INTRODUCTION
Introduction
Since 2012, Vietnamese banks have faced a temporary liquidity shortage, prompting a heightened focus on liquidity risk management This shift aims to bolster customer trust and prevent potential failures within the banking system.
Liquidity in commercial banks refers to their ability to fund asset increases and meet obligations without incurring significant losses Banks play a crucial role in transforming short-term deposits into long-term loans, which exposes them to liquidity risk (BIS, 2008) The Basel Committee (2009) emphasized that a bank's viability hinges on its liquidity position Research by Diamond and Dybvig (1983) highlighted the essential role banks play in creating liquidity Furthermore, Moussa (2015) noted that the level of liquidity risk is closely tied to effective banking operations, indicating that poor liquidity management can lead to insolvency during high liquidity risk periods and low profitability during excessively low liquidity risk.
Liquidity risk in banks remains a critical and timeless subject, with extensive research conducted across various economies and regions As banking systems continue to evolve, ongoing studies in this field are essential for adapting to societal changes and national policy developments.
Liquidity risk is influenced by both internal and external factors Internal determinants include bank size, capital ratio, credit risk, net interest margin (NIM), deposit ratio, return on assets (ROA), and return on equity (ROE) In contrast, external factors encompass the inflation rate (INF) and economic growth (GDP).
This study examines the factors influencing bank liquidity risk in Vietnam and offers recommendations to enhance the stability of liquidity in Vietnamese commercial banks moving forward.
Objectives of the study
This research aims to identify the key factors influencing liquidity risk in Vietnamese commercial banks and to propose practical recommendations for strengthening their liquidity systems, thereby establishing a robust defense against liquidity risk.
Finding out and examining the determinants affecting liquidity risk of commercial banks in Vietnam
Verify the impact levels, impact directions of those determinants on the liquidity risk of commercial banks in Vietnam
Proposing practical recommendations and policies for a better liquidity system of Vietnamese commercial banks.
Research questions
This study will focus on the following questions:
The liquidity risk of commercial banks in Vietnam is influenced by several key factors, including market conditions, regulatory frameworks, and bank-specific characteristics Market volatility and economic stability play significant roles in determining liquidity levels, while strict regulations can either mitigate or exacerbate risk Additionally, internal factors such as management practices and asset-liability mismatches significantly impact liquidity Understanding the varying levels of these factors is crucial, as some may positively enhance liquidity, while others can pose substantial risks, ultimately affecting the overall stability of the banking system.
(iii) What solutions can be taken to improve and maintain the optimal liquidity level and avoid the sudden liquidity risk for Vietnamese commercial banks?
Research’s subject and range
The subject of this study is the liquidity risk of commercial banks in Vietnam
The sample of the research is collected from financial reports of 25 commercial banks which are listed on stock exchanges in Vietnam in the period from
Scientific and practical significance
The findings of this thesis serve as a valuable reference for administrators, policymakers, and scholars, aimed at enhancing the efficiency of banking operations and advancing research and management within the banking sector.
Research methodology
The model that will be estimated in this study is:
LIQ = β 0 + β 1 *ROA it + β 2 *NIM it + β 3 *SIZE it + β 4 *CAP it + β 5 *TLA it + β 6 *CEA it + β 7 *TDES it + β 8 *INF t + β 9 *GDP t +ε it
LIQ serves as the dependent variable reflecting a bank's liquidity position, highlighting its capacity to manage liquidity shocks effectively A higher liquidity ratio signifies that the bank is better equipped to handle unexpected withdrawals.
2004), and, in contrast, the lower the risk of liquidity level
Return on assets reflects the efficiency of the banks in assets utilizing (Tran et al., 2019)
NIM indicates the efficiency of financial intermediation (Hamadi and Awdeh,
SIZE = Size of the bank = natural logarithm of total assets
Size can show the economies of scale The large banks benefit from economies of scale which reduces the cost of production and information gathering (Boyd and Runkhle, 1993)
𝑻𝒐𝒕𝒂𝒍 𝒂𝒔𝒔𝒆𝒕𝒔 This ratio measures the proportion of equity to total assets
The Total Loans to Assets (TLA) ratio indicates the proportion of total loans relative to a bank's total assets A higher TLA ratio signifies that a bank has a greater amount of illiquid assets, which increases its vulnerability to liquidity risk.
Operating expenses include personal expenses and other expenses (Moussa,
2015) CEA shows the percentage of operating expenses in the total assets
Deposits comprise demand deposits and term deposits TDES illustrates the proportion of total deposits to the total assets
INF = Rate of inflation INF shows the increase in the price index
GDP serves as a key indicator for evaluating a nation's economic health During recessions, banks typically maintain higher liquidity reserves to mitigate loan risks In contrast, during economic growth phases characterized by elevated interest rates, banks tend to decrease their liquidity reserves to boost lending activities (Tran et al., 2019).
To enhance the reliability of research findings and address the limitations of individual methods, this study employs a combination of qualitative and quantitative approaches The quantitative method is utilized to identify relationships and correlations among variables, while the qualitative method serves to validate the results of the data analysis.
The research framework involves developing a study sample and collecting data through secondary data collection methods The author gathers information from annual reports, cash flow statements, and published business results from commercial banks' websites, focusing on the period from 2010 to 2020.
Data processing involves the calculation of collected data as variables using Microsoft Excel Subsequently, these variables are analyzed through econometric models, including Pooled OLS, Fixed Effects Model (FEM), and Random Effects Model (REM), utilizing Stata 14 statistical software.
Quantitative method: Implementing the multivariate regression model
The qualitative method will employ various techniques, including description, synthesis, comparison, and analysis, to identify the key determinants affecting the liquidity risk of commercial banks and to develop an appropriate research model.
Step 1: Review background theory and previous studies
Step 5: Analyze the regression results and discuss the research results
Step 2: Build model and research methods
Step 3: Analyze the impact of the determinants on liquidity risk
Step 4: Test the regression model
Step 6: Suggest policy implications and limitations of research
This study analyzes panel data from 25 Vietnamese joint-stock commercial banks listed on stock exchanges between 2010 and 2020, utilizing information gathered from their financial statements Additionally, macroeconomic data was sourced from the State Bank and the General Statistic Office websites to enhance the analysis.
Research content
This study comprises five distinct chapters, each designed to present specific tasks Below, we provide a brief overview of each chapter and its objectives.
This chapter outlines the significance of the paper, establishing the research objectives, questions, subject, and scope Additionally, it clarifies the meaning and contributions of the research, setting the foundation for the subsequent analysis.
