1377 Factors Affecting The Liquidity Risk Of Commercial Banks In Vietnam 2023.Docx

HO CHI MINH CITY, 2022 STATE BANK OF VIETNAM MINISTRY OF EDUCATION & TRAINING HO CHI MINH UNIVERSITY OF BANKING HO CHI MINH UNIVERSITY OF BANKING NGO DANG QUYNH NHI FACTORS AFFECTING THE LIQUIDITY RIS[.]

RESEARCH MOTIVATIONS

Banking is one of the most sentitive industries not only in Vietnam but also throughout the word and it plays an role in the economic development Banks do not only affect but also facilitate the intergration economic activities activities such as mobilizing resources,production activities, public finance distribution and evendistribution of social welfare.Therefore, banking management is always a matter of special concern by government carrying out management and supervision activities.

Liquidity risk occurs when a bank is insolvent, cannot convert assets into cash in time or cannot borrow to meet the needs of payment contracts Thus, if a bank does not have the necessary capital to meet the needs of the market, it may become insolvent If liquidity risk occurs, the impact will not only be limited to a single bank, but also strongly affect other banks and the entire financial system A typical example of the banks’ heavy influence on economy is the global financial crisis that happened in 2007 which led to a series of bankruptcies, bringing, the economic stagnation to its peak. According to Bank for International Settlements, during global financial crisis, many banks struggled to sustain adequate liquidity, a number of banks still failed, being forced into mergers even when receiving extraordinary support from the central banks. Several years before the crisis, liquidity and its management was not really a priority, funding was available at low cost However, this crisis has totally changed market conditions that captured the importance of related liquidity issues measurement thus its management.

Liquidity risk has an impact on a bank's performance as well as its reputation (Jenkinson, 2008) Furthermore, a low liquidity position may result in regulatory penalties As a result, maintaining a sound liquidity structure becomes critical for a bank Liquidity risk has emerged as a major worry and challenge for banks in the modern period A bank

2 with good asset quality, strong earnings, and sufficient capital may fail if it is not maintaining adequate liquidity.

From that, it shows that the importance of assessing the liquidity risk of Vietnamese commercial banks at this stage is very important Light liquidity risk will reduce the bank’s profitability, if severe, it can lead to bankruptcy Therefore, one of the most important tasks of bank managers is to ensure reasonable liquidity and provision for liquidity risk A bank is considered to have good liquidity if it has easy access to available capital at a reasonable cost and at the right time However, a large amount of capital reserve will directly affect the profitability of the bank’s investment Liquidity risk is influenced by many factors, both internal and external to the bank.Therefore, studying the impact of factors will be very important to limit liquidity risk in banking activities.

Stemming from the above reasons, the author has chosen to carry out the research topic

"Factors affecting the liquidity risk of Vietnamese commercial banks" to study to show the factors that have affected the liquidity risk of the Bank, besides, there are proposed methods to improve the liquidity of Vietnamese commercial banks.

OBJECTIVES OF STUDY

General objective

Understanding the impact of a number of factors on the liquidity of Vietnamese commercial banks in the context of Vietnam in the period 2009-2019

Specific objective

Build models based on previous studies.

- Verify the impact of these factors on the liquidity risk of Vietnam Joint Stock Commercial Bank.

- Proposing solutions and recommendations for joint stock commercial banks to improve the liquidity risk of Vietnam joint stock commercial banks, limiting the impact on the liquidity risk of commercial banks.

RESEARCH QUESTION

In order to achieve the research purpose as mentioned above, this thesis is conducted to answer the following questions in turn:

- What are the determining factors affecting the liquidity risk of Vietnamese commercial banks?

- What model and method to measure the liquidity risk of Vietnamese commercial banks?

- How is the impact of the liquidity risk of Vietnamese commercial banks?

- Based on the research results, What solutions to improve the liquidity risk ofVietnamese commercial banks?

SUBJECT AND SCOPE OF THE STUDY

Research Subject

The object of this research is the financial capacity of commercial banks, the factors affecting the liquidity of commercial banks in Vietnam

Scope of the study

- The research using quantitative research method and using Stata software to carry out quantitative analysis.

- The research data is data from financial statements, annual reports of 31 Commercial banks in Vietnam in the period from 2009-2019.

RESEARCH METHODOGY

To overcome the limitations of each method and increase the reliability of the research results, the study employs both qualitative methods, as well as other methods, at the same time a quantitative method is used to detect the relationship and correlation between variables, while a qualitative method is used to validate the data analysis results.

-Methods of data collection: developing research models, designing research samples and collecting data for research To have data for the research, the author used

4 the method of collecting secondary data by taking the data published on the websites of commercial banks such as annual reports, cash flow statements, etc currency, business results, in the period 2009-2019.

-Data processing method: In this study, quantitative research method was used with the support of Stata software The author conducts regression analysis and tests on the acquired panel data in order to construct an appropriate model In particular, regression analysis of panel data using the least squares method (POOLED OLS), random effects method (REM), and fixed effects approach was employed in the study (FEM) The author employs test like Preusch and Pagan and the Hausman test to identify the best model based on panel data regression To deal with issues like variation of variable errors and autocorrelation, the study applies the feasible generalized least squares (FGLS) approach on panel data The author then employs the S-GMM strategy to resolving endogenous issues.

-Qualitative method: used to compare results from empirical analysis with results from previous studies to explain research objectives and research questions.

CONTRIBUTIONS

Theoretically, the thesis complements the building of a research model on liquidity risk of Vietnamese commercial banks Based on reliable data from the audited financial statements to provide empirical evidence on the impact of determinants on the liquidity risk and appropriately selected research models, the study will show the importance of building a sound liquidity system.

In practical terms, the research results of the thesis can be considered as a source of reference, a policy suggestion to help propose solutions to complete the policy framework in the management and administration of Vietnam's commercial banking system in order to improve ability to face liquidity shocks and improve the competitiveness of the current commercial banking system in Vietnam Along with that

5 is the scientific basis for commercial banks and the State Bank to propose appropriate policies to improve the operational efficiency of the banking industry.

DISSERTATION STRUCTURE

The papers divided into 5 chapters, as follows:

Describe reasons for selectingthe topic; review some previous studies;state research objectives, research questions, subjects and scope of research as well as research methods and contributions in practice and the structure of research.

Chapter 2 presents the theoretical basis of the Bank's financial performance, summarizes previous research models on the factors affecting the Bank's financial performance as a basis for building the research model in the next chapter.

Presenting the research model, research methods, methods of data sampling and processing, building and testing scales to measure the impact factors on liquidity risk.

Chapter 4: RESEARCH RESULTS ANF DICUSSION

Presenting the result from the estimation model and discuss the obtained result.

Summarize the main findings of the study, the significance and contribution of the research to the banking sector in particularand to theeconomy in general At the same time,giving some suggestions to Improve the liquidity of the bank This research should create a basis for others to continue to explore and develop, while showing some limitations of research and proposing further research directions.

The first chapter provided an overview of the study's topic The author has specified the research objectives, clearly defined the subject and scope of study, research methodologies, and ultimately the thesis structure, which includes five chapters, after examining the requirement of the research.

LITERATURE REVIEW

THEORY OF LIQUIDITY RISK OF JOIN-STOCK COMMERCIAL

According to Article 4 of the Law on Credit Institutions (Law No 47/2010/QH12),a commercial bank is a place that does money business and provides financial and credit services in accordance with the law.

Commercial banks are one of the financial intermediaries that play an important role in establishing the financial environment.With core activity is transferring money from capital surplus to capital shortage, banks make the idle money tobe fully utilized and make money available to consumers and businesses that they might not be able to earn, or at least not for a very long time.Besides, banks also create creditworthiness of customer by safeguarding money so that good money is only for good loans and not lost on bad loans.In other words, banks connect individuals, businesses and other institutions together that helps keeping the economy going Therefore, if banks fall, it will cause a collapse for a whole system of the economy, and because banks and money are that essential to maintain not only economies but entire societies, they are extremely regulated and must operate by strict procedures and principles.

Furthermore, banks build consumer creditworthiness by safeguarding money so that good money is used for good loans and not wasted on bad loans In other words, banks connect individuals, businesses, and other institutions, which helps to keep the economy going.

As a result, if banks fail, the overall economic system will collapse, and because banks and money are so important to the survival of not only economies but entire societies,they are highly regulated and must adhere to strict procedures and principles.

2.1.2.1 Bank liquidity risk theory definition

Bank for International Settlement defines liquidity as the ability of bank to finance increased assets and meet obligations when due, without incurring unacceptable losses.Therefore, liquidity risk arises when the bank is unable to meet capital needs at some point of time; or must raise capital from other sources with high costs to meet its obligation; or due to other subjective reasons that affects the solvency of the bank, accordingly it will lead to undesireable consequences.In other words, this is the type of risk that occurs in cases when the bank is insolvent due to falling to promptly liquidate assets in a short period of time and at less than market prices.

