy heavily relies on bank loan capital and experiences substantial credit growth.
“boiled” by soaring economy The ICOR report of the General Statistics Office (Note
7) cited by the author determined that Vietnam continues to be in a higher position compared to other countries in the region, which demonstrates the poor efficiency of investment capital The model results also find that the NPL ratio, which has a lag of
1 year after GDP, then has a positive interlinkage with GDP, due to the deterioration of credit quality during the period of economic growth
A study by Hang, Hoang et al (2018) analyzing 20 commercial banks in Vietnam from 2006 to 2017 reveals that an increase in Loan Loss Provision heightens credit risk at banks Additionally, the research indicates that the unemployment rate negatively impacts this dependency variable, supporting findings from Amit Ghosh's (2015) study.
A comprehensive study by Tam, Le et al (2021) analyzed data from 35 commercial banks between 2012 and 2022 using Pooled OLS, FEM, and REM models, revealing that macroeconomic factors like real estate market growth, real interest rates, and exchange rate fluctuations significantly influence credit risk measurement Conversely, an increase in return on assets (ROA) negatively affects banks' capacity to manage credit risk Notably, bank-specific factors such as asset size, financial leverage ratio, credit growth rate, outstanding loan/mobilized capital ratio, and market capitalization showed no significant impact This indicates that Vietnam's commercial banking system exhibits similar characteristics across banks with comparable financial leverage, while the State Bank governs the maximum credit growth rate Additionally, other macroeconomic variables, including economic growth and inflation rate, do not affect the credit risk of Vietnamese commercial banks.
Research gap
Credit risk in commercial banks is a significant concern acknowledged by researchers globally, with studies highlighting the influence of both bank-specific and macroeconomic factors on credit quality Traditionally, Non-Performing Loans (NPL) have been the primary focus for assessing credit risk; however, Loan Loss Coverage (LLC) is equally vital for ensuring bank stability and soundness LLC serves as a critical indicator of the quality of bank accruals from an accounting perspective This article explores the factors affecting credit risk in Vietnamese commercial banks by examining both NPL and LLC as dependent variables, offering a fresh viewpoint on the determinants of credit risk in this context.
Objectives of the thesis
● General objectives: To exploit the factors affecting credit risks at commercial banks in Vietnam, thereby proposing appropriate recommendations and policies to minimise credit risks
- First, synthesise the theoretical basis of factors impacting credit risks of commercial banks
- Second, study the practical impact of macro and bank specific factors on 24 commercial banks
- Third, draw conclusions and make practical recommendations applicable to commercial banks to minimise credit risks.
Research object and scope
- Research object: The impact of macroeconomic and bank specific factors on credit risks of commercial bank
- Research scope: The research was conducted on 24 commercial banks in Vietnam
Research methodology
In order to assess the influence of factors on credit risks of commercial banks in Vietnam, the author uses a combination of the following methods:
The research data is derived from the consolidated audited financial statements of Vietnamese commercial banks at year-end, encompassing a panel dataset of 24 banks, as detailed in Table 1 (Appendix) This dataset includes both cross-sectional observations at a single point in time and time-series data for individual banks from 2015 to 2022 Additionally, macroeconomic information is sourced from the General Statistics Office of Vietnam's website.
The research employs quantitative methodology using regression methods for panel data, focusing on traditional techniques such as Fixed Effects Model (FEM), Random Effects Model (REM), and Feasible Generalized Least Squares (FGLS) However, the presence of lagged variables can introduce endogeneity, leading to issues such as autocorrelation and variance instability in FEM and REM To mitigate these challenges, Arellano and Bond (1991) introduced Generalized Method of Moments (GMM) regression Consequently, this study primarily utilizes the System GMM estimation method (SGMM) to ensure robust and effective estimates.
Thesis structure
Thesis contribution
This thesis aims to identify and assess the factors influencing banks' credit risk, specifically focusing on macro-level determinants and bank-specific indicators that affect the credit quality of commercial banks in Vietnam To mitigate credit deterioration and enhance the economic stability of the country, the study provides management recommendations for banks and suggests regulatory improvements for the government It addresses gaps in existing research by exploring a broader range of variables associated with credit risk, aiming to offer a comprehensive analysis that builds on previous studies.
OVERVIEW
Definition of credit risk
Credit is the primary revenue source for commercial banks, making it the most significant type of risk they face Despite advancements in risk measurement, forecasting tools, and portfolio diversification, managing credit risk continues to be a challenging endeavor.
Credit risk, as defined by Bernd Schmid, comprises two key elements: default risk and spread risk Default risk refers to the possibility that a borrower may fail to make timely interest or principal payments, while spread risk involves the potential decline in market value due to shifts in the debtor's credit quality.
Pricing Models: Theory and Practice, 2004)
Credit risk occurs when financial intermediaries are unable to collect both the principal and interest on loans, or fail to receive timely payments of these amounts (Ngoc, To 2019; Money and Banking, 2019).
Credit risk refers to the possibility that a borrower or counterparty may not fulfill their obligations as per the agreed terms, as defined by the Basel Committee on Banking Supervision in their 1999 principles for managing credit risk.
In the 2021 Circular 11/2021/TT-NHNN issued by the State Bank of Vietnam, credit risk in banking refers to the potential losses faced by credit institutions or foreign bank branches when customers are unable to repay either part or all of their debts as stipulated in contracts or agreements.
Credit risk refers to the potential financial loss a commercial bank faces when a customer fails to meet their obligations This risk is a constant presence in the banking industry and is an inherent aspect of financial operations Essentially, banks operate in a sector where profitability is closely linked to managing the risks associated with lending and credit.
Indicators to evaluate credit risk of commercial banks
Non-performing loans (NPLs) lack a universally accepted definition across global economies According to the IMF's 2004 Compilation Guide on Financial Soundness Indicators, a loan is classified as non-performing if interest and/or principal payments are overdue by 90 days or more, if interest payments have been capitalized, refinanced, or delayed for 90 days or longer, or if payments are less than 90 days overdue but there are credible concerns, such as bankruptcy filings, that full repayment may not occur.
The Basel Committee on Banking Supervision (BCBS) prefers the term "impaired loans" over "non-performing loans," indicating that impairment occurs when it is likely that full repayment will not be received The BCBS does not focus on customer delinquency, as different countries have varying definitions of bad debt, often classifying debts overdue by 31 or 61 days Consequently, while 90 days is frequently used as a benchmark for categorizing debts as bad, this criterion is not universally applied.