LITERATURE REVIEW
Theoretical foundations of bank liquidity
2.1.1 The definition of bank liquidity
The Basel Committee (2008) defines bank liquidity as the capacity of a bank to fund asset increases and timely meet obligations without incurring losses A bank is considered to be in a sound liquidity position when it maintains an adequate amount of available capital or can quickly raise capital through borrowing or asset sales This is crucial as banks often face immediate liquidity demands Additionally, banks frequently encounter liquidity issues stemming from their customers' financial challenges; when customers experience liquidity shortages, they may resort to taking loans or withdrawing deposits, which can negatively impact the bank's liquidity.
2.1.2 Supply and demand of liquidity
Trương Quang Thông (2010) assumed that supply of liquidity is the bank's source to meet liquidity needs, including:
- Cash and cash equivalents: This is considered one of the most major and essential sources for the immediate liquidity needs of banks
To enhance liquidity, banks primarily rely on customer deposits, which serve as a critical funding source To attract more deposits, financial institutions can implement various strategies, including adjusting deposit rates, offering promotional incentives, and enhancing their overall reputation in the market.
- Credits refunded by the customers: This is the main profit-making activity of banks, using deposit capital to commit credits and earn interest payments
- Selling assets: When the bank meets liquidity demand, it can exchange a part of assets into cash
- Market Loans: Banks can borrow money in the monetary market from other commercial banks, which can afford large and immediate liquidity needs
- Borrowing from the central bank: Banks can also borrow short-term loans from the central bank at a discount rate
Trương Quang Thông (2010) assumed that liquidity demand is the need to pay for the committed financial obligations of the bank, including:
Depositors' withdrawal demands are the primary source of liquidity needs for banks, characterized by their permanence, immediacy, and unconditional nature This includes various types of deposits such as non-term, payment, and term deposits, as well as any deposits eligible for early withdrawal Consequently, banks must consistently maintain liquidity reserves to effectively accommodate these withdrawal requests.
- Providing credits to customers: Credit activities bring a large source of income but also contain risks such as capital loss, affecting the solvency of the banks
- Repaying due loans: This includes repayments of loans borrowed from other banks, central banks, and other repurchase agreements
- Operating and tax expenses: This includes all of the expenses related to the bank’s operation such as salaries, bonuses, electricity and water use, advertising expenses, tax
- Dividend payments: Dividend payments for banks’ shareholders
Net liquidity status is calculated by subtracting the value of liquidity supply from the value of liquidity demand of banks (Trương Quang Thông, 2010) In which,
A liquidity surplus occurs when the supply of liquidity surpasses its demand, resulting in an excess of available liquidity resources Thông (2013) suggests that this situation indicates economic inefficiencies and a scarcity of business opportunities, as well as banks' challenges in connecting with the market and customers To address this surplus, banks should consider investing in riskier assets that offer higher returns.
A liquidity deficit occurs when the demand for liquidity surpasses its supply, leading to a shortage of available liquidity sources This shortage can have severe repercussions for a bank, including missed business opportunities, loss of customers, decreased market share, and erosion of public trust Consequently, bank managers must quickly determine effective strategies to mobilize liquidity sources and assess the associated costs.
- Liquidity balance: The liquidity is called balanced only when the liquidity supply equals liquidity demand However, this is nearly impossible in reality
Banks must establish a tailored liquidity policy that balances liquidity and profitability, as maintaining high liquidity often requires sacrificing potential profit opportunities Conversely, a low liquidity level can enhance income but may expose banks to liquidity deficits and unforeseen risks Ultimately, banks with strong liquidity positions are those that can readily access sufficient, immediate, and cost-effective liquidity sources to meet their demands.
Theoretical foundations of liquidity risk
Liquidity risk, as defined by the Basel Committee on Banking Supervision, arises when a financial institution does not have enough capital to fulfill its obligations without disrupting its daily operations and overall financial health This situation leads to challenges in providing adequate cash for immediate liquidity needs (Tran et al., 2019) Essentially, liquidity risk can also be understood as the inability to promptly liquidate a position at a fair price (Muranaga and Ohsawa).
Goodhart (2008) identified two key aspects of liquidity risk: maturity transformation and the inherent liquidity of a bank’s assets Maturity transformation occurs when there is a duration mismatch between a bank’s assets and liabilities, such as borrowing short-term funds while lending long-term (Duttweiler, 2009) Inherent liquidity refers to the ability to sell assets without significant loss in value, which can be compromised if matured liabilities cannot be covered promptly, forcing banks to sell assets at lower prices These two facets are interconnected; banks with high liquidity assets can better manage maturity transformation, while those with short-term maturing assets may not require a large volume of liquid assets.
Nguyen (2013) highlighted the connection between liquidity risk and the sensitivity of financial assets to interest rate fluctuations When interest rates change in the financial market, banks offering lower interest rates may lose depositors to those with higher rates Additionally, borrowers may postpone debt payments or seek additional funds from banks that provide credit at lower rates As a result, interest rate volatility negatively affects banks' cash inflows and outflows, widening the gap between liquidity demand and supply, which increases liquidity risk.
The "domino effect" significantly impacts banks, as highlighted by Tran et al (2019) When one bank faces liquidity issues or nears bankruptcy, its interconnectedness with other banks leads to widespread repercussions The closer the relationships among banks, the more severe the consequences For instance, if depositors rush to withdraw their funds from a struggling bank, it may prompt concerns about the stability of related banks, causing more depositors to withdraw their money This chain reaction can jeopardize the entire banking system.
2.2.3 Liquidity risk and performance of banks
Liquidity problems may affect a bank’s earnings and capital and in extreme circumstances may result in the collapse of an otherwise solvent bank (Arif and Anees,
In times of liquidity crises, banks often resort to borrowing at exorbitant interest rates or selling assets at reduced prices, leading to diminished earnings Additionally, a rise in debt can elevate the debt-to-equity ratio, complicating the banks' ability to sustain an optimal capital structure.
The liquidity position of a bank is quite essential from a marketing perspective
If a bank has a liquidity problem, it may decrease its competitiveness compared to other banks in many business opportunities, such as funding a potential project (Arif and Anees, 2012)
Choosing a bank with high liquidity risk is unwise for depositors, as it may lack the reliability needed to safeguard their funds Depositing in a financially troubled bank increases the risk of not being able to withdraw their money Additionally, a decline in liquidity stability hampers a bank's ability to attract deposits, which are crucial for maintaining liquidity, ultimately leading to more severe liquidity challenges.
Liquidity risk can be assessed through two primary methods: liquidity risk ratios and the financing gap According to Trương Quang Thông (2013), the financing gap represents the difference between total assets and total equity, applicable to both current and future scenarios In contrast, liquidity risk ratios are derived from a bank's balance sheet data and are commonly utilized to forecast liquidity risk trends (Vodova, 2013).