It can be understood that liquidity risk occurs when commercial banks are not able to pay at a certain time, or have to mobilize capital at high costs to meet payment needs; or for other subjective reasons, cause the insolvency of commercial banks,which will lead to undesirable consequences (Duttweiler, 2009) From the definition of liquidity of a bank, so far, there are a number of different definitions of liquidity risk such as: according to Nguyen Dang Don:“Liquidity risk is a type of risk appearing in the case of a bank lack of ability to pay, not converting in time cash-generating assets or the inability to borrow to meet the requirements of payment contracts” In easier terms, liquidity risk can be defined as the risk of being unable to liquidate a position timely at a reasonable price (Muranaga and Ohsawa, 2002).

Liquidity is considered an important factor in determining the safety of a bank's operations as well as the stability of the whole banking system Liquidity risk is contained within systematic risk When one bank experiences liquidity risk, it will have a negative impact on liquidity risk on other banks, the extent and spread of liquidity risk,account is huge.

Many studies have been relatively consistent in showing that risk Liquidity risk may come from the asset or liability side, or from off-balance sheet activities of commercial banks' balance sheets (Valla and Escorbiac, 2006).

Goodhart (2008) assumed that there are two basic facets of liquidity risk: maturity transformation (the maturity of a bank’s liabilities and assets) and the inherent liquidity of a bank’s assets (the extent to which assets can be sold without incurring a significant loss of value under any market condition).

According to Nguyen Van Tien (2010), there are three preconditions that cause banks to face frequent liquidity risks:

- First, banks mobilize and borrow capital in a short time, then every week they repay them for a longer-term loan As a result, many banks face a mismatch in terms of maturity between assets and liabilities.

- Second, the sensitivity of financial assets to interest rate changes When interest rates rise, many depositors will withdraw their money and look for another deposit with a higher interest rate Those with credit needs will postpone or withdraw the entire credit limit balance at the agreed low interest rate.

- Thus, changes in interest rates simultaneously affect the flow of deposits as well as the flow of loans, and ultimately the bank's liquidity Thirdly, banks always have to meet their liquidity needs perfectly Liquidity problems will erode public confidence in banks. Another cause of influence that does not come from within but is influenced from without is called the "domino effect" (Tran et al., 2019) The effect is roughly understood as when banks have a close relationship with each other through transactions in the interbank market When a bank loses liquidity and faces bankruptcy risk, other banks will also be affected The degree of the contagion depends on the size of the transaction between banks In addition, when depositors withdraw money from a bank,others may assume that all other banks will also face liquidity difficulty and withdraw all money from these banks This phenomenon triggers the domino effect that causes trouble to the whole banking system.

Vu Thi Hong (2015) uses data from 37 Vietnamese commercial banks in the period from

2006 to 2011 to study the factors affecting the liquidity risk of banks in Vietnam using FEM and REM models The results show that equity ratio, loan-to deposit ratio, profit ratio and bad debt ratio affect liquidity In addition, the study also shows that the liquidity of the bank is guaranteed if the owner's equity is maintained stably.

Liquidity risk can be measured in two ways: liquidity gap and liquidity ratios Liquidity gap is first mentioned in Risk Management in Banking by J.Bessis (2009) which is estimated by the difference between assets and liabilities at both present and future dates.However, the drawback of this method is that it is very hard to collect data since a minority banks publish their annual reports with liquidity gaps which leading to a strongly unbalanced dataset that is unable to estimate.

Another measurement is liquidity ratios various balance sheet ratios that can be used to identify the main liquidity trends A number of studies has indicated various liquidity ratios as follows:

This ratio provides general information about the liquidity of the bank That is of the total assets of the bank, what is the proportion of liquid assets? This ratio is high, meaning the liquidity of the bank is very good.

Liquid assets (Deposit + short-term mobilized Capital)

LITERATURE REVIEW

There have been many different studies related to the liquidity risk of commercial banks, in which the author is mainly influenced by the research of: BIS (2009) defines liquidity as the ability to meet the needs of using available capital for business activities at all times, such as deposit payment, lending, payment, and banking capital transactions According to Duttweiler (2009), liquidity is the ease with which a particular asset is converted to cash and when a company wants to convert an asset into cash, the market is still capable of accepting transactions Bank liquidity can be divided into two categories: natural liquidity and artificial liquidity.

Natural liquidity is created by a bank's assets with a specified maturity Artificial liquidity is created through the ability to convert assets to cash before maturity.

The research of Chung-Hua Shen et al on bank liquidity risk and operational efficiency The study uses asymmetric panel data of commercial banks in 12 developed economies for the period 1994–2006 and uses the two-stage least squares (2SLS) regression mothod.

The results show that liquidity risk is a decisive factor in the bank's performance.

Liquidity risk can reduce bank profitability because of the high cost of reserve funds Truong Quang Thong (2013) investigated the determinants of liquidity risk by regressing a sample of 212 observations with a fixed-effect model His findings revealed that total assets have a non-linear impact on the bank's liquidity risk To begin with, an increase in assets results in a decrease in liquidity risk However, when total assets reach a certain level, the liquidity risk rises Furthermore, two factors have a significant impact on liquidity risk: the external funding dependence ratio and the liquidity reserve to total assets ratio.

Dang Van Dan (2015) said that the financing gap represents a warning sign of a bank's future liquidity risk If the financing gap is positive and the bank has a large financing gap, then the bank will be forced to reduce cash reserves and reduce liquid assets or borrow additional money in the money market, leading to liquidity risk of the bank will rise

Saunders et al (1990) studied the relationship between bank ownership structure and risk acceptance based on the data of US banks in the period 1979 -1985 Using the Pooled OLS model for seven models corresponding to seven types of RRs with the same independent variables, the results show that the larger the foreign ownership ratio, the higher the RR of the bank The ratio of equity / total assets is almost inverse with the RR of the bank, while the ratio of fixed assets / total assets tends to change depending on the period on the bank's risk.

Foos et al (2010) used data from Bank scope from more than 10,000 private banks in the period 1997–2005 to examine how loan growth affects bank risk through three hypotheses about the relationship between past loan growth and loan losses, bank profits, and solvency The author suggests that when bank lending activities thrive, it will lead to credit damage in the near future, as well as a reduction in the impact on interest income and capital ratio From there, the accumulated losses generate new risks, especially in the liquidity situation However, the study only focuses on 14 large countries, and it is easy to recover from the crisis, so it does not have a comprehensive view of the economy and the research period is quite short.

According to research by Rose (2001), banks have good liquidity when they have a reasonable amount of available capital or can quickly raise capital through borrowing or selling assets Commercial banks always keep a certain number of liquid assets in reserve on their balance sheets Measuring the proportion of these types of assets compared to the operational scale of commercial banks is considered a method to assess the liquidity of commercial banks.

According to research by Munteanu (2012) data collected by 27 banks in Romania in the years 2002-2010 aims to highlight the difference between crisis years The measures used in the study are Loans/Total Assets and Current Assets/Deposits and sources of short-term funding The results for the identified and different factors for the analyzed two-paying rule are consistent with previous literature of the same topic.

Previous empirical studies such as San and Heng (2013), Ongore & Kusa (2013) used different estimation methods to measure and evaluate the factors affecting the financial performance of commercial banks These studies measure the financial performance of commercial banks by three financial metrics: Return on Equity (ROE), Return on Assets (ROA) and Net Interest Margin (NIM)

In 2006, Valla and Escorbiac also published the results of their study However, this study in essence also focuses on some internal and macro factors affecting the liquidity of banks in the UK as studied by Aspachs et al (2005).

Ibrahim, S S (2017) studies the effect of liquidity on profitability at commercial banks in Iraq in the period 2005–2013 The results of OLS regression analysis show that: The general ratio has a positive effect, while the quick ratio has a negative effect, implying that profits can be increased if short-term bank liquidity is guaranteed profits for the bank, but if in the short term, the bank holds too many assets that are instantly solvable, such as cash, demand deposits at other credit institutions, other assets on the market Besides payment ratios, a bank's liquidity management is also shown through a number of other indicators The capital adequacy ratio, the loan ratio, and the general ratio are used to test this problem The higher the loan and investment ratios, the more profitable the bank is, but for the capital adequacy ratio, if the bank keeps this ratio too high, the capital invested in low-risk business contracts will bring a low rate of return.

Ghenimi, A., Chaibi, H., & Omri, M A B (2020) study: liquidity risk determinants:

Islamic banking versus conventional for the period 2005–2015 The results show that credit risk, ROE, liquidity gap, and CAR are common liquidity risk determinants in both banking systems These results can be explained by Islamic law (forbidden to pay or receive interest (Riba), the inefficiencies of Islamic money markets (lack of liquidity) and lack of diversification (banks in Islam in general are highly dependent on real estate) This therefore suggests that regulators should focus more on risk management strategy and management performance Therefore, Islamic banks should manage this risk differently from conventional banks while complying with Islamic Sharia.