In Vietnam, non-performing loans (NPLs) are defined under Circular 11/2021/TT-NHNN as loans categorized in groups 3, 4, and 5 According to Aril, A (2019), the Financial Stability Institute (FSI) recommends classifying loans as non-performing if (1) principal and interest payments are overdue by 90 days or more, (2) interest payments have been capitalized, refinanced, or rolled over for 90 days or more, or (3) there is evidence suggesting reclassification as non-performing, such as a debtor filing for bankruptcy, regardless of the 90-day payment status.
1.2.2 Loan Loss Coverage Ratio (LLC)
The loan loss coverage ratio serves as a crucial metric for a bank's protection against the risks associated with bad debts According to Hieu and Nguyen (2020), this ratio can be defined in three distinct ways One of the primary definitions equates the loan loss coverage ratio with the loan loss provision ratio, which is calculated using a specific formula.
Provision for bad loans/Total liabilities This index shows the percentage of all existing loans that can be covered by the bank's bad loan provision
The loan loss coverage ratio is determined using the formula: Provision for bad debts divided by Total bad debt balance This ratio indicates the extent to which a bank's bad debt provision can cover all existing bad loans, including those categorized as group loans.
According to Circular 11/2021/TT-NHNN issued by the State Bank on July 30, 2021, financial institutions are required to maintain a general reserve amount equal to 0.75% of the total debt balance from group 1 to group 4 loans Additionally, specific provision ratios must be applied for each debt group to ensure adequate risk management.
Finally, the loan loss coverage ratio can also be calculated through the formula
Profit after tax/Total bad debt balance With this formula, the ability to use net profit to cover total bad loans is clearly shown
According to Hieu Nguyen (2020), a higher non-performing loan (NPL) coverage ratio indicates that a bank is better equipped to manage hard-to-recover debts through provisions, serving as a protective buffer against future financial shocks This means that banks with a robust loan loss coverage ratio are more secure, even if all bad debts become uncollectible Conversely, commercial banks with a low loan loss coverage ratio face significant risks, as insufficient coverage can lead to severe losses from credit risk exposure.
Factors affect credit risk in commercial bank
Empirical evidence strongly connects the macroeconomic environment to a bank's credit quality This paper identifies economic growth, exchange rates, unemployment rates, and inflation rates as key macroeconomic factors influencing credit risk.
Economic growth denotes growth in the production of goods and services over a period of time It is commonly assessed by using indicators such as GDP or GNP
Economic growth is a crucial indicator of a country's status and potential, as it leads to increased employment rates, investment opportunities, and improved living standards In a thriving economy, business activities and trade flourish, driving up capital demand and generating surplus capital This environment enhances the financial stability of households and businesses, improving their repayment capabilities Consequently, in developed economies, the likelihood of credit risk is significantly lower compared to periods of economic recession.
Prior research highlights the strong connection between economic growth and the credit quality of banks, particularly in Vietnam, where previous GDP significantly influences commercial banks (Duong et al., 2016) This relationship is attributed to the unique characteristics of Vietnam's economy and banking sector The rising credit growth rate in Vietnam is linked to its booming economy, indicating a heavy reliance on bank loans Additionally, the high non-performing loan (NPL) rate in Vietnam may stem from the misallocation of capital into high-risk funds, which increases the likelihood of credit risk.
Some scholars argue that credit risk can be reduced through economic growth, as favorable conditions facilitate the exchange of goods and services, allowing households and businesses to enhance their income and improve their debt repayment capabilities (Naili, 2022; Abid, 2014; Ahmed, 2021) Conversely, during economic downturns, borrowers may face difficulties in generating sufficient cash flow to meet their debt obligations, leading to a decline in banks' credit quality.
According to Krugman, the exchange rate is the price of one currency in terms of another When the exchange rate rises, the value of the local currency falls, making
The rising costs of 13 imported goods are placing a significant burden on businesses reliant on imported raw materials and intermediate products This price increase not only elevates production costs but also contributes to a higher consumer price index, resulting in inflation and increased borrowing expenses Consequently, the likelihood of credit risk is on the rise.
Research indicates that exchange rate fluctuations significantly impact banks' risk-taking behavior According to Tam (2020), these fluctuations positively influence credit risk, as a depreciating local currency increases input costs for businesses Consequently, higher expenses for importing materials lead to reduced profits, ultimately impairing the ability of enterprises to repay their debts to banks.
Chaibi (2015) offers a unique perspective on the relationship between exchange rates and credit risk, noting that an increase in exchange rates correlates with a rise in non-performing loans (NPL) in France The author explains that a real appreciation of the local currency makes domestic products more expensive, diminishing the competitiveness of export-oriented companies and adversely impacting their ability to service debts Conversely, in Germany, the relationship is negative; an appreciation of the exchange rate enhances the capacity of borrowers with foreign currency loans to meet their debt obligations, resulting in a decrease in NPL.
The unemployment rate significantly influences banks' credit risk, as it reflects the number of individuals seeking work but unable to find employment (Ngoc, Nguyen, 2006) A rising unemployment rate results in reduced budget capital due to decreased tax revenues and increased support for unemployed workers, ultimately leading to economic recession and diminished investment capital This situation adversely impacts the credit market, as job loss translates to decreased income and heightened difficulty in meeting financial obligations, thereby increasing the likelihood of defaults on loans at commercial banks.
Most of the scholars conducted on this relationship provide positive direction (Chaibi, 2015; Ghosh, 2015; Dimitrios, 2016; Louzis, 2012; Naili, 2022)
According to Louzis (2012), the unemployment rate is a crucial indicator of consumer non-performing loans (NPLs), as rising unemployment diminishes households' ability to manage their debts Supporting this, Naili (2022) highlights that a bleak labor market, characterized by a scarcity of job opportunities, leads borrowers to be less inclined to repay their debts, thereby increasing NPL levels in commercial banks Furthermore, individuals with low incomes face elevated interest rates due to their uncertain employment status, which further hampers their ability to service loans.
Inflation, as defined by the International Monetary Fund, refers to the rate at which prices increase over a specific time frame, often assessed through the overall rise in prices or the cost of living within a country This economic indicator illustrates how much more expensive a particular set of goods and services has become, typically measured annually Additionally, inflation significantly influences credit risk for commercial banks, affecting their lending practices and financial stability.
High inflation leads to rapidly rising prices, causing speculation and hoarding that disrupts the circulation of goods and complicates business operations As purchasing power diminishes, consumers find it increasingly difficult to afford products, which forces many businesses to suspend production or limit their market reach Additionally, high inflation drives up interest rates, restricting access to credit for companies trying to sustain operations, ultimately resulting in a decline in credit quality and solvency for these enterprises.
Numerous studies in banking literature explore the connection between inflation rates and credit risk According to Lobna Abid et al (2014), the inflation coefficient serves as a crucial indicator of non-performing loans (NPL) in consumer loan portfolios This relationship is attributed to the positive effects of declining inflation on household financial conditions, which enhances loan repayment capabilities Ghosh (2015) further elucidates this connection, noting that rising inflation without a corresponding increase in nominal income leads to a decrease in real income.