There are 4 liquidity risk ratios often used in empirical research:
The liquidity ratio reflects the proportion of total liquid assets to total assets, with a higher ratio indicating a stronger liquidity position for the bank A high liquidity ratio signifies that the bank possesses a substantial amount of liquid assets, serving as a safeguard against unexpected liquidity shortages This metric has been utilized in research studies by Moussa (2015), Bourke (1989), and Tran et al (2019).
The L2 ratio is a critical indicator of a bank's liquidity risk, assessing its liquid assets against various liquidity sources, including individual and household deposits, enterprise funds, and short-term capital A high L2 ratio signifies strong liquidity, indicating that the bank is well-positioned to withstand liquidity challenges Research by Tran et al (2019) and Fola (2015) supports the importance of the L2 ratio in evaluating a bank's financial health.
L2 ratio as a dependent variable in their paper
The loan-to-assets ratio indicates the percentage of a bank's total assets that are allocated to loans, with a higher ratio suggesting that the bank is heavily invested in credit provision This elevated ratio signifies a greater liquidity risk, as difficulties in managing credit risks, such as bad debts or overdue loans, could lead to challenges in meeting customer deposit withdrawals or interest payments In simple terms, a higher loan-to-assets ratio correlates with an increased liquidity risk for the bank.
This ratio was used in Lucchetta's (2007) and Vodová's (2011) studies to measure the liquidity risk of banks
The L4 ratio assesses how effectively a bank's assets can satisfy the liquidity demands of loans, particularly focusing on highly liquid assets like customer deposits and short-term funding Similar to the L3 ratio, a higher L4 ratio indicates an increased liquidity risk for the bank This metric has been referenced in prior studies, including those by Vodová (2011), Tran et al (2019), and Fola (2015).
While liquidity risk ratios are commonly used to assess liquidity risk, alternative methods, such as the "Financing Gap" concept proposed by Saunder and Cornett (2006), offer valuable insights This approach has been further validated by research from Chen et al (2018) and Arif & Anees (2012) In Vietnam, scholars like Trương Quang Thông (2013) and Đặng Văn Dân (2015) have built upon these foundational studies, employing the financing gap to measure liquidity risk effectively The financing gap (FGAP) is calculated using a specific formula, emphasizing its relevance in liquidity risk assessment.
The financing gap represents a warning sign of a bank's future liquidity risk
A positive financing gap (FGAP) indicates that a bank may need to decrease its cash reserves and liquid assets or seek additional borrowing in the money market, which can elevate liquidity risk Consequently, as the financing gap widens, the bank's reliance on money market loans increases, heightening the potential for liquidity challenges.
Literature review
A study by Valla et al (2006) examined the impact of internal factors and macroeconomic variables on banks' liquidity risk in France over a 12-year period from 1993 The regression analysis revealed that the scale of banks negatively affects liquidity risk, while factors such as economic growth, lending support, short-term interest rates, profit, and credit growth exhibit inverse relationships with liquidity risk.
Vodova (2011) demonstrated that banks' liquidity risk in the Czech Republic is influenced by both internal and external factors, based on data from 2001 to 2009 The study revealed that a higher capital adequacy ratio, lower bad debt ratio, and reduced interbank interest rates negatively impact liquidity risk, while liquidity risk is positively associated with inflation rates, economic growth rates, and financial crises.
The research of Moussa (2015) assumed that liquidity risk is an essential variable for the bank and the banking system components Moussa used a sample of
Between 2000 and 2010, a study of 18 banks in Tunisia assessed liquidity risk using two measures: liquid assets to total assets and total loans to total deposits Utilizing both static and dynamic panel methods, the findings revealed that factors such as financial performance, capital to total assets, operating costs to total assets, GDP growth rate, inflation rate, and delayed liquidity significantly influence bank liquidity risk Conversely, variables including bank size, total loans to total assets, financial costs to total credits, and total deposits to total assets were found to have no significant impact on liquidity risk.
A study by Lucchetta (2007) analyzed the liquidity risk of 5,066 banks from 1998 to 2004, utilizing internal and macroeconomic variables as control factors This approach distinguished Lucchetta’s research from others by focusing on the interbank lending process to estimate the relationships among banks in the interbank market.
In a study conducted by Trương Quang Thông (2013) in Vietnam, the determinants of liquidity risk were analyzed using a fixed-effect model based on a sample of 212 observations The findings revealed a non-linear relationship between total assets and bank liquidity risk, indicating that an initial increase in assets reduces liquidity risk, but beyond a certain threshold, further asset growth leads to increased liquidity risk Additionally, liquidity risk is significantly influenced by two key factors: the external funding dependence ratio and the liquidity reserve to total assets ratio.
Fola's 2015 research on Ethiopian commercial banks from 2002 to 2013 examined the relationships between credit growth, economic growth, margin interest rates, bad debt, and inflation with banks' liquidity risk Contrary to expectations, the findings revealed that only credit growth positively impacted liquidity risk, while economic growth, margin interest rates, bad debt, and inflation exhibited negative correlations with liquidity risk.
Munteanu (2012) analyzed a panel of 27 banks in Romania over the period from 2002 to 2010, emphasizing the differences between the pre-crisis years (2002-
The study examines liquidity rates during the years 2007 and the crisis period from 2008 to 2010, utilizing measurement methods such as Net Loans/Total Assets and Liquid Assets/Deposits and Short-term Funding The findings reveal both shared and unique determinants affecting the two analyzed liquidity rates, aligning with existing literature Notably, the Z-score, a key indicator of bank stability, significantly impacts bank liquidity risk during the crisis years.
Calomiris et al (2015) demonstrated the importance of maintaining cash reserves for banks to ensure stable liquidity This practice enables banks to effectively manage large-scale withdrawals, thereby preventing potential illiquidity crises.
Aspachs et al (2015) conducted a comprehensive analysis of the determinants influencing UK banks' liquidity policy, examining both idiosyncratic and macroeconomic factors Their research revealed that increased potential support from the central bank during liquidity crises correlates with lower liquidity buffers held by banks Additionally, they found that UK banks tend to adopt a counter-cyclical liquidity policy, maintaining lower liquidity during economic upturns The study also tested the hypothesis that these counter-cyclical liquidity buffers are influenced by financial constraints on banks' lending policies, finding substantial support for this notion.
RESEARCH METHOD
Proposing a research model
This paper examines liquidity risk theory and its influencing factors, building on previous research The chosen regression model is based on Moussa's 2015 study published in the highly regarded International Journal of Economics and Financial Issues, which boasts a reliable H-index of 23.
The study of Moussa used a sample of 18 banks in Tunisia in the period from
Between 2000 and 2010, a study estimated two measures of liquidity risk: liquid assets to total assets and total loans to total deposits Utilizing both static and dynamic panel methods, the research revealed that factors such as financial performance, capital to total assets, operating costs to total assets, GDP growth rate, inflation rate, and delayed liquidity significantly impact bank liquidity risk Conversely, size, total loans to total assets, financial costs to total credits, and total deposits to total assets were found to have no significant effects on bank liquidity risk.