Research by Nguyen Phuc Quy Thanh (2020) on the liquidity status and operational efficiency of 31 Vietnamese commercial banks, including state-owned commercial banks and private commercial banks (excluding joint-venture banks, 100% foreign- owned banks, and bank branches) on foreign goods in the period 2005-2015 The thesis has focused on researching theoretical issues and non-parametric methods (DEA) in measuring efficiency and using the Tobit model to analyze the factors affecting the performance of 32 commercial banks Vietnam in the period 2007–2017 On the basis of qualitative analysis combined with quantitative analysis in evaluating the efficiency and determining the factors affecting the performance of commercial banks in Vietnam, the study can give some recommendations I propose to improve the operational efficiency and competitiveness of the current commercial banking system in Vietnam in accordance with the requirements of innovation and the trend of international economic integration The thesis concludes that liquidity status has a positive impact on the performance of Vietnamese commercial banks during the research period When the liquidity status of commercial banks ensures the solvency of obligations when they come due without significant losses, it will contribute to improving the operational efficiency of banks But when this index is too high, it shows that a large amount of capital is not participating in the production process and causes waste forthe bank to serve as a basis for providing solutions and recommendations to improve operational efficiency ofVietnamese commercial banks.

HYPOTHESES DEVELOPMENT

2.3.1.1 LLR – Provision for credit risk

Provision costs for credit losses reflect the quality of the loan or credit risk, if higher provision costs reflect reduced quality of loans and increased exposure to credit risk get a raise Truong Quang Thong (2013), Lucchetta (2007) found a positive correlation between the credit risk provision ratio and the liquidity risk of banks.

H1: the ratio of provisions for credit risk to total outstanding loans has a positive effect on the bank's liquidity risk.

2.3.1.2 LDR: Loan-to-deposit ratio

The higher this ratio means that the bank lends more than the capital it can mobilize. Therefore, when facing liquidity risk, it will be difficult for banks to mobilize cheap capital if they lend too much, which reduces the bank's liquidity, which means increased liquidity risk When this ratio is low, banks can easily mobilize from different sources, such as the interbank market, issue valuable papers, etc., with cheap capital, which increases the bank's liquidity.

H2: A positive relationship exists between liquidity risk and the loan/deposit ratio.

Banks use equity and debt to finance their business operations Unlike loans, which are payable in nature, equity is considered the bank's own funds, representing the ability to fend for themselves in the event of an accident The larger capital banks tend to hold less liquid assets, so the greater the liquidity risk and vice versa This ratio represents the capital adequacy and the safety and financial soundness of a bank This low ratio shows that the bank uses high financial leverage, which contains a lot of risks and can make the bank's profits decrease when the cost of debt is high An empirical study on the impact of the CAP variable on liquidity has different results such as: Thakor (1996), Bunda

(2003), Rupullo (2003), Rupullo (2003), Rupullo (2003) 2003), Gorton & Huang

(2004), Indriani (2004), Aspachs et al (2005), Inoca Munteanu (2012), Chikoko Laurine

(2013), Gorton & Winton (2017) all show that equity over total assets has a positive relationship with liquidity account Therefore, we expect the equity ratio to be positively correlated with the bank's liquidity risk.

H3: The equity ratio has a positive effect on the bank's liquidity risk.

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) The larger the total assets of a bank, the less liquidity risk it is exposed to Large banks can rely on the interbank market, or on liquidity support from the lender of last resort (Vodava1, 2013) The results of some empirical studies show that size has a positive effect on liquidity (O Aspachs et al, 2005), (Chikoko Laurine, 2013) However, some studies have opposite results, size has a negative impact on liquidity (Bunda & Desquilbet, 2008), (Doriana Cucinelli, 2013), (Vodová P, 2013) From the above theories, arguments and empirical research results, the author hypothesizes about the positive relationship between asset size and liquidity of banks.

H4: There exists a positive effect between liquidity risk and bank size.

TLA shows the percentage of total loans in relation to total assets Since loans are illiquid assets, a high TLA ratio indicates that the number of liquid assets held by banks is low and banks easily experience liquidity problems.

H5: Total loans ratio has a positive relationship with liquidity risk

This coefficient is measured by taking after-tax profits on the total equity, which reflects the bank's management effectiveness in the use of equity The bank's profits are mainly generated from traditional businesses, which is the interest rate difference between lending and capital mobilization Therefore, the more assets a bank hold to meet its liquidity needs, the lower its profitability will be and vice versa (Aspachs et al, 2005)

H6: Return on equity ratio has a positive effect on the bank's liquidity risk

Interests receivables (by borrowers), Interests incurred (paid by the bank to the creditors and depositors) NIM indicates the efficiency of financial intermediation

H7: Marginal interest income has a positive effect on the bank's liquidity risk

Inflation rate is one of the important macro factors in the economy, the INF both shows the trend of the economy and is an indicator for the State Bank to adjust economic policies in line with the trend direction of the economy during that period Research by Moussa, M.A.B (2015), Truong Quang Thong (2013), Samuel Siaw (2015) and Tran Thi Thanh Dieu (2020) shows that there is a positive impact between the inflation rate and the bank's liquidity risk.

H8: Inflation rate has a positive relationship with liquidity risk

Economic growth index is one of the macro factors affecting all business activities of all economic sectors, if the high economic growth rate shows that the business activities of the economic sectors are better Therefore, promoting high lending activities increases credit balance and effective loan recovery, reducing credit risks.

Research by Vodova Pavla (2011) and Tran Thi Dieu Thanh (2020) has a positive relationship between economic growth rate and bank liquidity risk In contrast, the study of Moussa, MAB (2015), Truong Quang Thong (2013), Samuel Siaw (2015), Godfrey Marozva (2016) has a negative impact between economic growth rate and liquidity risk of banks row.

H9: Economic growth rate has a positive relationship with liquidity risk

Also known as the total means of payment, includes: the amount of cash in circulation, term deposits, demand deposits and savings deposits of individual and corporate customers at credit institutions On the balance sheet of the State Bank, money supply M2 is a liability and assets are the factors affecting money supply In addition, money supply is also a factor for the State Bank to control inflation and stabilize the money market.

Research by the authors: Truong Quang Thong (2013), Vodova Pavla (2011) and Tran Thi Thanh Dieu (2020) show that there is a positive relationship between money supply ratio and liquidity risk of banks.

H10: Money supply has a positive relationship with liquidity risk

Through the analysis of the theoretical basis, the theoretical bases related to liquidity as well as the previous studies have helped the thesis to have a clearer overview of the factors affecting the liquidity of Vietnamese commercial banks This is the basis for the author to build a research model on the impact of factors on the liquidity ofVietnamese commercial banks made in chapter 3.

RESEARCH METHODS

DATA COLLECTION

We use a data set of Vietnamese commercial banks produced annually between

2009 and 2019 to study the factors affecting the liquidity risk of commercial banks During the study period, this sample excludes banks that were acquired by the State Bank of Vietnam, as well as merged and consolidated banks These banks' operations are subject to huge oscillations that take longer to normalize and their financial statistics may change drastically Furthermore, banks with missing data for the previous five years are excluded from the sample After all was said and done, 31 commercial banks were obtained Vietstock collects financial data from banks.

Table 3.1 List of Commercial banks in Vietnam

1 ABB An Binh Commercial Joint Stock Bank

2 ACB Asia Commercial Joint Stock Bank

3 BAB Bac A Commercial Joint Stock Bank (BaoViet Bank)

4 BID Joint Stock Commercial Bank For Investment And

5 BVB Bao Viet Joint Stock Commercial Bank (BaoViet Bank)

6 CTG Vietnam Joint Stock Commercial Bank For Industry and

7 DAF Dong A Commercial Joint Stock Bank (Dong A Bank )

8 EIB Vietnam Export Import Commercial Joint Stock Bank

9 HDB Ho Chi Minh City Development Joint Stock

10 KLB Kien Long Commercial Joint Stock Bank

11 LPB Lien Viet Post Joint Stock Commercial Bank

12 MBB Military Commercial Joint Stock Bank (MB)

13 MSB Vietnam Maritime Commercial Joint Stock Bank

14 NAB Nam A Commercial Joint Stock Bank (Nam A Bank )

15 NCB National Citizen Bank (NCB)

16 OCB Orient Commercial Joint Stock Bank (OCB)

17 PGB Petrolimex Group Commercial Joint Stock Bank (PG

18 PVB Vietnam Public Bank (PVcomBank)

19 SCB SaiGon Joint Stock Commercial Bank (SCB)

20 SEA Southeast Asia Commercial Joint Stock Bank

21 SGB SaiGon Bank For Industry and Trade (SAIGONBANK)

22 SHB SaiGon – HaNoi Commercial Joint Stock Bank (SHB)

23 STB SaiGon Thuong Tin Commercial Joint Stock Bank

24 TCB Vietnam Technological And Commercial Joint Stock

25 TPB Tien Phong Commercial Joint Stock Bank (TPBank)

26 VAB Vietnam Asia Commercial Joint Stock Bank

27 VBB Vietnam Joint Stock Commercial Bank (Vietbank)

28 VCA Viet Capital Commercial Joint Stock Bank (Viet Capital

29 VCB JointStock Commercial Bank For Foreign Trade Of

30 VIB Vietnam International Commercial Joint Stock Bank

31 VPB Vietnam Prosperity Joint Stock Commercial Bank

RESEARCH MODELS

The research model is formatted according to panel data According to Frees, E.