Naili (2022) confirmed that rising inflation leads to an increase in non-performing loans (NPLs), particularly affecting floating rate loans The study concluded that high inflation adversely impacts household income, diminishing the real value of earnings and consequently reducing the ability to repay debts.
A study by Chaibi (2015) demonstrates that inflation does not influence non-performing loans (NPL) in France The research indicates that while inflation can negatively affect the real value of outstanding loans, this effect is offset by other factors, ultimately neutralizing the overall impact of inflation risk on NPL.
The factors of credit risk should not be sought exclusively among macroeconomic variables, which are exogenous to the banking industry (Louzis,
The unique features of commercial banks, coupled with their initiatives to enhance performance and manage risk, are anticipated to significantly influence credit risk This study focuses on key bank characteristics—namely bank size, profitability, capitalization, and diversification—to explore their relationship with credit risk.
DATABASE AND METHODOLOGY
Sample and data sources
This research analyzes a sample of 24 commercial banks in Vietnam, focusing on various performance metrics such as the Return on Assets (ROA) ratio, total assets for scale, and income diversification based on the ratio of interest to non-interest income The study is limited to listed banks due to data availability, as many commercial banks do not publicly disclose their annual financial statements Additionally, the research excludes foreign bank branches and joint-venture banks, which have limited operations in the Vietnamese market The objective is to examine factors influencing credit risk within commercial banks, with the research period spanning from 2015 to 2022, coinciding with the initial implementation of bad loan settlement measures outlined in Resolution 42/2017/QH14 by the State.
This research analyzes bank-specific data from the audited consolidated financial statements of 24 Vietnamese commercial banks from 2015 to 2022, sourced from the Vietstock Financial - Securities portal Additionally, the author calculates unavailable data, such as market capitalization and income diversity, using appropriate formulas to ensure comprehensive analysis.
Macroeconomic data on GDP per capita was sourced from The World Bank, while the unemployment rate, inflation rate, and exchange rate were primarily obtained from national official sources, including the General Statistics Office of Vietnam and the State Bank of Vietnam Most bank-specific and macroeconomic variables are reported on an annual basis.
Variable Mean Min Max Std Dev Skewness Kurtosis
Figure 1: Descriptive statistic for the sample period 2015 - 2022
Figure 1 shows the descriptive statistics for all banks for the period covering
From 2015 to 2022, the analysis reveals that the average values of Log(NPL) and Log(LLC) are approximately 0.19 and 1.89, translating to mean NPL and LLC ratios of 0.0155 and 77.62%, respectively Additionally, the standard deviations for these variables are nearly 0.21 and 0.25.
22 respectively These indexes demonstrate that the credit risk of commercial banks in Vietnam is at a low level and the distinction among them is relatively huge
During challenging times, banks achieved a notable profitability rate of 8.13%, with an average size of VND 181,970,086 million However, they exhibited relatively low diversification, deriving 80% of their income from interest-related activities This level of diversification is significantly higher compared to global standards; for instance, in the United States, the proportion of interest income to total income is below 50% and is on a declining trend, while European commercial banks report an approximate ratio of 55.55%.
Regarding the macroeconomic indicators used in this thesis, the statistics indicate an average GDP per capita of VND 5.24 million, an exchange rate of VND
The unemployment rate reached a peak of 3.24% and a low of 2.40%, currently standing at 2.40%, while the inflation rate has remained stable, fluctuating between 1.4% and 3.24%, with a current rate near 2.14% The exchange rate has shown some variability but remains close to the average throughout the review period.
The author conducts a statistical analysis of the data set by examining key metrics such as Kurtosis, Skewness, maximum and minimum values, fluctuations, and standard deviations Notably, GDP Growth, Exchange Rate, Bank Performance, and Bank Diversification exhibit negative Skewness, indicating a leftward deviation from the mean In contrast, variables like inflation, the unemployment rate, bank capitalization, and bank size also show negative values but tend to deviate to the right of the mean Furthermore, when all variables are positive, the Kurtosis coefficient reveals a peak distribution that aligns more closely with the mean than with a normal distribution.
Methodology
Credit risk accumulations often develop over extended periods, suggesting that the impact of credit losses on the macroeconomic cycle may not be immediate To tackle the issue of persistent credit risk, I propose a dynamic model that incorporates a lagged value of credit risk.
23 bank-characteristic variables that are most likely to be endogenous with credit risk
A deterioration in the balance sheets and asset quality of banks is reflected in rising credit risk, which may diminish leverage and lower earnings for the banks
Using a fixed effect model with a lagged dependent variable is inappropriate due to the correlation between the error term and the lagged credit risk component, leading to inconsistent estimates To resolve this issue and address the endogeneity problem, I utilize the systems-GMM estimation method developed by Arellano and Bover (1995) and Blundell and Bond (1998).
(1) NPL it = 𝛼 i,t + β 1 NPL i.t-1 + β 2 Bank-specific i,t + β 3 Macro-level it + 𝜺 it
(2) LLC it = 𝛼 i,t + β 1 LLC i.t-1 + β 2 Bank-specific i,t + β 3 Macro-level it + 𝜺 it
This model treats banking industry-specific variables as endogenous, utilizing GMM-style instrumentation similar to the lagged dependent variable, while the macroeconomic factors influencing credit risk are predetermined and instrumented using an IV approach.
Hypothesis development
This literature review highlights the relationship between credit risk and various determinants at both the bank-specific and macroeconomic levels This section of the thesis will explore the factors influencing banks' credit risk, as identified in previous studies, and will be incorporated into the model developed in this paper.
Vietnam's GDP per capita has experienced remarkable growth, rising from over $2,500 in 2015 to around $4,200 in 2022, according to Forbes (VTVNews, 2022) This significant increase highlights the vitality of Vietnam's economy and establishes a solid foundation for the banking sector and robust credit expansion Based on this observation, the thesis proposes the following hypotheses.
H1: Economic growth has a negative impact on credit risk
The Federal Reserve's operational policies are anticipated to significantly impact the currencies of emerging nations, especially Vietnam, which has a large, open economy This influence is reflected in various sectors, including the stock market, real estate, corporate bonds, and foreign currency reserves While the appreciation of the Vietnamese Dong (VND) benefits the import market, it negatively affects exports Overall, these dynamics shape the exchange rate trends in Vietnam.
From 2015 to 2022, the value of the VND experienced a decline due to inflation This situation raises a critical question regarding the impact of inflation on credit risk, sparking an ongoing debate that necessitates further exploration.