The model that will be estimated in my research is:
LIQ = β 0 + β 1 *ROA it + β 2 *NIM it + β 3 *SIZE it + β 4 *CAP it + β 5 *TLA it + β 6 *CEA it + β 7 *TDES it + β 8 *INF t + β 9 *GDP t +ε it
Variables’ descriptions
3.2.1 The dependent variable – Liquidity risk measure ( LIQ )
This study utilizes the liquidity risk measure known as the L1 ratio, which is the ratio of liquid assets to total assets, to assess a bank's capacity to handle liquidity shocks A higher liquidity ratio signifies that the bank is better equipped to manage unexpected withdrawals, as outlined by Chagwiza (2014).
The L1 ratio (LIQ) is identified as the most effective measure of liquidity risk, as it compares liquid assets to total assets, clearly indicating a bank's liquidity position A low L1 ratio signifies a significant liquidity shortage and an increased risk of liquidity issues Additionally, the data required for calculating the L1 ratio is more readily available, particularly in the context of Vietnamese commercial banks, which often lack full transparency and disclosure of certain indicators For these reasons, the author has chosen the L1 ratio (LIQ) as the dependent variable for this study.
Return on assets reflects the efficiency of the banks in assets utilizing (Tran et al., 2019)
NIM denotes the difference between the interest income earned from borrowers and the interest paid to creditors and depositors NIM indicates the efficiency of financial intermediation (Hamadi and Awdeh, 2012)
Size can show the economies of scale The large banks benefit from economies of scale which reduces the cost of production and information gathering (Boyd and Runkhle, 1993)
SIZE = size of the bank = natural logarithm of total assets
The equity to total assets ratio indicates a bank's leverage strategy, with a lower ratio signifying a higher level of leverage While a high leverage strategy can amplify risks for the bank, it also tends to diminish profitability, especially when borrowing costs are elevated (Tran et al., 2019).
The Total Loan Assets (TLA) ratio, representing the proportion of total loans to total bank assets, serves as a key independent variable in this research to assess liquidity risk, as previously discussed in Chapter 2.
Operating expenses include personal expenses and other expenses (Moussa,
2015) CEA shows the percentage of operating expenses in the total assets
Deposits comprise demand deposits and term deposits TDES illustrates the proportion of total deposits to the total assets
INF shows the increase in the price index This variable does not come from the intrinsic of banks but relates to macroeconomic
3.2.2.9 The economic growth rate ( GDP )
Gross Domestic Product (GDP) serves as a key indicator for evaluating a nation's economic health During recessions, banks typically maintain higher liquidity reserves to mitigate loan risks, while in times of economic growth with elevated interest rates, they tend to lower these reserves to boost lending activities (Tran et al., 2019) Like inflation, GDP is not exclusive to any single bank but reflects the overall economy.
Table 3-1 outlines the descriptions and calculation methods for all variables utilized in the model, highlighting their significance in previous research and the author's anticipated impact on liquidity risk.
Moussa (2015), Tran et al (2019), Lucchetta (2007)
SIZE Size of the bank log(𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠) -
Trương Quang Thông (2013), Vodová (2011), Đặng Văn Dân
Tran et al (2019), Trương Quang Thông (2013), Đặng Văn Dân
Tran et al (2019), Trương Quang Thông (2013), Đặng Văn Dân
Notes: This table reports the variables descriptions for commercial banks in Vietnam over the period
From 2010 to 2020, the study analyzes liquidity risk (LIQ) in relation to various financial metrics, including return on assets (ROA), net interest margin (NIM), bank size (SIZE), equity ratio (CAP), total loans ratio (TLA), operating expenses ratio (CEA), total deposits ratio (TDES), inflation rate (INF), and economic growth rate (GDP).
Research methodology
This research employs a mixed-methods approach, incorporating both qualitative and quantitative techniques Utilizing Stata 14 software, the author conducts regression analysis to evaluate the impact of various factors within a model that includes one dependent variable and nine independent variables The methodology is outlined through a series of systematic steps.
Step 1: The author makes a review of previous research related to the liquidity and liquidity risk of banks in both Vietnam and foreign countries Therefore, those research results will be analyzed and considered as a base for identifying variables and building a research model
Step 2: According to the theoretical foundations and empirical studies, the author builds an appropriate model and relevant research methods are applied to calculate the chosen variables having impacts on liquidity risk in the next step
Step 3: Based on the foundation built in step 2 and the collected data of variables, the author exercises models regression Then the results about independent variables’ impact on the outcome variable are all analyzed
Step 4: In this step, the author conducts some tests such as testing for multicollinearity, autocorrelation, and heteroskedasticity In case there is any defect in the selected model, it will be replaced by the FGLS model which can overcome those defects
Step 5: The author analyzes the regression model and makes discussion about the research results
Step 6: The author suggests policy implications and limitations of the research
These steps are summarized in Figure 3-1 shown below:
3.3.2 Research sample and descriptive statistics
This research analyzes data from 25 commercial joint-stock banks in Vietnam, listed on stock exchanges, over an 11-year period from 2010 to 2020 Financial data for these banks is sourced from their published reports on finance and securities portals, while macroeconomic data is obtained from the State Bank and the General Statistics Office of Vietnam.