W (2004), panel data describes surveys of a group of repeatedly surveyed individuals over time In this article, data collected from 31 joint-stock commercial banks in

Vietnam will be researched The data is collected from audited financial statements and posted on the websites of commercial banks and some other financial information sites

The study was carried out over a 10-year period from 2009 to 2019.

Research on liquidity is very important for financial markets and banks, especially since the 2008 economic crisis According to Aspachs (2005) and Nikolau (2009), liquidity is not solely dependent on objective external factors that are important (such as efficient markets, infrastructure, low transaction costs, a large number of buyers and sellers, and the transparency of transactional assets) Importantly, it is influenced by internal factors, especially the reactions of market participants in the face of uncertainty and changes in asset values Until now, research by some authors, such as Aspachs et al

(2005), Rychtárik (2009), Praet and Herzberg (2008), has only focused on adjusting the internal liquidity ratios within banks.

According to the research of author Dang Van Dan (2013), the funding gap method is the most appropriate method in quantitative research The funding gap index reflects the most basic aspect of a bank's liquidity.

Besides, the research paper "Factors affecting liquidity risk of Vietnam's commercial banking system" by author Truong Quang Thong (2013) also shows that, in addition to internal factors of banks, such as size of total assets, ratio of liquidity reserve to total assets, interbank lending to total assets, dependence on external funding sources, ratio of equity to total capital, provision credit risk on total outstanding loans, external factors such as economic growth, inflation in the economy, and money supply in the economy are also factors affecting the liquidity of Vietnam's commercial banking system.

Most research results suggest that, in order to ensure the best liquidity risk management, comes from balancing internal and external factors However, external factors also contribute to important support for liquidity in an account Therefore, the following study emphasizes the importance of the impact of macro factors as well as industry-wide indexes in reducing liquidity risk Inheriting the above studies and applying the funding gap method, we get the following model:

FGAP it = α+ β1LDR it + β2CAP it + β3NIM it + β4SIZE it + β5TLA it + β6ROE it + β7LLR it β8AGDP t + β9CR3CR3 t β10INF t + β11M2 t + β12GDP t + ε it

In which: FGAPit is funding gap (liquidity gap), equal to average total credit balance minus average total mobilized capital, this index measures the liquidity risk of commercial banks. α = Intercept

LDR it = Loan to deposit ratio of bank i at time t

CAP it = Equity capital of Bank i at time t

NIM it = Net Interest Margin of bank i at time t

SIZE it = Bank size of bank i at time t

TLA it = loan-to-total assets ratio of bank i at time t

ROE it = return on equity i at time t

LLR it = Loan loss reserves of bank i at time t

AGDP t = Industry growth at time t

CR3 t = Industry concentration at time t

INF t = Inflation rate at time t

GDP t = Economic growth (GDP) at time t ɛ it = Error term where i is cross sectional and t time identifier

DESCRIPTION VARIABLE AND RESEARCH HYPOTHESIS

3.3.1 Dependent variable – Funding gap (FGAP)

FGAP: Research by Chung Hua Shen et al (2009) has shown the strength of using funding gap to measure liquidity risk versus liquidity ratio Those are liquidity ratios calculated from the balance sheet, so they are often used to predict the direction of liquidity, while the funding gap is calculated as the difference between assets and capital present and future time The use of funding gap to measure is also supported by Gatev and Strahan (2006), Sauders and Corrnet (2007), Arif and Anees (2012), Dang Van Dan (2013)

(Credit balance - Capital mobilization) total assets

The larger a bank's funding gap (FGAP), the greater the need for money market loans and the greater likelihood of liquidity problems based on this funding.

2001) Previous studies by many authors such as Aspachs et al (2003); Bonfim and Kim (2011), Indriani (2004) all show a negative correlation between the ratio of short-

FGAP term mobilized capital and bank liquidity.

Total loans Total short-term deposits

CAP Banks use equity and debt to finance their business operations, unlike loans which are payable in nature, equity is considered the bank's own funds, representing for the ability to fend for themselves in the event of an accident Larger capital banks tend to hold less liquid assets, so the greater the liquidity risk and vice versa Therefore, we expect the equity ratio to be positively correlated with the bank's liquidity risk.

Net interest margin (NIM) is a measure of the net return on the bank’s earning assets, which include investment securities, loans, and leases It is the ratio of interest income divided by total assets NIM is used in many research papers such as: San and Heng (2013), Ongore & Kusa (2013), Gul and Zaman (2011), Robin and Bloch (2018), Bao (2016, Cuong (2017), Duong and Nguyen (2021).

Bank size is measured by taking the natural logarithm of total assets (SIZE) If SIZE has a positive correlation with a bank's liquidity, it means that the more the bank expands, the more liquidity will increase, opening up opportunities for banks to continue to mobilize different capital sources to improve its liquidity Size can show the

Total Assets economies of scale The large banks benefit from economies of scale which reduces the cost of production and information gathering (Boyd and Runkhle, 1993) The formula of bank size is:

SIZE = Log (Total asset) 3.3.2.5 Loan-to-total assets ratio

TLA: In Vietnam, as well as the banking system of emerging economies, banks often focus on using capital sources in traditional activities of lending Ordinary loans have low liquidity Therefore, large and unpredictable withdrawals can lead to loss of bank liquidity.

ROE measures the rate of return on the ownership interest (shareholders' equity) of the common stock owners It measures a firm's efficiency at generating profits from every unit of shareholders' equity (also known as net assets or assets minus liabilities) ROE shows how well a company uses investment funds to generate earnings growth Used in many research papers such as: San and Heng (2013), Ongore and Kusa (2013),

Gul and Zaman (2011), Robin and Bloch (2018) Bao (2016, Cuong (2017) ROE is measured according to the following formula:

LLR: Provision costs for credit losses reflect the quality of the loan or credit risk, if higher provision costs reflect reduced quality of loans and increased exposure to credit risk get a raise Truong Quang Thong (2013), Lucchetta (2007) found a positive correlation between the credit risk provision ratio and the liquidity risk of banks.

Provision for credit risk Total outstanding balance

INF: The variable inflation rate is calculated by the inflation rate of the year of observation

INF = Inflation Index 3.3.3.2 Economic Growth

GDP: GDP shows the growth of economic activity in the country (Ayadi and

According to Friedman (1963), the money supply speed must correspond to the economic growth rate, an excessive money supply will be the source of inflation

Changing money supply, through different tools of the central bank can affect the

GDPLLR ROE liquidity of commercial banks M2 also known as the total means of payment, includes: the amount of cash in circulation, term deposits, demand deposits and savings deposits of individual and corporate customers at credit institutions

The table below shows the descriptions of all the variables used in the model as well as their calculation methods Also, the evidence that these variables were used in previous research and the author’s expectation about their effect on liquidity risk are both shown in Table 3-2.

Table 3.2 Summary of research on financial performance Status Symbol Variable name

MEASURE ES Empirical evidence in previous studies

Credit balance - capital mobilization/tota l assets

Ferrouhi & Lahadiri(2014), Shen and associates (2009);Saunders and Cornett(2006),Arif and Nauman Anees (2012)

1 LDR Loan to deposit ratio

Total loans/Total short-term deposits

Net Interest Margin/Total Assets

4 SIZE Size bank Log (Total asset) + Ahmad, R., Ariff, M., &

5 TLA Loan-to- total assets ratio

+ Bunda và Desquilbet (2008); Shen and associates (2009).

Provision for credit risk/Total outstanding balance

Cucinelli (2013) , Trương Quang Thông (2013), Lucchetta (2007)

Ratio of total bank assets to GDP

The proportion of assets of the three largest commercial banks

(2004), Claeys & Vander (2008), Athanasoglou and associates

Inflation Index - Gul & cộng sự (2011), Vong and associates (2009)

Anbarvà Alper(2011); Ferrouhi (2014); Growe and cộng sự (2014); Ayaydin and Karakaya (2014)

RESEARCH PROCESS

This research uses both qualitative methods and quantitative methods The author uses Stata14 software to regress and determine the effect of factors in the model which consists of 1 dependent variable and 7 independent ones The particular process is demonstrated through the following 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 31 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 variable shaving 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 heteroscedasticity 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.

The least squares estimation model, commonly known as the OLS estimation model is based on the principle of minimizing the sum of the squares of the model's residuals and is used to estimate the coefficients of the explanatory variable on the mean of the dependent variable The residual is defined as the difference between the actual and expected values of the dependent variable as a function of the explanatory factors.