H2: Exchange rate has a positive impact on credit risk
The unemployment rate serves as a crucial indicator of economic health in Vietnam, increasing during economic downturns and decreasing as the economy recovers and grows This fluctuation directly influences consumers' ability to repay loans from commercial banks Based on these observations, the following hypothesis is formulated.
H3: Unemployment rate has a positive impact on credit risk
In recent years, Vietnam has successfully maintained moderate and steady inflation; however, the post-Covid-19 era has seen major countries grappling with unexpectedly high inflation, which has increased pressure on Vietnam's economic system This high inflation directly affects businesses that rely on imported food and energy, diminishing their ability to repay loans Consequently, this paper hypothesizes that
H4: Inflation has a positive impact on credit risk
In Vietnam, large banks often prioritize lending to major firms and state-owned enterprises due to their advantages in the borrowing process, which can lead to a streamlined loan approval system However, this focus increases the risk of loan defaults, prompting banks to enhance their loan loss provisions and coverage to safeguard their financial stability To validate these observations, this study proposes specific hypotheses for empirical testing.
H5: Bank size has a negative/positive effect on credit risk
The banking sector is crucial to Vietnam's economy and has been enhancing its operational efficiency, resulting in a steady increase in industry earnings However, the COVID-19 pandemic caused stagnation and even declines in earnings from 2019 to 2021 With the pandemic's impact diminishing in 2022, many banks began to recover their profitability and reduce the pressure for increased loan provisioning Notably, over half of the banks lowered their credit risk provisioning costs compared to the previous year This indicates that in Vietnam's financial landscape, improved profitability is associated with reduced credit risk in banks This study aims to explore the relationship between bank profitability and credit risk.
H6 : Bank profitability has a negative effect on credit risk
In recent years, Vietnamese banks have sought to increase their charter capital to improve operational safety and support commercial expansion This initiative provides banks with additional resources for growth while enhancing their ability to maintain safety indicators The competition for capital among banks, particularly the "Big 4" state-owned institutions, has become increasingly intense.
26 commercial banks, but even more stimulating among joint-stock commercial banks
Several banks are planning to increase their charter capital by thousands of billions of VND to support credit and liquid asset investments, as well as to enhance infrastructure, technology, and branch networks This significant capital increase acts as a "buffer," providing banks with the resources needed to tackle challenges and expand operations, ultimately benefiting both the economy and enterprises Given this context, it is valuable to further investigate this relationship through the following hypothesis.
H7: Bank capitalisation has a negative impact on credit risk
According to the project on restructuring the system of credit institutions according to Decision No 254/QD-TTg in 2012 and Decision No 986/QD-TTg in
In 2018, the banking industry's development strategy emphasized income diversification, leading to a gradual shift in banks' business models to reduce reliance on credit activities while enhancing revenues from non-credit services The COVID-19 pandemic introduced new challenges, prompting banks to implement measures such as increasing risk provisioning and restructuring income sources Consequently, the growth of non-interest income has positively impacted the bank credit system.
This study employs the ratio of interest income to total income as a proxy for bank diversification, challenging previous literature The author anticipates a positive relationship between credit risk and bank diversification, suggesting that increased diversification may lead to enhanced credit risk management in banking institutions.
H8: Bank diversification has a positive impact on credit risk
Research model
This paper categorizes the variables into two primary groups: macroeconomic determinants and bank-specific factors To facilitate the exploration of regression models, all variables are presented in their logarithmic forms.
27 function This model will mitigate the high volatility of the data, and more accurately represent economic variables when expressing significance as a percentage
This paper explores the representation of banks' credit risk through the non-performing loan (NPL) ratio and the loan loss coverage (LLC) ratio Extensive literature has examined credit risk in commercial banks, highlighting that it can be assessed using both precise and approximate metrics Key indicators for measuring credit risk include non-performing loans, loan loss provisions, and loan loss coverage, with the NPL ratio being particularly significant for evaluating loan portfolio quality (Mahyoub, M 2021) Therefore, this research emphasizes the importance of incorporating the NPL ratio alongside an additional indicator to provide a comprehensive view of banks' credit risk Additionally, loan loss coverage plays a vital role in assessing capital adequacy and prudent risk management, serving as a critical market signal While the State Bank of Vietnam (SBV) has established regulations for specific and general provisions, the actual risk provision amount is influenced by individual bank discretion According to Circular 11/2021/TT-NHNN, the specific reserve amount deducted from customers is determined using a specified formula based on the principal balance of debts.
C i : Deductible value of collaterals, financial leasing assets, negotiable instruments, other valuable papers in discounting activities, resale and purchase of Government bonds of the i-th debt
Commercial banks can adjust their specific deduction rates to influence the deduction level and the size of their risk provisions Consequently, the NPL coverage ratio serves as an effective measure for assessing a bank's resilience in situations where non-performing loans (NPLs) cannot be recovered This is particularly relevant in the context of the Vietnamese banking system.
While a significant level of government engagement is present, it may still present certain challenges The author addresses these concerns by incorporating NPL and LLC into the study, ensuring the findings of the model remain consistent and comprehensive.
This paper analyzes the ratio of Non-Performing Loans (NPLs) derived from the financial statements of commercial banks, focusing on the criteria of Total Loans categorized from groups 3 to 5 in relation to Gross Loans for the year 2015.
In 2022, commercial banks classified their loans according to Circular No 02/2013/TT-NHNN issued by the State Bank of Vietnam (SBV) Additionally, the non-performing loans coverage ratio is determined by the proportion of provisions for bad loans to the total balance of bad debts.
Economic growth is essential for understanding the factors influencing credit losses Historically, the GDP growth rate has been the primary metric for assessing economic progress However, a more comprehensive measure that accounts for population growth is GDP per capita, which is calculated by dividing the GDP by the mid-year population This ratio serves as a valuable proxy for evaluating economic growth.
This article examines the impact of exchange rates on credit risk, focusing specifically on the USD/VND exchange rate The USD is the primary currency for international raw material transactions and serves as a safe haven for investors during economic uncertainty Fluctuations in the dollar's value significantly affect economies worldwide Therefore, to analyze the relationship between exchange rates and credit loss, the USD/VND exchange rate is utilized as a key variable, aligning with the research conducted by Chaibi (2015).
The rising unemployment rate is theoretically diminishing the general population's ability to repay debt, subsequently affecting the credit quality of banks This assertion is supported by various studies, including those conducted by Chaibi (2015), Ghosh (2015), Dimitrio (2016), and Louzis.
In a study by Naili M et al (2022), the unemployment rate in Vietnam was analyzed using the formula: Unemployment rate (%) = (Number of unemployed people / Workforce) x 100 This calculation helps to quantify the relationship between unemployment and economic factors, with the unemployment rate expressed in percentage terms.