According to Green (1991), in order to determine the size of the sample, we can apply the formula: n ≥ 50 + 8m In which, n is the minimum sample size required
Step 1: Review background theory and previous studies
Step 5: Analyze the regression results and discuss the research results
Step 2: Build model and research methods
Step 3: Analyze the impact of the determinants on liquidity risk
Step 4: Test the regression model
Step 6: Suggest policy implications and limitations of research
FGLS and m is the number of independent variables Since there are 9 independent variables in this research, the sample size required is 122, whereas the observations of the study
= 25 (banks) * 11 (years) = 275 > 122 Therefore, it can say that the research sample meets the minimum requirement to perform the regression
The commercial banks in the research scope are listed in Table 3-2 below:
Table 3-2: List of commercial banks in the research sample
No Bank Code Bank Name
1 ABB An Binh Commercial Joint Stock Bank
2 ACB Asia Commercial Joint Stock Bank
3 BID Joint Stock Commercial Bank for Investment and Development of Vietnam
4 BVB Viet Capital Commercial Joint Stock Bank
5 CTG Vietnam Joint Stock Commercial Bank for Industry and Trade
6 EIB Vietnam Export Import Commercial Joint Stock Bank
7 HDB Ho Chi Minh City Development Commercial Joint Stock Bank
8 KLB Kien Long Commercial Joint Stock Bank
9 LPB Lien Viet Post Commercial Joint Stock Bank
10 MBB Military Commercial Joint Stock Bank
11 MSB Maritime Commercial Joint Stock Bank
12 NAB Nam A Commercial Joint Stock Bank
13 NVB National Citizen Commercial Joint Stock Bank
14 OCB Orient Commercial Joint Stock Bank
15 PGB Petrolimex Group Commercial Joint Stock Bank
16 SGB Saigon Bank for Industry and Trade
17 SHB Saigon - Hanoi Commercial Joint Stock Bank
18 SSB Southeast Asia Commercial Joint Stock Bank
19 STB Saigon Thuong Tin Commercial Joint Stock Bank
20 TCB Vietnam Technological and Commercial Joint Stock Bank
21 TPB Tien Phong Commercial Joint Stock Bank
22 VAB Vietnam Asia Commercial Joint Stock Bank
23 VCB Joint Stock Commercial Bank for Foreign Trade of Vietnam
24 VIB Vietnam International Commercial Joint Stock Bank
25 VPB Vietnam Prosperity Commercial Joint Stock Bank
Variable Obs Mean Std Dev Min Max
This table presents the descriptive statistics of panel regression results for commercial banks in Vietnam from 2010 to 2020, focusing on liquidity risk as the dependent variable (LIQ) Key independent variables include return on assets (ROA), net interest margin (NIM), bank size (SIZE), equity ratio (CAP), total loans ratio (TLA), operating expenses ratio (CEA), total deposits ratio (TDES), inflation rate (INF), and economic growth rate (GDP) Detailed definitions of these variables are provided in Table 3-1.
Total number of observations = 25 (banks) * 11 (years) = 275
The number of variables = 10 (including LIQ – dependent variable and 9 other independent variables)
LIQ (Mean = 32.5%) The liquid assets represent 32.5% of total assets with a high standard deviation (12.3%) The greatest value is 0.754 belonging to BVB in
2018 while the smallest one is 0.100 (SGB in 2014)
ROA (Mean = 0.8%) The return represents 0.8% of total assets, which shows a low average return on assets of banks The standard deviation of this variable is 0.7%
NIM (Mean = 2.7%) The interest margin represents 2.7% of total assets with a low standard deviation which is 1.1%
SIZE (Mean = 14.032) The sizes of most banks are considered small and medium Moreover, their sizes do not have much variance since the standard deviation is 0.501
CAP (Mean = 9.4%) The capital represents on average 9.4% of total assets
The differences in this ratio between banks are not significant, which can be observed from a low standard deviation (4.09%)
TLA (Mean = 57.4%) The loans represent on average 57.4% of total assets
But there is a large variety of this variable between observed banks (0.123)
CEA (Mean = 1.6%) Operating expenses represent 1.6% of total assets There is a slight variance of 0.5% of CEA between banks
TDES (Mean = 75.1%) Deposits represent 75.1% of total assets, which shows banks’ high ability to attract deposits from customers And the standard deviation of
The average inflation rate during the research period was 5.8%, accompanied by a high standard deviation of 4.8%, indicating significant year-to-year fluctuations This variability is further illustrated by the range between the lowest inflation rate of 0.63% and the highest rate of 18.58%.
GDP (Mean = 6%) The average growth of gross domestic product from 2010 to 2020 is 6% with a little standard deviation (1.1%) Therefore, we can say that the variances between years are insignificant
The study utilizes panel data, also known as longitudinal data, which integrates time series and cross-sectional data Time series data consists of observations from a single subject over various time intervals, while cross-sectional data captures observations from multiple subjects at a specific point in time Panel data, therefore, combines these two approaches, providing observations for multiple subjects across multiple time instances The fundamental structure of panel data is represented by the regression model y_it = α + β x_it + u_it, where y_it denotes the dependent variable, α represents the intercept, β is a vector of parameters for the explanatory variables, and x_it is a vector of observations on those variables, with t ranging from 1 to T and i from 1 to N.
3.3.3.1 Models used for panel data
There are three predominant approaches to regress panel data, which are Pooled OLS, Fixed effect model (FEM), and Random effect model (REM)
The Pooled OLS model is specified as below: y it = α 1 + β 1 x 1it + β 2 x 2it + β 3 x 3it + + β k x kit + u it
Where y it is the dependent variable of observation i, x kit is a 1 x k vector of observations on the explanatory variables, t = 1, , T, i = 1, N
According to Brooks (2008), FEM can be specified as: y it = C i + β x it + u it
In this study, the dependent variable (y) represents the observation for unit i at time t, while the independent variable (x) corresponds to the same unit and time Each research unit has its own intercept (C), and the slope (β) is determined by the relationship between the independent factors (x) and the error term (u).
The random-effects model can be written as: y it = C i + β x it + u it
In contrast to the Fixed Effects Model (FEM), which maintains a constant intercept (C_i) over time, the Random Effects Model (REM) assumes that the intercepts for each cross-sectional unit stem from a shared intercept (C) that remains uniform across all units and time periods Additionally, REM incorporates a random variable (ε_i) that introduces variability across different cross-sectional units while remaining stable over time.
After exercising the three models above, a model will be pointed out as the most suitable one for the research by using some tests
F-test is used to evaluate the appropriateness between POLS and FEM Its hypotheses are:
H0: POLS is the suitable model for research
H1: FEM is the suitable model for research
Prob > F > 0.05 (p-value > 0.05): H0 is accepted => POLS is the suitable model for research
Prob > F < 0.05 (p-value < 0.05): H0 is rejected => FEM is the suitable model for research
When using the POLS model after conducting the F-test, the Hausman test is not required However, if the FEM model is selected, the Hausman test is essential to determine whether to use FEM or REM The test evaluates specific hypotheses to guide this decision.
H0: There is no correlation between εit and independent variables
H1: There is a correlation between εit and independent variables
Prob > chi2 > 0.05 (p-value > 0.05): H0 is accepted => REM is the suitable model for research
Prob > chi2 < 0.05 (p-value < 0.05): H0 is rejected => FEM is the suitable model for research
3.3.3.3 Tests for defects of the chosen model
In this research, the Modified Wald test is used to check whether there is a heteroskedasticity phenomenon or not in the chosen model with the following hypotheses:
H0: There is no heteroskedasticity phenomenon
H1: There is a heteroskedasticity phenomenon in the model
Prob > chi2 > 0.05 (p-value > 0.05): H0 is accepted => There is no heteroskedasticity phenomenon in the model
Prob > chi2 < 0.05 (p-value < 0.05): H0 is rejected => There is heteroskedasticity phenomenon in the model
The Wooldridge test is used to check for the appearance of autocorrelation in the model Its hypotheses are:
H0: There is no first order autocorrelation
H1: There is sequence autocorrelation phenomenon
Prob > F > 0.05 (p-value > 0.05): H0 is accepted => There is no first order autocorrelation in the model
Prob > F < 0.05 (p-value < 0.05): H0 is rejected => There is sequence autocorrelation phenomenon
When defects are identified in the model, the Feasible Generalized Least Squares (FGLS) method will be employed to address these issues effectively.