The fixed effects model (FEM) is used to depict the effect of the explanatory variable on the dependent variable while taking into consideration the individual features of the dependent variable when the observed crossover units are not uniform

As a result, FEM assumes that the partial regression coefficients are the same across cross units but that the regression intercepts are different.

For each cross-unit, the REM model estimates different intercepts as well as the overall effect of the explanatory factors Each cross-intercept unit is made up of a common intercept that remains constant across time and subject, as well as a random variable that is a subject-varying but time-variable error component.

3.5.4 Feasible Generalized Least Square (FGLS)

For each cross-unit, the REM model estimates different intercepts as well as the overall effect of the explanatory factors Each cross-intercept unit is made up of a common intercept that remains constant across time and subject, as well as a random variable that is a subject-varying but time-variable error component.

We can use the OLS to estimate the structure of the variable variance instead of assuming it Feasible GLS is the name of this approach (FGLS) If the regression model contains the phenomena of variable variance or autocorrelation, or if both of these phenomena are present at the same time, the generic least squares method of testing (FGLS) is the best strategy to overcome these phenomena in the model The OLS approach will be used to estimate the model in the FGLS method (even in the case of the existence of autocorrelation and variable variance) The variance- covariance matrix of the error will be calculated using the model errors Finally, apply this matrix to transform the original variables and estimate the values of the model's parameters.

This chapter has already provided readers with a detailed overview of where to look for analytical data and how to apply it to the thesis model Furthermore, we continue to demonstrate estimated methods such as Pooled OLS, FEM, REM, and, most notably, FGLS Furthermore, this thesis demonstrates hypothesis tests in relation to the estimated model Then, this chapter will most likely summarize the existence of endogenous, autocorrelation, and heteroscedasticity As a result, this chapter is regarded as the foundation for subsequent chapters.

RESEARCH RESULTS AND DISCUSSION

DESCRIPTIVE STATISTICAL

In order to test the factors effecting the liquidity risk of commercial banks in Vietnam, the study of building a model consists of 13 variables, of which 1 dependent variable is funding gap (FGAP), and 7 independent variables These factors include LDR, CAP, NIM, TLA, ROE, LLR, and bank size (SIZE) Macroeconomic factors include the economic growth rate (GDP), inflation (INF), and money supply (M2) Bank market factors are industry growth (AGDP) and industry concentration (CR3) The variables were surveyed through 260 observations of 31 commercial banks in Vietnam during 2009–2019 Table 4.2 below shows the numerical values describing the quantitative variables in the research model and reflecting the characteristics of every variable.

Table 4.1 Summary of Descriptive statistics

Variable Obs Mean Std Dev Min Max

Source: Calculation results from Stata software

Average total credit balance minus average total mobilized capital (FGAP) of 31 commercial banks in Vietnam from 2009-2019 has an average value of 37,67 % with a standard deviation of 18.65%, with the smallest value being - 2,42 % of Vietnam Maritime Commercial Joint Stock Bank (MSB) in 2014 and the highest value is 71,99% of Vietnam Asia Commercial Joint Stock Bank (VietABank) in 2010.

Return on equity (ROE) has an average value of 8.58% with a standard deviation of 7.25%, with a minimum value of 0.14% of the Kien Long Commercial Joint Stock Bank (Kienlongbank) in 2019 and the highest value is 28.79% Vietnam Technological and Commercial Joint Stock Bank (Techcombank) in 2011.

Net interest margin (NIM) has an average value of 3.06% with a standard deviation of 1.25%, with a minimum value of 0.5% of SaiGon Joint Stock Commercial Bank (SCB) in

2017 and the highest value is 8.13% of Vietnam Prosperity Joint Stock Commercial Bank (VPBank) in 2018.

Economic Growth (GDP) has a mean of 6.35%, with a standard deviation of 0.6% The minimum value is 5.25% of Vietnam Export Import Commercial Joint Stock Bank (Eximbank) in 2012 and the maximum value is 7.08% of Vietnam Export Import Commercial Joint Stock Bank (Eximbank) in 2018.

Bad dept (LLR) has a mean of 1.31%, with a standard deviation of 0.54% The minimum value is 0 Bao Viet Joint Stock Commercial Bank (BaoViet Bank) in 2010 and the maximum value is 5.57% of Nam A Commercial Joint Stock Bank (Nam A Bank) in 2014.

Bank size (SIZE) is measured by the natural logarithm of total assets with an average value of 8.029 with a standard deviation of 0.466 In which, the smallest value equal to 7.121 belongs to Bao Viet Joint Stock Commercial Bank (BaoViet Bank) in 2011, the maximum value is 9.118 of Joint Stock Commercial Bank for Investment and Development of Vietnam (BIDV) in 2018.

Equity capital (CAP) has a mean of 0.14, with a standard deviation of 0.186 The minimum value is 0.023 of Bac A Commercial Joint Stock Bank (BaoViet Bank) in 2012 and the highest value 1.052 of Vietnam Prosperity Joint Stock Commercial Bank (VPBank) in 2010.

Loan to deposit ratio (LDR) have a mean of 0.854, with a standard deviation of 0.172. The minimum value is 0.363 of Vietnam Maritime Commercial Joint Stock Bank

(MSB) in 2014 and the maximum value is 1.396 of Orient Commercial Joint Stock Bank (OCB) in 2011.

Loan-to-total assets ratio (TLA) have a mean of 53.64%, with a standard deviation of17.27% The minimum value is 22.01 % of Vietnam Maritime Commercial Joint Stock Bank

(MSB) in 2014 and the maximum value is 76.53% of Vietnam Joint Stock Commercial Bank (Vietbank) in 2017.

Industry growth (AGDP) have a mean of 5.29%, with a standard deviation of 7.55% The minimum value is 0.36% of SaiGon Bank for Industry and Trade (SAIGONBANK) in 2018 and the maximum value is 41.7% of Vietnam Joint Stock Commercial Bank for Industry and Trade (Viettin Bank) in 2017.

Industry concentration (CR3) have a mean of 14.84%, with a standard deviation of 0.52% The minimum value is 13.57% of An Binh Commercial Joint Stock Bank (ABB) in

2012 and the maximum value is 15.32% of An Binh Commercial Joint Stock Bank (ABB) in 2013.

Inflation rate (INF) has a mean of 5.61%, with a standard deviation of 4.73% The minimum value is 0.63% of Asia Commercial Joint Stock Bank (ACB) in 2015 and the maximum value is 18.68% Vietnam Export Import Commercial Joint Stock Bank (Eximbank) in 2011.

Money supply (M2) has a mean of 15.01%, with a standard deviation of 6.45% The minimum value is 4.4% of Ho Chi Minh City Development Joint Stock Commercial Bank(HDBank) in 2013 and the maximum value is 33% of Viet nam Export Import CommercialJoint Stock Bank (Eximbank) in 2010.

CORRELATION ANALYSIS OF VARIABLES

Table 4.2 Correlation between FGAP and independent variables

H NIM H GDP H INF H SIZE ' CAP

The starred coefficient estimates are significant at 1%(***), 5%(**), 10%(*)

Source: Calculation results from Stata software

Table 4.2 shows that the correlation coefficients for the independent variables ROE, GDP,SIZE, LDR, TLA, and AGDP are 0.1861, 0.2368,0.3047, 0.3125, 0.905 and 0.3577 respectively, at the 1% significance level NIM and CR3 has a correlation coefficient of 0.1582 and 0.148 at the 5% significance level, and LLR has a correlation coefficient of 0.1111 at the10% significance level All of the variables listed above have a positive relationship with the dependent variable FGAP, INF and CAP have correlation coefficients of -0.2297 and -0.1295,indicating a negative correlation with FGAP at the 1% significance level.

MULTICOLLINEARITY TEST

Multicollinearity is simply a phenomenon caused by a strong correlation relationship between independent variables in a linear regression model The study tested the hypothesis that there was no multicollinearity phenomenon by using the VIF criterion with the results presented in the following table:

Source: Calculation results from Stata software

In this research, multi-collinearity test will use VIF index, if VIF is less than 10, there will be no multi-collinearity; otherwise, variables with VIF > 10 should be removed Because theVIF of all independent variables is less than 4, the model's multicollinearity is considered insignificant.

ESTIMATED THE POOLED OLS, FEM, REM MODELS

In this section, the author employs three methods to quantify the influence of the model, including Pooled-OLS, REM, and FEM regression, and then chooses the best appropriate model to continue the defect tests The author will next calculate the model's effect level, significance level for each coefficient, and level of explanation for the liquidity risk of 31 Vietnamese commercial banks.

Table 4.4 Estimated of Pooled OLS, FEM and REM

Source: Calculation results from Stata software.

Table 4.5 Check for suitable model selection

FEM and REM OLS and REM

There is no between difference between different subject or time points

There is no correlation between the characteristic error between the subject and the explanatory variables

The error of the estimate does not include the deviations between subjects.

Source: Calculation results from Stata software.