Inflation is a crucial indicator of economic health, making it vital to consider its impact on the credit systems of banks In Vietnam, the common method for calculating inflation involves analyzing price changes of goods over time The inflation rate is determined using the formula: inflation rate = (P1 - P0) / P0 x 100%.
P0 is the price of basket of goods in year 0
P1 is the price of basket of goods in year 1
This study utilizes the inflation rate data compiled by the General Statistics Office as a key representative variable This approach aligns with the methodologies established by Naili M et al (2022), Abid (2014), Chaibi (2015), and Ghosh (2015).
● Bank specific variables a) Bank Size
Research indicates that bank size is a crucial factor when analyzing credit risk at the bank level Traditionally, the natural logarithm of total assets, derived from financial statements, has been used to measure bank size However, previous studies have shown inconsistent results regarding the correlation between bank size and credit quality To address this gap, the current study utilizes total assets to validate the relationship between bank size and credit effectiveness, aligning with the methodologies of prior researchers such as Louzis et al (2012), Ghosh (2015), and Dimitrios (2016).
Previous literature indicates that key indicators of bank performance include return on assets (ROA), return on equity (ROE), return on deposits, and net interest margin (NIM), with ROA being the most frequently utilized measure Consequently, it is essential to integrate ROA into models assessing the impact on banks' credit ROA is calculated as the ratio of net income to total assets (Louzis et al., 2012; Ghosh, 2015).
Dimitrios, 2016; Naili M et al, 2022) c) Bank Capitalization
Bank capitalization can be assessed through various methods, with the most common being the evaluation of equity, which acts as a risk buffer for commercial banks Higher equity levels enhance a bank's resilience, especially during challenging business environments A robust equity capital base allows banks to diversify their loan portfolios, strengthen their market presence, and improve their ability to raise funds and extend credit This article utilizes the ratio of total equity to total assets as a proxy for measuring a bank's capitalization (Louzis et al., 2012; Ghosh).
EMPIRICAL RESULTS AND DISCUSSION
Credit risk in Vietnam commercial banks 2015 - 2022
● The period before the COVID-19 pandemic
Since late 2007, Vietnam's non-performing loan (NPL) ratio has surged, particularly worsening after 2011 In May 2015, the State Bank of Vietnam (SBV) reported a staggering bad debt ratio of 17.21%, amounting to VND 465 trillion in non-recoverable loans By the end of 2016, this figure escalated to nearly VND 600 trillion, representing 13.3% of GDP To combat the rising bad loans threatening the stability of the banking system and economic activities, the government approved the establishment of the Vietnam Asset Management Company (VAMC) The SBV also implemented measures to supervise commercial banks and guide them away from high-risk lending, particularly in real estate As a result, by February 2017, the banking sector had resolved VND 611.59 trillion in non-performing loans, significantly enhancing credit quality, with VAMC handling 43% of the bad debts.
The State Bank of Vietnam (SBV) has mandated the implementation of Basel II through Circular 41/2017/TT-NHNN, requiring banks to adopt this framework by 2020 The application of Basel II enables banks to effectively quantify risks associated with all activities and transactions, allowing them to establish strategic operational directions that optimize profits while minimizing credit risks.
● The outbreak of the COVID-19 pandemic
The Covid-19 pandemic, which began in early 2020, has had profound effects on the global economy, leading to significant challenges for businesses and threatening their financial stability As a result, the non-performing loan (NPL) ratio across the banking industry has risen sharply since the onset of the pandemic.
As of the end of 2021, the State Bank of Vietnam reported an internal bad loan ratio of 1.9%, reflecting a year-on-year increase of 0.21% When factoring in debts sold to the Vietnam Asset Management Company (VAMC), the total bad loan ratio rose to 3.9% Additionally, the gross bad debt ratio, which encompasses internal bad debts, unresolved debts sold to VAMC, and potential bad debts from restructuring, surged to 7.31% at the end of 2021, up from 5.1% at the end of 2020, nearing the 7.4% figure recorded at the end of 2017.
Figure 3: Non performing loan and loan loss coverage ratio 2016 – 2021
Source: The author researched and summarised
The Covid-19 pandemic, particularly the Delta variant's fourth wave in 2021, exacerbated the pre-existing issue of rising bad debts within the credit institutions system, leading to significant losses for businesses and impacting livelihoods Financial reports from 2021 indicate a notable increase in bad debts at several banks, with VPBank experiencing a 60% rise, Vietinbank at 49%, VIB at 58%, and HDB at 43% Overall, the average bad loan balances of 28 listed commercial banks, along with Agribank, saw a 17.3% increase compared to 2020.
The COVID-19 pandemic significantly affected financial metrics, notably causing sharp fluctuations in the provisioning ratio By the end of 2021, the loan loss coverage rate had surged to its highest level, reflecting the ongoing challenges faced by the sector.
Balance sheet non-performing loans Gross non-performing loans
According to MIB data, the banking sector has experienced significant growth over the past five years, with an average increase of approximately 145%, more than doubling previous figures Notably, banks like Vietcombank and MB have achieved record highs, reaching 424% and 268%, respectively This upward trend is also evident among state-owned banks, such as Vietinbank and BIDV, as well as in joint-stock commercial banks like Sacombank, VPBank, and LienVietPostBank, which have consistently maintained strong performance levels.
● The period after the COVID-19 epidemic
In 2022, the global economy, including Vietnam, began to recover from the recession caused by the COVID-19 pandemic, presenting new opportunities for the banking industry in 2023 However, this recovery is accompanied by significant credit risk pressures, as highlighted by Hieu, Nguyen (2023), who notes that geopolitical uncertainties, such as the Russia-Ukraine war and tightening monetary policies in the U.S and Eurozone to combat inflation, are impacting Vietnam's economy Following nine consecutive interest rate hikes by the Fed since March 2022, further increases are anticipated, adversely affecting the VND exchange rate against the USD and domestic interest rates Additionally, domestic challenges persist due to the lingering effects of the pandemic, particularly in managing credit risks associated with declining stock and real estate markets High interest rates, rising inflation, and slow growth, compounded by crises in these markets, will have direct repercussions on banks' operations in both the short and long term.
According to data from Mirae Asset Securities (2022), banks are experiencing a decline in credit quality, with bad debts becoming evident in financial statements following the conclusion of the restructuring period for COVID-19-affected customers, as outlined in Circular 14/2021/TT-NHNN, which ended in June The internal non-performing loan (NPL) ratio reported by the State Bank at the end of 2022 highlighted these growing concerns.