This chapter presents a model grounded in theoretical foundations and prior research, detailing the variables utilized, their definitions, calculation methods, and anticipated impacts on banks' liquidity risk It also outlines the research sample and the procedural steps taken during the study Additionally, the chapter illustrates various models, including Pooled OLS, FEM, REM, and FGLS, along with the criteria used to select the most appropriate model for the analysis.
RESEARCH RESULTS
Research results
Table 4-1: Correlations between variables LIQ ROA NIM SIZE CAP TLA CEA TDES LIQ 1.000
This table presents the correlations among various variables derived from panel regression analyses of commercial banks in Vietnam from 2010 to 2020 The dependent variable, LIQ, measures liquidity risk, while ROA represents the return on assets ratio Additionally, NIM indicates the net interest margin, SIZE refers to bank size, CAP denotes the equity ratio, TLA represents the total loans ratio, and CEA signifies the operating expenses ratio.
TDES is the ratio of the total deposits, INF is the inflation rate, GDP is the economic growth rate
The definition of variables is presented in Table 3-1
Table 4-1 illustrates the impact and level of influence of independent variable pairs, highlighting their correlations within the model The correlation coefficient ranges from -1 to +1, with -1 signifying a perfect negative correlation, +1 indicating a perfect positive correlation, and 0 denoting no correlation Notably, a variable's correlation with itself will always equal 1 As all coefficients have absolute values below 0.8, this indicates that multi-collinearity is not a significant concern.
This table presents the Variance Inflation Factor (VIF) ratios from panel regression analyses conducted on commercial banks in Vietnam between 2010 and 2020 The analysis focuses on liquidity risk, represented by the dependent variable LIQ Key independent variables include the return on assets (ROA), net interest margin (NIM), bank size (SIZE), equity ratio (CAP), total loans ratio (TLA), operating expenses ratio (CEA), total deposits ratio (TDES), inflation rate (INF), and economic growth rate (GDP) For detailed definitions of these variables, please refer to Table 3-1.
Table 4-2 presents the VIF ratio, a key indicator for identifying multicollinearity within a model A low VIF suggests minimal multicollinearity, while a VIF below 10 indicates that multicollinearity can be deemed negligible The findings in Table 4-2 reveal that all VIF ratios are below 10, confirming the absence of multicollinearity in the model (Hoàng & Chu).
4.1.2 Regression results of three models: Pooled OLS, FEM, and REM
This study aims to provide a clear comparison among three statistical models: the Pooled Ordinary Least Squares (POLS) model, the Fixed Effect Model (FEM), and the Random Effect Model (REM), by estimating each model's results.
Table 4-3: Results of Pooled OLS, FEM, and REM
This table presents the findings from three models—POLS, FEM, and REM—analyzing the liquidity risk of commercial banks in Vietnam from 2010 to 2020 The dependent variable, LIQ, measures liquidity risk, while key independent variables include ROA (return on assets), NIM (net interest margin), SIZE (bank size), CAP (equity ratio), TLA (total loans ratio), and CEA (operating expenses ratio).
TDES represents the ratio of deposits, while INF indicates the inflation rate and GDP reflects the economic growth rate A detailed definition of these variables can be found in Table 3-1 The analysis employs three methodologies: (1) Pooled Ordinary Least Squares (POLS), (2) Fixed Effects Model (FEM), and (3) Random Effects Model (REM).
The analysis presented in Table 4-3, calculated using Stata, reveals that only three out of nine independent variables—CAP, TLA, and TDES—are statistically significant in the POLS model These variables exert a negative influence on the dependent variable, LIQ, indicating that increases in CAP, TLA, and TDES will lead to a decline in bank liquidity and an increase in liquidity risk.
The estimation results of the FEM model show similarities to the Pooled OLS model in terms of CAP and TLA, but notable differences exist While the Pooled OLS model identifies TDES as statistically significant, this effect on LIQ is absent in the FEM model Conversely, the FEM model incorporates two additional statistically significant variables, SIZE and GDP, both of which contribute to a decrease in LIQ due to their upward trends.
The results of the REM model closely align with those of the FEM model, showing greater similarity than with Pooled OLS in terms of the number of significant variables and their directional effects Notably, liquidity (LIQ) exhibits negative relationships with all the statistically significant variables, including size (SIZE), capital (CAP), total liabilities (TLA), and GDP.
4.1.3.1 Choosing between Pooled OLS and FEM
F-test is used to evaluate the appropriateness between POLS and FEM Its hypotheses are:
H0: POLS is a suitable model for research
H1: FEM is a suitable model for research
The result illustrates that Prob > F = 0.0000 which means p-value < 0.05 Therefore, H0 should be rejected and the FEM model is the selected one
4.1.3.2 Choosing between FEM and REM
To determine the appropriate model between Fixed Effects Model (FEM) and Random Effects Model (REM), the Hausman test is essential This statistical test evaluates the suitability of FEM over POLS, with specific hypotheses guiding its analysis.
H0: There is no correlation between εit and independent variables
H1: There is a correlation between εit and independent variables
Accordingly, the estimated result shows that Prob > chi2 = 0.0008 (p-value < 0.05) Therefore, H1 is accepted and FEM is the more preferred model between these two
4.1.4 Tests for defects of FEM model
In this research, the Modified Wald test is used to check whether there is a heteroskedasticity phenomenon or not in the chosen model with the following hypotheses:
H0: There is no heteroskedasticity phenomenon
H1: There is a heteroskedasticity phenomenon in the model
The test illustrates that Prob > chi2 = 0.0000 (p-value < 0.05), proving that H0 is rejected and there is a phenomenon of heteroskedasticity in the model
The Wooldridge test is used to check for the appearance of autocorrelation in the model Its hypotheses are:
H0: There is no first-order autocorrelation
H1: There is a sequence autocorrelation phenomenon
In result, Prob > F = 0.1253 (p-value > 0.05), so H0 is accepted which means that the model is not suffering from autocorrelation phenomenon
Since there is a defect (heteroskedasticity) in the chosen model – FEM, the FGLS model will also be regressed in order to overcome that defect of FEM
Table 4-4: Results of FGLS model LIQ Coef Std Err z P > |z| [95% Conf Interval]
The table presents the findings from the FGLS model applied to commercial banks in Vietnam for the period from 2010 to 2020, focusing on liquidity risk measured by the dependent variable LIQ Key independent variables include ROA (return on assets), NIM (net interest margin), SIZE (bank size), CAP (equity ratio), TLA (total loans ratio), CEA (operating expenses ratio), TDES (deposits ratio), along with macroeconomic factors such as INF (inflation rate) and GDP (economic growth rate) Detailed definitions of these variables are provided in Table 3-1.