SELECTION TEST OF 3 MODELS POOLED OLS AND FEM

❖ Check the fit between Pooled OLS model and FEM

The Wald F-test is used by the author to retest the concordance between the Pooled Ols model (classical linear regression model) and the FEM (fixed-effects regression model) at the significance level, with the hypothesis H0: The Pooled OLS model is a better fit The following are the outcomes of the model:

Model 1, model 2, and model 3 all produce the same results:

Because Pro>chi2 = 0.0000 < 0.05, we reject hypothesis H0 and accept hypothesis H1, indicating that the FEM model is more appropriate.

❖ Check the fit between FEM and REM models

Since FEM is the more suitable model compared to POLS, the Hausman test must be carried on to make a choice between FEM and REM The hypotheses of this test are: H0: There is no correlation between εit and independent variablesit and independent variables

H1: There is a correlation between εit and independent variablesit and independent variables.

Accordingly, the estimated result shows that Prob > chi2 = 0.1686 (p-value > 0.05). Therefore, H0 is accepted and REM is the more preferred model between these two.

❖ Testing the fit between Pooled OLS and REM models

Next, to choose between Pooled OLS and REM models, the research uses Breusch Pagan test with hypothesis H0: Pooled Ols model is more suitable The model gives the following results:

The results of model 1, model 2, and model 3 give the same results:

The results show that the coefficient Pro>chibar2 = 0.0000 < 0.05 should reject the hypothesis H0, accept the hypothesis H1 is to use the REM model.

The model defect tests are used to improve the reliability and relevance of the research results Tests on three common flaws in quantitative research, namely multicollinearity, variable variance, and autocorrelation As a result, the topic has confirmed that the phenomenon of multicollinearity does not exist in Table 4.3 using

5 1 the variance exaggeration factor - VIF test The topic continues in this section with the Modified Wald test to check variable variance and the Breusch - Godfrey test to detect autocorrelation.

The phenomenon of variance changes can affect the effectiveness of the model estimation, thereby affecting the reliability of the coefficient test The previous group tested the variance with the method of Breusch and Pagan (1979) with the hypothesis: H0: The model does not have variable covariance.

H1: The model has variable covariance.

Test Breush-Pagan Lagrangian multiplier test Chibar2 Prob > chibar2

From the results, it is shown that P, value = 0.0000 < 0.05, rejecting H0 at 5% significance level, there is a phenomenon of variance in the research model.

The phenomenon of variance changes can affect the effectiveness of the model estimation, lose the reliability of the coefficient test The previous group tested the variance by Wooldridge method (2002) with the hypothesis:

H0: The model does not have a 1 correlation phenomenon

H1: The model has first-order correlation phenomenon

Thus, with the test results showing that Prob = 0.0000 < 0.05 (significant level of 5%), both models have autocorrelation of residuals, which means that the regression model is estimated by the regression model The REM method does not guarantee the hypothesis Therefore, the study will not use this model for discussion but will continue to perform error handling by the GLS (Generalized Least Squares) estimation method for random effects models to obtain an accurate model.

ESTIMATED THE FGLS

As a result of the fixed effects model's variable variance and autocorrelation, the study employs the feasible generalized least squares (FGLS) method to overcome these issues.

Table 4.6 FGLS model results FGAP Coef Std Err Z P>z [9CR35% Conf Interval]

Source: Calculation results from Stata software.

The z-value indicates that the variables NIM, GDP, INF, SIZE, CAP, LDR, TLA, LLR,

CR3 and M2 are statistically significant with adequate significance levels, with the exception of the AGDP and ROE variable This conclusion is comparable to the three models examined: Pooled OLS, FEM, and REM In summary, the study regressed the most generally used table data models, including Pooled OLS, FEM, and REM Then, to overcome the flaws in the REM model, apply the FGLS model and discover that it is appropriate because all of the independent variables affect the dependent variable.

Table 4.7 Regression results of 4 methods

Source: Calculation results from Stata software.

Notes : This table reports the comparison between regressed models for commercial banks in

Vietnam over the period 2009 to 2019 Number (1) is Pooled OLS, number (2) is FEM, number

(3) is REM, and number (4) is FGLS.

The regression results show that the group of internal factors, including CAP, LDR, SIZE, TLA, and LLR, affect the bank's liquidity risk The group of bank market factors

CR3 and the group of macro factors including INF, GDP, and M2, also play a role in

5 5 explaining the bank's liquidity risk.

RESULTS DISCUSSION

The author has already regressed three estimated models: Pooled OLS, FEM, and REM, using panel data from 31 Vietnamese commercial banks during a 10-year period from 2009 to

2019 REM has been demonstrated to be the best appropriate model through various tests. REM, on the other hand, has issues with autocorrelation and variable variance As a result, the FGLS model is regressed as a replacement for REM and used to describe the research findings on liquidity risk.

FGAP= -0.754+ 0.0623 NIM+ 0.046 SIZE+ 0.154CAP- 0.317LDR+ 1.66TLA- 2.055LLR+ 1.116CR3 +0.341INF- 1.189CR3GDP+ 0.138M2

+ Net interest margin (NIM) Net interest income on average profitable assets is a measure of a bank's interest margin (NIM) This study's findings are consistent with those of previous studies such as Moussa, M.A.B (2015), Vodova, Pavla (2011), and Tran Thi Thanh Dieu (2020) The same results show that if the interest margin increases by 1 unit, the liquidity risk increases by 0.622 units When the bank's NIM is higher, it shows that the bank is operating very well and earning interest income from loans as well as other sources of income derived from borrowing very effectively The bank's interest expense is well managed The bank's efficiency and liquidity risk increase because as NIM increases, most of the interest income and similar income comes from the bank's lending activities, so the bank's lending activity increases When credit risk occurs, the bank will lack a large source of liquidity to deal with the risk, increasing the bank's liquidity risk.

Figure 2 Relationship between FGAP and SIZE

In the absence of other factors, bank size has a positive correlation with the independent variable that is statistically significant at the 1% significance level This result is consistent with most of the points the author has collected for the positive impact of bank size on bank liquidity risk, such as: Ahmad, R et al (2008); Akhtar et al (2011) and Truong Quang Thong (2013). The positive effect shows that the more the bank expands, the higher the liquidity risk. Theoretically, the larger the total assets of the bank, the less liquidity risk However, the larger the bank, the more benefits the implicit guarantees and advantages, which can reduce the cost of raising capital and regulating capital That makes them bolder to invest in riskier assets like loans, there by increasing the funding gap In reality, in Vietnam, large-scale banks such as VCB, CTG, and BIDV take advantage of their large scale to reduce their reserves of liquid assets to invest in highly liquid assets, thus adjusting This may increase the liquidity risk for the bank If the bank size increases by 1 unit, the funding gap increases by 0.046 units.

+ Equity ratio (CAP) The ratio of equity to total assets has a positive correlation with the bank's liquidity risk at a 1% significance level This result is consistent with most of the points the author has collected, showing the positive impact of bank size on bank liquidity risk,such as: Moussa, M A B (2015); Cucinelli (2013); and Truong Quang Thong (2013) The results show that if the equity capital increases by 1 unit, the bank's gap will increase by 0.154

5 7 units When equity is increased to improve the bank's financial capacity but the growth rate of credit and other service revenues cannot keep up with the growth rate of equity, the return on investment will decrease equity As a result, managers are under increasing pressure.

Figure 3 Relationship between FGAP and LDR

The ratio of total loans to total short-term deposits has a negative correlation with the variable FGAP, which is statistically significant at 1% The LDR ratio results of this study are consistent with the research results of Samuel Siaw (2015), and Truong Quang

Thong (2013) This result shows that an increase of 1 unit in the loan-to-deposit ratio will reduce the bank's funding gap by 0.317 units.

If the LDR ratio is high, the bank will have high profitability, but there are also trade- offs,such as higher liquidity risk Credit is a bank's main earning asset, but it is very inflexible compared to other assets An increase in LDR is not necessarily a "warning" indicator of liquidity, but it will assess the level of risk management for banks On the contrary, if the LDR ratio is low, the bank's liquidity is good, it can grow freely, it is easy to decide to invest and lend, and when many customers withdraw their deposits at the same time, it is not difficult to

5 8 meet However, a low LDR does not mean a bank is safe, because safety is not only reflected in liquidity risk but also other types of risks such as credit quality and term risk.

Figure 4 Relationship between FGAP and TLA

The ratio of outstanding loans to total assets is a factor that considers the influence of lending activities on a bank's liquidity, also known as the loan ratio (TLA) The results of the group study on TLA have a positive correlation with a significance level of 1%.

The results of the group study on TLA are consistent with the research results of other authors such as: Truong Quang Thong (2013); Bonin & Associates (2008); and Mohamed (2015) This result shows that if the bank increases the lending ratio by 1 unit, it will increase the bank's liquidity risk to 1,660 units When the bank's lending activity increases, the bank will bear a huge credit risk and this will lead to an increasing liquidity risk for the bank.