In Q4 2022, the non-performing loan (NPL) ratio stood at 1.92%, remaining within a safe threshold However, new bad debts surged by over VND 56,000 billion, significantly exceeding figures from Q3 2021 This rise in the NPL ratio is primarily attributed to the increasing likelihood of debt defaults and potential capital losses.
5) This debt group has increased by more than VND 30000 billion, up 70% compared to the end of 2021
Figure 4: Non performing loans ratio 2022
Source: The author researched and summarised
FiinGroup's analysis highlights the banking system's vulnerability to potential bad debt risks stemming from its real estate credit portfolio, which includes loans to real estate investors, home buyers, and cross-bad debts linked to real estate bonds Statistics from Saigon Liberation Magazine indicate that real estate-related bad debts represent around 20% of the total bad debts within credit institutions.
Figure 5: The ratio of real estate outstanding to total outstanding loans
Source: The author researched and summarised
The credit quality of the real estate market has deteriorated due to liquidity constraints and falling profits, compounded by preferential home loans and the maturity of principal and interest repayments in 2023, amid a decline in personal income following COVID-19 Currently, approximately 70% of bank loan collateral is tied to real estate, making it challenging to issue collateral and sell debts through market mechanisms to address bad debts when the real estate market faces difficulties.
In response to rising bad debts, many commercial banks have significantly increased their risk provisioning to bolster their defenses against potential losses Notably, Vietcombank reported a remarkable bad debt coverage ratio of 317% by the end of 2022, followed by MB at nearly 300%, ACB at 155%, TPBank at 135%, and Techcombank at 125% However, not all banks maintain risk provision ratios that align with the growth of their bad loans, as evidenced by statistics indicating that 10 out of 27 banks have provision ratios below 50% Some banks have even opted to spread their structural debt provisions according to permissible mechanisms rather than fully setting them aside.
Techcombank BIDV VietinBank VPBank VIB ACB MSB
The ratio of real estate outstanding to total outstanding loans
Summary statistics
2.3.1 Analyse correlation coefficients between variables
Before conducting the empirical procedure and descriptive statistics for the sample, it is essential to perform a correlation analysis This study utilizes Pearson’s pairwise correlation matrix and variance inflation factor to assess relationships among variables As shown in Figure 6, the pairwise correlation indicates a relatively low correlation among all explanatory variables.
LLC NPL EG ER UR IF SIZE ROA CAP DIV LLC 1.0000
Figure 6: Pearson's pairwise correlation matrix
This thesis evaluates multicollinearity in the sample by conducting a Variance Inflation Factor (VIF) analysis As shown in Figure 7, all VIF values for the explanatory variables are relatively low and remain within the acceptable range, with none exceeding 4 Consequently, it can be concluded that the dataset does not exhibit multicollinearity.
Figure 7: Variance inflation factor (VIF)
This study employed the Levin, Lin, and Chu (2002) panel unit root test and the Fisher-ADF tests to examine unit roots in a panel dataset The tests aimed to determine the stationarity of variables, with the null hypothesis indicating non-stationarity Results, illustrated in Figure 8, revealed that nearly all variables were stationary, except for SIZE and IF To achieve stationarity for these two variables, their first differences, denoted as dSIZE and dIF, were utilized.
Figure 8: Panel unit root test results
Source: Stata Note: For this test the null hypothesis of unit root is tested against the alternative of stationarity
Empirical Results
As the first step, this study conducts Pooled OLS, Fixed Effects Model (FEM) and Random Effects Model (REM) Table exhibit the results from these models a LLC
F test Prob>F = 0.0000 < 0.05: Between FEM and Pooled OLS, FEM is accepted
Prob>chibar2=0.000 < 0.05: Between Pooled OLS and REM:
Table 7: Summary results of OLS, FEM and REM of model with dependent variable
Source: Stata Note: The standard errors are reported in parentheses The bold coefficients denote the statistically significant values Asterisks indicate the significance at the 1 percent (***),
The Hausman test is essential for identifying the appropriate methodology for static panel data analysis, specifically between the fixed effects model and the random effects model Following this, the Xtserial command is utilized to perform the Wooldridge test for autocorrelation Finally, the variance change is assessed using specific commands to ensure the robustness of the analysis.
- imtest, white to perform White test for variable variance, after OLS regression using reg command
- xttest3 to perform Modified Wald testing for FEM model
- xttest0 to perform the Breusch and Pagan Lagrangian multiplier test in the REM model
Figure 9: Results of Hausman, White and Modified Wald tests of model with dependent variable LLC
At a 1% significance level, the null hypothesis is rejected, confirming the acceptance of fixed effects in the LLC's regression model, indicating that the Fixed Effects Model (FEM) is the most appropriate for this research Nevertheless, the White and Modified Wald tests reveal that the FEM models exhibit issues with autocorrelation.
This study employs ivregress 2sls regression and the Durbin Wu-Hausman test to examine endogenous phenomena affecting the dependent variable, LLC The findings indicate that the inclusion of variables dSIZE and ROA leads to the presence of endogeneity in the model.
Variable Wu-Hausman statistic p-value Conclusion
UR 0.0661 0.7975 Variable is exogenous dIF 0.0345 0.8529 Variable is exogenous
Figure 10: Test results for endogenous variables of the model with dependent variable LLC
Similarly, the model with the dependent variable NPL is carried out with Pooled OLS, FEM, and REM regression Table 8 summarises the results obtained from Stata software
F test Prob>F = 0.0039 < 0.05: Between FEM and Pooled OLS,
Prob>chibar2=0.0008 < 0.005: Between Pooled OLS and
Figure 11: Summary results of OLS, FEM and REM of model with dependent variable NPL
Figure 12: Results of Hausman, White and Modified Wald tests of model with dependent variable NPL
The REM model is identified as the most appropriate model for analyzing the dependent variable NPL; however, the Breusch and Pagan test indicates the presence of autocorrelation issues within the model To address this, the study employs ivregress 2sls regression and the Durbin Wu-Hausman test, revealing that ROA is the endogenous variable when NPL is treated as the dependent variable The results of the endogenous variable tests are summarized in Table 13.
Variable Wu-Hausman statistic p-value Conclusion
UR 0.0641 0.8004 Variable is exogenous dIF 0.3055 0.5813 Variable is exogenous dSIZE 2.5787 0.1107 Variable is exogenous
Figure 13: Test results for endogenous variables of the model with dependent variable NPL
3.3.2 Dynamic Panel Data Two-Steps System GMM Estimation Method
The Generalized Method of Moments (GMM) estimation method effectively addresses model defects, such as variable variance and the issue of endogeneity, where explanatory variables are not entirely independent and exhibit a two-way relationship This endogeneity can undermine the effectiveness of Fixed Effects Model (FEM) and Random Effects Model (REM) estimations To tackle this, the study employed the xtabond2 command for endogenous variables using a two-step selection process, yielding significant results as illustrated in Figure 14.