The analysis of the independent variables reveals that out of nine, six are statistically significant, including ROA, CAP, TLA, CEA, TDES, and GDP Notably, ROA and CEA positively influence liquidity (LIQ), whereas the relationships between LIQ and the remaining four variables—CAP, TLA, TDES, and GDP—are negative.
Notes: This table reports the comparison between regressed models for commercial banks in Vietnam over the period 2010 to 2020 The dependent variable is LIQ, which is the liquidity risk measure
The article discusses key financial ratios and metrics used in banking, including ROA (return on assets), NIM (net interest margin), SIZE (bank size), CAP (equity ratio), TLA (total loans ratio), CEA (operating expenses ratio), TDES (deposits ratio), INF (inflation rate), and GDP (economic growth rate), with definitions provided in Table 3-1 The analysis employs various econometric models: Pooled OLS, FEM, REM, and FGLS, to evaluate these financial indicators.
Table 4-5 presents a comparison of four models estimated in this study: Pooled OLS, FEM, REM, and FGLS Initially, the analysis involved Pooled OLS, FEM, and REM, with FEM emerging as the most suitable model after testing However, due to issues of heteroskedasticity in FEM, the FGLS model was introduced as an alternative The FGLS model not only addresses the shortcomings of FEM but also demonstrates a higher number of significant variables compared to the other models Consequently, FGLS is identified as the most appropriate model for this research.
Based on these results, the research model has the following regression equation:
LIQ = 0.747 + 1.568*ROA it – 0.576*CAP it – 0.763*TLA it + 2.544*CEA it –
0.108*TDES it – 0.588*GDP it + ε it
Variable Estimated effect Level of statistics significance
Estimated effect in previous research
CONCLUSION AND RECOMMENDATIONS
Research conclusion
An analysis of data from 25 commercial banks listed on stock exchanges over an 11-year period from 2010 identifies six key determinants of banks' liquidity risk, measured by the liquidity index (LIQ) Notably, a higher LIQ value reflects increased liquidity; thus, the results should be interpreted inversely The determinants include: Return on Assets Ratio (-), which negatively impacts liquidity risk; Equity Ratio (+), Total Loans Ratio (+), and Total Deposits Ratio (+), which positively influence liquidity; and Operating Expenses Ratio (-) and Economic Growth Rate (+), which also affect liquidity risk.
The liquidity risk of banks is positively influenced by several key variables, including the equity ratio (CAP), total loans ratio (TLA), total deposits ratio (TDES), and economic growth rate (GDP) The regression results for CAP align with findings from previous studies by Trương Quang Thông (2013), Vũ Thị Hồng (2015), and Moussa (2015) Additionally, Moussa's research confirms the positive impact of TLA on liquidity risk TDES is utilized in two models within Moussa's study, further supporting its relevance in assessing liquidity risk.
In 2015, two models were analyzed to assess the impact of TDES on liquidity, yielding both positive and negative results; however, these findings were statistically insignificant Additionally, the results regarding GDP align with Vodová's (2011) research conducted in the Czech Republic.
In Vietnam, two key factors inversely affect the liquidity risk of commercial banks: Return on Assets (ROA) and the Operating Expenses Ratio (CEA) While ROA is often overlooked in studies on bank liquidity, this research presents contrasting findings compared to Moussa (2015) Similarly, the impact of CEA on liquidity risk differs from Moussa's conclusions, suggesting that variations in banking systems between the two countries may account for these discrepancies.
Policy implications and recommendations for Vietnamese commercial
As banks expand their size, it is crucial to implement a strategic roadmap that enhances operational efficiency while maintaining control over growth to improve liquidity However, history has shown that a larger bank does not inherently mitigate liquidity risks Consequently, during the expansion process, banks must prioritize increasing their holdings of highly liquid assets to effectively manage and hedge against potential liquidity challenges.
To enhance loan quality, banks must prioritize a rigorous credit approval process, as lending is their primary revenue source but also poses liquidity risks Utilizing deposit capital to issue loans can threaten liquidity if not managed carefully It is crucial for banks to apply stringent criteria to all loan applications to mitigate the risk of bad debts Additionally, imposing excessive loan quotas on credit officers can lead to reckless lending practices, increasing both credit and liquidity risks for commercial banks.
To effectively raise capital and enhance access to funding, banks must regularly evaluate their relationships with equity owners and diversify their capital sources Building strong connections with key capital suppliers, such as partners, agents, potential customers, and payment systems, can serve as a safety net during liquidity challenges Furthermore, commercial banks should implement a capital increase strategy tailored to their size and the varying economic conditions This capital growth must be complemented by efficient utilization and a commitment to stable development.
The State Bank must enhance its guidance and advisory role for commercial banks by consistently analyzing market trends and making accurate forecasts This approach will provide commercial banks with essential insights to inform their liquidity policy development.
The State Bank must conduct regular inspections and effective supervision of commercial banks to ensure the safe and sustainable development of the banking system Emphasizing the required reserve ratio and monitoring credit provision by these banks is essential to mitigate sudden risks.
To enhance the effectiveness of banking inspections, it is crucial to strengthen the organization of the banking inspection apparatus from central to local levels, ensuring the independence of management and professional activities within the State Bank Additionally, the State Bank should implement the fundamental principles of effective banking supervision outlined by the Basel Committee and maintain a prudent approach to inspections.
The government must enhance the legal framework governing liquidity risk management in commercial banks, as the current regulations are still basic and require significant improvements It is essential to establish comprehensive guidelines on liquidity risk to aid banks in their operations Additionally, targeted policies are necessary to stimulate economic growth, utilizing both monetary and fiscal measures to effectively manage the growth rate and control inflation in Vietnam.
Limitations of the study and next research directions
This study aims to identify the factors influencing liquidity risk variation in Vietnamese commercial banks and to propose solutions to enhance their ability to mitigate this risk While the objectives have been successfully met, the research acknowledges certain limitations related to time constraints, data sources, research methodologies, and the author's inexperience.
The research focuses on 25 commercial banks over an 11-year period from 2010 to 2020, which limits its ability to provide a comprehensive and representative overview of the current liquidity risk situation in Vietnam.
Secondly, there are several liquidity risk measures besides LIQ that are still not estimated in the research
Several intrinsic and macroeconomic factors, including External Funding Dependence (EFD), the ratio of Credit Loss Provision to Total Loans (LPTL), Long-Term Lending Interest Rate (LTR), unemployment rates, monetary market interest rates, and the impact of financial crises, are often overlooked but play a crucial role in economic analysis.