Based on the chart, the liquidity risk of the whole system has tended to increase since 2013 as banks increased the proportion of loans to total assets It will be worrisome when the bank is only interested in lending a lot without paying attention to the balance of assets and liabilities, which can cause a liquidity shortage that increases the risk of liquidity However, if banks’ balance between deposits and loans in the short term, this problem will be solved.

Figure 5 Relationship between FGAP and LLR

The results of the credit risk provision ratio (LLR) have a negative relationship with the dependent variable FGAP with a significance level of 1% The results of the LLR ratio are consistent with other research results from: General Hua Chen (2009); Cucinelli (2013); Truong Quang Thong (2013); and Lucchetta (2007) This result shows that when the credit risk provision ratio increases by 1 unit, it will lead to a decrease in liquidity risk of 2,055 units If an increase in a bank's provision for credit losses indicates an increase in the bank's lending and credit exposure, the bank will invest less in assets, reducing liquidity This means that the bank's liquidity risk is increased.

The basic objective of making provision for credit risks is to avoid the risk of loss from possible financial crises in the future Provisioning is a method used by banks to recognize a loss compared to the original value of a loan This loss is included in deductible expenses, reducing profits and at the same time reducing liability its tax payment However, the level of provision for credit risks has tended to decrease since 2014, which has increased the funding gap, making the bank's liquidity more strained.

+ Industry concentration (CR3) The study's findings are consistent with those of two other authors: Samuel Siaw (2015) and Godferry Marozva (2016) Research results show that when industry concentration increases by 1 unit, liquidity risk increases by 1,164 units.

The proportion of assets of the three banks with the largest total assets compared to the remaining banks fluctuated When CR3 increases, it shows that competition in the banking

6 0 market is quite fierce Banks have gradually revealed inadequacies and weaknesses such as increasing bad debt ratios and inadequate liquidity, leading to increased liquidity risk.

+ Inflation Rate (INF) The inflation rate (INF) has a positive relationship with the dependent variable FGAP with a statistical significance of 1% The results of the inflation rate are consistent with the results of other authors such as: Moussa, M.A.B

CONCLUSIONS AND RECOMMENDATIONS

CONCLUSION

Research results from 31 commercial banks in Vietnam during the period of 2009–

2019 show that the bank's liquidity risk is affected by the following factors: Net interest margin (+), Bank size (+), Equity ratio (+), Industry concentration (+), Inflation rate (+), Money supply (+), and Total loans ratio (+) are all positive indicators.

Research results show that bank size has a positive relationship with liquidity risk; that is, not every increase in bank size will reduce liquidity risk Large assets need to focus Pay attention to investing in assets with high liquidity Avoid cases where you are only focusing on business investment to increase profits On the other hand, in the process of increasing the bank's assets, the bank may have to use loans, so the bank may also face the risk of payment due, so the business situation is not efficient In the case of a liquidity shortage, if banks hold highly liquid assets or have good liquidity, they will avoid financial instability Therefore, banks need to develop and comply with policies to ensure safety indicators in their operations; minimize and strictly control high-risk assets; maintain autonomy in holding highly liquid assets; and appropriately allocate assets.

The more the credit balance, the greater the credit risk, and the bank's liquidity risk is also highly affected in this instance The items composing assets on the State Bank's balance sheet are those that effect the change in money supply M2, with the outstanding credit balance of credit institutions lending to the economy being one of the most important factors affecting the money supply ratio.

Next, the ratio of equity to total assets has a positive relationship with the bank's liquidity risk When commercial banks increase equity to improve the bank's financial capacity to improve liquidity If sudden capital withdrawal needs arise but the growth rate of credit and other service revenues cannot keep up with the growth rate of equity, this will reduce the return on equity of the companies’ bank Therefore, the increase in owner's equity will create back pressure on the bank's management to find ways to increase profits through credit expansion, financial investment, etc This also contributes to the increased risk for the bank.

Liquidity risk is inversely proportional to industry concentration When CR3 rises, it indicates that the banking market is becoming increasingly competitive Inadequacies and weaknesses in banks have steadily surfaced, such as an increasing bad debt ratio and insufficient liquidity, resulting in greater liquidity risk.

The ratio of loans to total assets has a positive relationship with liquidity risk when banks conduct lending activities in an uncontrolled manner or due to lax due diligence, not following the process mainly in pursuit of growth goals bank credit Therefore, it is possible that the bank will be led to a situation of overdue debt Bad debt may increase if risks occur, reducing the bank's operational efficiency and profits That will force the bank to reduce cash reserves and liquid assets or borrow additional money in the money market to compensate for liquidity Therefore, when equity capital increases, credit growth also increases, which at the same time reduces the bank's liquidity.

For inflation growth, the State Bank and other macro-management agencies need to have appropriate policies to help achieve the desired level of inflation From there,banks make plans for re-establishing the scale, capital structure, etc., in line with lending activities according to the market's needs without affecting the bank's liquidity.Considering the inflation factor, when there is inflation in the economy, prices will escalate or the bank itself will have the ability to increase lending interest rates, so customers will limit borrowing to supplement capital, but use another channel, so credit growth can be beneficial The bank's liquidity can be preserved or reduced in risk.Liquidity risk is positively related to net interest income earned on successful assets When a bank's NIM is greater, it means the bank is doing well and earning interest income from loans as well as other kinds of income earned from borrowing, and the bank's interest expense is under control Because most of the interest and similar income originates from the bank's lending activities, the bank's efficiency and liquidity risk grow when NIM rises When credit risk arises, the bank will not have a substantial source of liquidity to deal with the risk Hence, the bank's liquidity risk will increase.

RESOLUTION

Collateral has a negative impact on credit risk Therefore, if the collateral increases then the credit risk decreases In corporate lending, collateral not only serves as a safety standard for loans, but it also serves as a credit risk screening tool for businesses. However, in order to promote the active role of collateral in limiting creditrisks, commercial banks in Vietnam need to make reasonable use of collateral tools when lending loans to enterprises.

Banks need to ensure sustainable credit growth, as shown by close supervision in the process of appraisal and credit issuance as well as post-credit supervision Banks need to avoid over-granting credit, lowering credit standards; It is necessary to develop and clearly define risk appetite, thereby proactively building a credit portfolio with the expected allocation rate and choosing a plan suitable to the bank's profit target and loss tolerance.

According to the results of the study, the size of the bank has a positive impact on the efficiency of banking operations; The larger the bank, the greater the credit risk. However,increasing the size of the bank allows the bank to diversify its financial activities and provide the scope of products and services,creating an expanding and competitive market Joint stock commercial banks can increase their size by increasing loan mobilization through methods: deposit mobilization, issuance of valuable papers, and other institutional credit loans.

Several experimental studies have shown that GDP growth has a positive impact on credit risk Banks in countries with more competitive banking sectors, where banking assets makeup a large share of GDP, often have smaller and inferior profit margins. Overall, the bank's high asset-to-GDP ratio implies that financial development plays an important role in the economy This relative importance may reflect higher demand for banking services Therefore, attract more potential competitors to the market As the market becomes more competitive, banks need to adopt different strategic moves to maintain their profits, leading to increased credit risk.

In certain periods, the State Bank of Vietnam needs to have appropriate

7 0 monetary policies to control low inflation, helping to reduce bad debts.

Maintain safe liquidity, strictly comply with the regulations of the State Bank on the ratio of loans to mobilized capital at an appropriate level A greater emphasis is on focused on mobilizing and growing corporate capital Diversify your loan products and avoid high-risk lending.

Firstly, the thesis' major goal is to investigate factors impacting bank liquidity in commercial banks, assess the interaction of these factors on liquidity, and provide legal solutions to increase bank liquidity for commercial banks in Vietnam Despite the fact that the study met its aims, research data and research techniques have several limitations due to time constraints: To begin with, research data is scarce Over the ten years from 2009 to 2019, the analytical data was collected only on a year-by-year basis from the financial statements of 31 commercial banks The study did not split the study time into stages so that the impact of the independent factors on the dependent variable could be

Second, the research has not handled the research objectives in a variety of ways from which to compare and identify more successful research methodologies, nor has it provided many good proposals and beneficial solutions to maximize bank profits.

Thirdly, we have tested and drawn some important results to propose some recommendations to improve the performance of the Vietnamese commercial banking system in the research model However, the regression model's results did not produce an R-squared value, which was supposed to indicate how well the model compared to the entire banking system.

To sum up, the author hopes to conduct further research in order to provide a more general measurement of liquidity risk across the entire Vietnamese banking system, as well as to develop a model with improved testing and identify additional factors affecting bank liquidity, in order to provide a useful reference for students' research and banks in developing policies to improve liquidity.

Based on the study's limitations, the author suggests that future research should focus on increasing the number of research observations, increasing the internal and macro variables, and using various research models and methods to assess the impact of determinants on liquidity risk in a more comprehensive manner.