Figure 14: Regression analyses using GMM estimation techniques
Source: Stata Note: The standard errors are reported in parentheses The bold coefficients denote the statistically significant values Asterisks indicate the significance at the 1 percent (***),
The Sargan (or Hansen) test is employed to assess the suitability of instrumental variables in the GMM model, with its null hypothesis indicating that exogenous instrumental variables are uncorrelated with errors A higher P-value in the Hansen test signifies better instrument validity Additionally, the Arellano-Bond test identifies autocorrelation in errors at first differences, disregarding first-order autocorrelation (AR1) and focusing on second-order autocorrelation (AR2) The results, shown at the end of Figure 14, indicate that all P-values exceed 5%, confirming the appropriateness of the models used Consequently, the estimates derived from the System GMM method are deemed reliable.
Discussion
Bank size +/- + Accept +/- - No statistical significance
Figure 15: Summary of model results
The findings presented in the GMM tables provide valuable insights into the factors affecting credit risk, as indicated by LLC and NPL.
Figure 14 confirms several hypotheses, particularly highlighting that economic growth has a negative coefficient at a 10% significance level, supporting hypothesis 1 and aligning with the findings of numerous scholars in this field The underlying rationale suggests that during favorable economic conditions, businesses and households tend to be more active, fostering positive expectations for banks regarding customers' loan repayment abilities Consequently, customers face fewer challenges in repaying their debts, which enhances banks' credit quality and allows them to lower their buffers to invest in more profitable avenues.
The exchange rate significantly negatively impacts the non-performing loans coverage ratio, aligning with Chaibi's (2015) findings in Germany, contradicting prior hypotheses This phenomenon occurs because an increase in the exchange rate, indicating an appreciation of foreign currency, enhances the financial capacity of exporting entities that receive USD This is evident in Vietnam's status as an export surplus country, with the United States being its largest export customer.
2022 figures show that US export turnover reached 109.1 billion USD, accounting for nearly 30% of total export turnover
The analysis reveals a significant relationship between the unemployment rate and credit risk, supporting hypothesis 3 Specifically, a 1% increase in the unemployment rate correlates with a 3.16% rise in the non-performing loan (NPL) ratio among Vietnamese commercial banks at a 10% significance level This finding aligns with previous research by Ghosh (2015), Chaibi (2015), Dimitrios (2016), and Naili (2022) As job opportunities diminish, borrowers struggle to repay loans, leading to an increase in non-performing loans In response to the challenging labor market conditions, banks have also raised their bad debt coverage ratios to mitigate credit risks effectively.
This research confirms the positive interlinkage between inflation and credit risk in commercial banks Under the escalating inflation circumstance, the purchasing
The weakening power of money has led to a depreciation in the real value of revenue, significantly affecting debtors' ability to repay loans During inflationary periods, both households and businesses face increased expenses to meet their needs, resulting in greater challenges in making timely loan repayments.
The literature indicates a positive correlation between bank size and loan loss coverage (LLC), while the relationship with non-performing loans (NPL) remains unclear This study aligns with hypothesis 5 and reflects the context of Vietnamese commercial banks, where large, capital-intensive state-owned enterprises (SOEs) dominate borrowing These SOEs often utilize large banks for their financial needs, which exposes banks to risks during economic downturns Additionally, banks tend to streamline appraisal processes for lending to these large SOEs, increasing the risk of overdue debts As of the end of 2021, state-owned enterprises accounted for VND 462,245 billion in outstanding bank loans, representing 4.4% of the total national debt.
In which, 4 state-owned enterprises have loans from large commercial banks such as: PetroVN, EVN, Vinacomin, Vinachem
This research indicates a significant negative relationship between Return on Assets (ROA) and credit risk in Vietnamese commercial banks, confirming the initial hypothesis at a 5% significance level This finding aligns with previous studies by Louzis (2012), Ghosh (2015), Dimitrios (2016), and Naili (2022), which suggest that banks with lower profitability tend to exhibit poor lending management skills, leading to an increase in non-performing loans (NPLs) Additionally, these banks experience inconsistent profit levels.
"Friendly" credit policies can facilitate access to multiple loans, addressing profit shortfalls; however, this may increase the risk of a high bad debt ratio in the future Conversely, banks that demonstrate higher profitability often adopt more cautious lending practices, conducting thorough appraisals of loans prior to approval, which ultimately enhances credit quality.
The variable of bank capitalization, representing equity relative to total assets, lacks a robust statistical foundation in both models This limitation may stem from the State Bank of Vietnam's regulations on capital adequacy ratio requirements for commercial banks.
According to Circular 22/2019/TT-NHNN, banks must maintain a capital adequacy ratio of at least 9% As a result, banks consistently uphold a comparable level of capital, which minimizes the impact of this variable on credit risk.
GMM estimates indicate that increased income from interest activities in banks correlates with a rise in non-performing loans (NPL), contradicting hypothesis 8 but aligning with Naili's research findings.
(2022) This can be approached through this following perspective A bank with a high ratio of interest income over total income will be dependent on credit activities
As a result, the likelihood that banks may experience credit risk climbs with the ratio
As banks diversify their operations into sectors like investment banking, insurance brokerage, venture capital, and trading, their lending activities may decline, potentially reducing non-performing loans Conversely, banks with a high ratio of income from interest-generating activities tend to show a negative correlation, where increased interest income leads to lower provisions for bad debts This relationship highlights the impact of diversified income streams on banks' financial health.
Banks engage in non-interest activities, such as investing in corporate bonds, which carry inherent risks These investments can heighten credit risk, particularly when businesses delay interest and principal payments Additionally, banks face cross risks as companies often turn to bond issuance to access alternative capital sources when they can no longer secure loans from banks.
To effectively diversify their business activities, banks must prioritize risk management across both markets This involves reallocating resources from credit risk management to other sectors, which can heighten credit risk exposure and necessitate increased provisions for bad debt coverage.
The analysis reveals a significant connection between past and present credit risk, aligning with the findings of Duong and Nguyen (2017) This suggests that Vietnam's banks struggle with effective credit risk management, as non-performing loans from previous periods persist and continue to impact current financial stability.
Conclusion and recommendation
Conclusion
Credit risk poses a significant threat to the stability of the banking industry, particularly in Vietnam This study conducts a qualitative analysis of the determinants influencing credit risks within Vietnamese commercial banks Utilizing panel data and the Generalized Method of Moments, alongside the Arellano-Bond approach (1991), the research addresses endogenous variables to examine the factors affecting credit risks from 2015 to 2022 The analysis encompasses both bank-specific and macroeconomic factors, providing insights into the dynamics of credit risk in Vietnam's banking sector.