Fourthly, because of the inexperience of the author, the proposed solutions may not really be practical and useful to the commercial banks, the State bank, and the Government
To address the study's limitations, the author suggests future research should focus on expanding the number of observations, incorporating additional internal and macroeconomic variables, and employing diverse research models and methods This approach aims to provide a more comprehensive assessment of the factors influencing liquidity risk.
In Chapter 5, the author summarizes the impact of independent variables on liquidity risk and proposes solutions to improve liquidity quality in Vietnamese commercial banks The chapter also highlights the study's limitations and outlines future research directions.
1 Đặng Văn Dân (2015) Các nhân tố ảnh hưởng đến rủi ro thanh khoản của ngân hàng thương mại tại Việt Nam Tạp chí Tài chính, 11(1), 60-66
2 Hoàng Trọng & Chu Nguyễn Mộng Ngọc (2008) Phân tích dữ liệu nghiên cứu với SPSS (Dùng với SPSS các phiên bản 11.5, 13, 14, 15, 16)-Tập 2
3 Vũ Thị Hồng (2015) Các yếu tố ảnh hưởng đến thanh khoản của các ngân hàng thương mại Việt Nam Tạp chí Phát triển và Hội nhập, 23(33), 32-49
4 Nguyễn Thị Tuyết Nga (2019) Những yếu tố tác động đến thanh khoản tại các ngân hàng thương mại Việt Nam, Tạp chí Tài chính, kỳ 1 tháng 7/2019
5 Nguyễn Thị Bích Thuận & Phạm Ánh Tuyết (2021) Nhân tố ảnh hưởng đến rủi ro thanh khoản tại các ngân hàng thương mại Việt Nam, Tạp chí Tài chính
6 Trương Quang Thông (2013) Các nhân tố tác động đến rủi ro thanh khoản của hệ thống ngân hàng thương mại Việt Nam Tạp chí Phát triển Kinh tế, 50-62
1 Anbar, A., & Alper, D (2011) Bank specific and macroeconomic determinants of commercial bank profitability: Empirical evidence from Turkey Business and economics research journal, 2(2), 139-152
2 Arif, A., & Anees, A N (2012) Liquidity risk and performance of banking system Journal of Financial Regulation and Compliance
3 Aspachs, O., Nier, E W., & Tiesset, M (2005) Liquidity, banking regulation and the macroeconomy Available at SSRN 673883
4 Basel Committee on Banking Supervisor (2008) Principles for Sound
Liquidity Risk Management and Supervision, https://www.bis.org/publ/bcbs144.pdf
5 Bourke, P (1989) Concentration and other determinants of bank profitability in Europe, North America and Australia Journal of Banking & Finance, 13(1), 65-79
6 Boyd, J H., & Runkle, D E (1993) Size and performance of banking firms:
Testing the predictions of theory Journal of monetary economics, 31(1), 47-
7 Brooks, C (2008) RATS Handbook to accompany introductory econometrics for finance Cambridge Books
8 Calomiris, C W., Heider, F., & Hoerova, M (2015) A theory of bank liquidity requirements Columbia Business School Research Paper, 2, p10
9 Chagwiza, W (2014) Zimbabwean commercial banks liquidity and its determinants International Journal of Empirical Finance, 2(2), 52-64
10 Chen, Y K., Shen, C H., Kao, L., & Yeh, C Y (2018) Bank liquidity risk and performance Review of pacific basin financial markets and policies, 21(01), 1850007
11 Cooper, M J., Jackson III, W E., & Patterson, G A (2003) Evidence of predictability in the cross-section of bank stock returns Journal of Banking &
12 Diamond, D W., & Dybvig, P H (1983) Bank runs, deposit insurance, and liquidity Journal of political economy, 91(3), 401-419
13 Duttweiler, R (2011) Managing liquidity in banks: A top down approach
14 Fola, B (2015) Factors affecting liquidity of selected commercial banks in
Ethiopia Masters of Science in Accounting and Finance Thesis
15 Goodhart, C (2008) Liquidity risk management Banque de France Financial
16 Green, S B (1991) How many subjects does it take to do a regression analysis Multivariate behavioral research, 26(3), 499-510
17 Hamadi, H., & Awdeh, A (2012) The determinants of bank net interest margin: Evidence from the Lebanese banking sector journal of Money,
18 Iannotta, G., Nocera, G., and Sironi, A (2007) Ownership Structure, Risk and
Performance in the European Banking Industry Journal of Banking & Finance, 31, 2127-2149
19 Lucchetta, M (2007) What do data say about monetary policy, bank liquidity and bank risk taking? Economic notes, 36(2), 189-203
20 Munteanu, I (2012) Bank liquidity and its determinants in
Romania Procedia Economics and Finance, 3, 993-998
21 Muranaga, J., & Ohsawa, M (1997) Measurement of liquidity risk in the context of market risk calculation a BIS volume entitled The Measurement of
22 Negash, D W & Veni, P (2019) Determinants of liquidity risk in selected commercial banks in Ethiopia International Journal of Advanced Research in
23 Nguyen Thi Lan Anh (2013) Study impact of lowering interest rates on financial market of Vietnam from 2012 to present
24 Perry, P (1992) Do banks gain or lose from inflation? Journal of retail banking, 14(2), 25-31
25 Saunders, A., Cornett, M M., & McGraw, P A (2006) Financial institutions management: A risk management approach (Vol 8) New York: McGraw-
26 Tran, T T., Nguyen, Y T., Nguyen, T T., & Tran, L (2019) The determinants of liquidity risk of commercial banks in Vietnam Banks and
27 Vodová, P (2011) Determinants of commercial banks' liquidity in the Czech
Republic Recent Researches in Applied and Computational Mathematics, 92-
28 Vodová, P (2013) Determinants of commercial bank liquidity in
Hungary Finansowy Kwartalnik Internetowy e-Finanse, 9(4), 64-71
29 Valla, N., Saes-Escorbiac, B., & Tiesset, M (2006) Bank liquidity and financial stability Banque de France Financial Stability Review, 9(1), 89-104
30 Wójcik-Mazur, A., & Szajt, M (2015) Determinants of liquidity risk in commercial banks in the European Union
APPENDIX A: DESCRIPTIVE STATISTIC AND CORRELATION
APPENDIX B: RESULTS OF POLS, FEM, REM, AND FGLS
Table B.1 Results of POLS model
Table B.2 Results of FEM model
Table B.3 Results of REM model
Table B.4 Results of FGLS model
NAME YEAR LIQ ROA ROE NIM SIZE CAP TLA CEA TDES