The study uses FGLS and OLS, FEM, and REM models to examine the factors that cause liquidity risk for commercial banks in Vietnam The thesis made a number of recommendations based on the research findings, including diversifying investment portfolios to limit focus on lending activities; controlling bad debt situations; improving credit quality; and improving credit quality actively coordinate closely with the media management agency to avoid losing people's trust in the bank in the SBV in general and commercial banks in particular, thereby contributing to maintaining stability and ensuring sustainability Vietnam's economy In addition, the author also points out the limitations of the study and future research directions to gain a deeper understanding of liquidity risk.

1 Ngân hàng Nhà nước Việt Nam, Danh sách các TMCP trong nước ( Đến

31/12/2019 ), Availaible from < https://www.sbv.gov.vn/webcenter/portal/vi/menu/fm/htctctd/nh/nhtm/nhtmnn>

2 Đặng Văn Dân, 2015 Các nhân tố tác động đến rủi ro thanh khoản của ngân hàng thương mại tại Việt Nam Tạp chí Tài chính, số tháng 11-2015, p 60-64.

3 Trương Quang Thông (2013), “Các nhân tố tác động đến rủi ro thanh khoản của các ngân hàng thương mại Việt Nam”, Tạp chí Phát triển kinh tế, số 276, 50-62

4 Ths Nguyễn Thị Bích Thuận - Ths Phạm Ánh Tuyết (2021 ), “Nhân tố ảnh hưởng đến rủi ro thanh khoản tại các ngân hàng thương mại Việt Nam”,

Availaible from < https://tapchitaichinh.vn/ngan-hang/nhan-to-anh-huong-den- rui-ro-thanh-khoan-tai-cac-ngan-hang-thuong-mai-viet-nam-334883.html >

5 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, vol 23, p 32-49.

6 Nguyễn Thị Mỹ Linh (2016), “Các yếu tố tác động đến tỷ lệ thanh khoản tại các ngân hàng thương mại Việt Nam”, Tạp chí Ngân hàng, số 9/2016;

8 Nguyễn Phúc Quý Thạnh (2020), “Trạng thái thanh khoản và hiệu quả hoạtđộng tại các ngân hàng thương mại Việt Nam”, Luận án tiến sĩ tài chính ngân hàng,Trường đại học ngân hàng thành phố Hồ Chí Minh.

9 Phan Thị Mỹ Hạnh – Tống Lâm Vy (2019), “Các yếu tố ảnh hưởng đến rủi ro thanh khoản của hệ thống ngân hàng thương mại Việt Nam”, Tạp chí nghiên cứu tài chính- Marketing số 51,06/2019

10 Funding liquidity and bank lending: Evidence from Vietnam Business and Economic Horizons, 15(2), 205-218.

11.Bonfim, D., & Kim, M (2012) Liquidity risk in banking: is there herding

European Banking Center Discussion Paper, 24, 1-31.

12 Valla, N., Saes-Escorbiac, B., & Tiesset, M (2006) Bank liquidity and financial stability Banque de France Financial Stability Review, 9(1), 89-104.

13 Vodová, P (2011, October) Determinants of commercial bank’s liquidity in Slovakia In Lessons learned from the financial crisis Proceedings of 13th

International Conference on Finance and Banking (pp 740-747).

14 Mugenyah, L O (2015) Determinants of liquidity risk of commercial banks in Kenya Doctoral dissertation, University of Nairobi

15 Mazreku, I., Morina, F., Misiri, V., Spiteri, J V., & Grima, S (2019) Exploring the liquidity risk factors in the Balkan Region banking system

16 Moussa, M A B (2015) The determinants of bank liquidity: Case of Tunisia International Journal of Economics and Financial Issues, 5(1), 249.

17 Chowdhury, M., & Zaman, S (2018) Effect of liquidity risk on performance of Islamic banks in Bangladesh IOSR Journal of Economics and Finance, 9(4), 1- 09.

18 Arif, A., & Anees, A N (2012) Liquidity risk and performance of banking system Journal of Financial Regulation and Compliance.

19 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.

20 Ibrahim, S 2017 The Impacts of Liquidity on Profitability in Banking Sectors of Iraq International Journal of Finance & Banking Studies (2147-4486).

21.Ghenimi, A., Chaibi, H., & Omri, M A B (2020) Liquidity risk determinants: Islamic vs conventional banks International Journal of Law and Management.

22 Duttweiler, R (2009) Managing Liquidity in Banks: A Top Down Approach

A1 Panel data description xtset bank NĂM panel variable: bank (unbalanced) time variable: NĂM, 2010 to 2019, delta: 1 unit A2 Variables statistics

sum fgap roe nim gdp inf size cap ldr tla llr cr3 m2 agdp

Variable Obs Mean Std Dev Min Max fga p 260 3767331 186548 -.0242 71 ro 99 e 260 0858327 0725248 0014 28 ni 79 m 260 0305638 0124852 005 08 gd 13 p 260 0634758 005966 0525 0708 in f 260 056075 0472604 0063 18 siz 68 e 260 8.028627 4662183 7.1214 9.1183 ca p 260 1399269 1856802 0238 1.0526 ld r 260 8542488 1726617 3633 1.396 tl a 260 5364031 1253704 2201 76 ll 53 r

(obs&0) nim gdp inf size cap ldr tla llr cr3 m2 agdp fgap roe nim gdp inf size cap ldr tla llr cr3 fgap 1.0000 roe 0.1861 1.0000 nim 0.1582 0.2241 1.0000 gdp inf

xtreg fgap roe nim gdp inf size cap ldr tla llr cr3 m2 agdp,fe fgap Coef Std Err t P>|t| [95% Conf.

R-sq: within = 0.8730 between = 0.7483 overall = 0.7863 corr(u_i, Xb) = -0.2364

Number of groups = 31 min = 3 avg = 8.4 max = 10

07644364 04708098 7249938 (fraction of variance due to u_i)

xtreg fgap roe nim gdp inf size cap ldr tla llr cr3 m2 agdp,re

R-sq: Obs per group: within 0.8693 min = 3 betwee n = 0.8786 avg = overal 8.4 l = 0.8685 max =

Wald chi2(12) = 1651.61 corr(u_i, X) = 0 (assumed) Prob > chi2 =

0.0000 fga p Coef Std Err z P>|z| [95% Conf.

05105737 04708098 54045188 (fraction of variance due to u_i)

A7 Pooled -OLS, FEM, REM regression

Random-effects GLS regression Number of obs = 260

Group variable: bank Number of groups = 31

esttab pooled fem rem, r2 star(* 0.1 ** 0.05 *** 0.01)

_B) ) roe -.0630106 -.0779192 0149086 0207709 nim 3096086 5404742 -.2308656 0901269 gdp -.5063072 -1.035023 5287158 4102051 inf 1810815 2153196 -.034238 056339 size -.0169494 -.0042827 -.0126667 0280782 cap 4241065 1713333 2527732 1104738 ldr -.2862833 -.2875382 0012549 0109041 tla 1.634816 1.640832 -.0060166 0236995 llr -1.762238 -1.674626 -.0876116 160343 cr3 0829922 2787999 -.1958078 1109805 m2 0459994 0765929 -.0305934 0284541 agdp 2666482 1985867 0680615 15233 b = consistent under Ho and

B = eg inconsistent under Ha, efficient under

Ho; obtained from xtr eg Test: Ho: difference in coefficients not systematic chi2(12) = (b-B)'[(V_b-V_B)^(-1)](b-B)

Prob>chi2 = 0.1686 (V_b-V_B is not positive definite)

Variable VIF 1/VIF inf 2.22 0.449758 ldr 2.09 0.478919 tla 2.07 0.484055 roe 1.99 0.501412 cap 1.72 0.580531 nim 1.68 0.594535 cr3 1.61 0.622431 size 1.58 0.633174 gdp 1.48 0.675846 m2 1.21 0.826407 llr 1.16 0.860603 Mean VIF 1.71

Breusch and Pagan Lagrangian multiplier test for random effects fgap[bank,t] = Xb + u[bank] + e[bank,t]

Var sd = sqrt(Var) fgap 0348001 186548 e 0022166 047081 u 0026069 0510574

Test: Var(u) = 0 chibar2(01) = 211.55 Prob > chibar2 = 0.0000

xtserial fgap roe nim gdp inf size cap ldr tla llr cr3 m2 agdp

Wooldridge test for autocorrelation in panel data

xtgls fgap roe nim gdp inf size cap ldr tla llr cr3 m2 agdp, panel(h) corr(ar1) force Cross-sectional time-series FGLS regression

Correlation: generalized least squares heteroskedastic common AR(1) coefficient for all panel s (0.3954)

Number of obs = Number of groups

esttab pooled fem rem xtgls, r2 star(* 0.1 ** 0.05 *** 0.01)

B Collect this data from 31 commercial banks collected between 2009CR3 - 2019CR3 8

MÃ CK NĂM ROE NIM GDP INF SIZE CAP LDR TLA FGAP LLR CR3 M2 AGDP