The exchange rate is the most significant macroeconomic determinant affecting businesses with USD-denominated debt and commercial banks, particularly as the FED maintains a stringent monetary policy that pressures exchange rates and heightens credit risks Other macroeconomic factors, including economic growth, unemployment rates, and inflation, also notably impact the credit risk faced by banks The findings align with real-world observations, indicating that stable and developed economic conditions lead to improved bad debt situations, while rising unemployment and high inflation contribute to a decline in credit quality for banks.
The determinants of bank characteristics include five key factors: bank performance (ROA), bank capitalization (CAP), bank size (dSIZE), bank diversification (DIV), and lagged credit risk (NPLt-1, LLCt-1) A significant positive correlation between last year's credit risk proxy highlights the critical role of risk management practices at commercial banks in Vietnam, underscoring their effectiveness and importance in maintaining financial stability.
Recommendation
4.2.1 Improve resilience to volatile macroeconomic situation
Research indicates that macroeconomic factors, including economic growth rate, exchange rate, unemployment rate, and inflation rate, significantly influence credit risk in Vietnamese commercial banks Based on these findings, the thesis presents recommendations for both commercial banks and the government.
The Governor aims to stabilize the macroeconomy by closely monitoring both domestic and foreign market developments, implementing effective strategies to address challenges posed by the lingering effects of COVID-19.
The State Bank of Vietnam (SBV) must prioritize the oversight and inspection of commercial banks to ensure the transparency and stability of the banking system Additionally, it is essential for the SBV to address existing challenges and enhance the effectiveness of the 2% interest rate support package for enterprises This involves implementing policies that enable banks to lower interest rates while maintaining stable liquidity, thereby fostering market confidence Such measures are crucial for supporting credit institutions in providing affordable credit to stimulate economic recovery in the near future.
The author advocates for commercial banks to enhance the availability of a 2% interest rate lending package aimed at supporting businesses struggling post-pandemic By facilitating access to affordable capital, banks can bolster business resilience and financial stability, thereby reducing credit risks Additionally, this approach fosters self-improvement during periods of geopolitical tension, which significantly impact global economies Implementing Basel II and Basel III standards will further enhance banking quality, establishing a sustainable framework to mitigate potential future losses.
4.2.2 Control the size of the commercial bank
Large bank sizes typically enhance loan loss coverage, reflecting variable credit risk in the banking sector Currently, many Vietnamese commercial banks may be engaging in "backyard" lending, primarily directing their resources towards a select group of large enterprises Despite regulations stating that a credit institution's total loan balance to a single customer cannot exceed 15% of its own capital, and the combined loan and guarantee balance must not surpass 25%, companies can circumvent these limits by utilizing subsidiaries to secure bank loans To address these challenges and improve credit quality, the author recommends several strategies based on the model's findings and the current landscape.
For the Governor , the thesis recommends tightening the Law on Credit
To manage the loan balance structure of credit institutions effectively, the Government should enforce diversity within the Board of Directors and strengthen the role of independent management through the Law on Credit Institutions With credit balances representing 21.2% of the total system credit, it is crucial for the Government to implement stricter regulations to oversee the flow of credit capital into the real estate market, particularly focusing on the emergence of new real estate types.
The author recommends that SBV enhance its information system to effectively track bank ownership and reduce centralized control Additionally, there should be a focus on improving the monitoring of cash flow movements, particularly given the rise of fintech applications and retail lending institutions associated with shell companies This approach will help address the potential for borrowing across various banking groups.
Commercial banks, particularly those with substantial capital and a concentrated customer base, must implement a minimum risk factor of 200% for real estate loans in accordance with Circular 41 Additionally, they should adhere to Basel II's capital adequacy ratio (CAR) of at least 8% to ensure liquidity and resilience in case any key client encounters repayment difficulties.
4.2.3 Improving profitability of commercial banks
Increased bank performance reduces the bank's credit risk giving the idea of suggestions and recommendations to improve profitability and increase the bank's resilience to risk
The Governor must ensure a balanced growth across all components of the economic market, treating not only the banking sector but also other companies and organizations that contribute to the nation’s financial stability with fairness Creating an optimal environment for innovation and nurturing a robust banking sector are key objectives of the government, given its critical role in effectively managing and regulating the economy.
The author recommends that the State Bank of Vietnam (SBV) implement a standardized system for developing internal credit rating systems across commercial banks Currently, each bank operates under its own regulations, resulting in inconsistencies when comparing and evaluating the same customers, which can adversely affect customer quality and diminish bank profits A normative system would provide banks with a unified framework to assess customer quality effectively and reduce the incidence of poor-quality loans.
In the current economic climate marked by substantial inflation, commercial banks must exercise prudence in their credit application procedures, particularly when selecting borrowers The impact of restrictive monetary policies is leading to higher lending rates and increased reserve ratios Consequently, banks should carefully monitor and evaluate loans while expanding their credit operations to avoid the pitfalls of prioritizing profits, which can result in elevated levels of non-performing loans.
From the modelling results, it can be seen that diversity in income sources has a significant effect on banks' credit risk Therefore, this is also a key issue that needs
58 to be monitored closely and take effective management steps in the future In the detail:
The article suggests that the Governor should implement strategies to encourage banks to boost their non-interest revenue It emphasizes the need for a comprehensive legal framework governing non-interest income and e-banking As digital transformation accelerates, non-interest profit activities are increasingly flourishing through banks' e-banking platforms Establishing a unified legal framework for e-banking will incentivize banks to expand their services and product offerings in the digital landscape.
The thesis suggests that the State Bank of Vietnam (SBV) should establish a professional training system focused on non-profit activities for commercial banks As modern non-interest services increasingly rely on advanced technologies, there is a pressing need for skilled human resources who are proficient in these technologies By developing this training system, the SBV aims to provide commercial banks with high-quality, well-trained personnel, enabling them to effectively harness and maximize the potential of the non-interest operational sector.
In 2023, commercial banks are focusing on enhancing profitability by promoting non-interest income services such as insurance consulting, e-banking, and payment services, especially amid high global and Vietnamese interest rates resulting from tightened monetary policy As capital costs rise and credit growth is limited by State Bank allocations, non-interest income becomes crucial for sustainable growth, offsetting declines in credit income To maximize effectiveness, banks must improve the quality of non-credit services through technology and align their non-interest income proportions across various categories, including service fees, foreign exchange trading, and securities trading, to implement targeted promotional strategies.
4.2.5 Completely settle outstanding bad loans
The GMM model results indicate that poor credit quality from the previous fiscal year significantly impacts the current year's credit risk To address this issue, the author proposes several recommendations.