Foreign ownership and national governance quality affect liquidity risk – case in Vietnam

Abstract

The paper aim to examine the impact of foreign ownership and national governance quality on liquidity risk in Vietnam’s listed commercial banks from 2010 to 2021, covering the COVID-19 outbreak. Using the two-step generalized method of moments regression model (GMM), the findings indicate that foreign ownership deteriorates liquidity risk, while national governance quality has a positive effect on liquidity risk. In particular, the study highlights the positive impact of COVID-19 on liquidity risk because of the unpredicted demand for cash withdrawals by customers, which led to a shortage of cash holdings in commercial banks. Consistent with agency and neo-institutional theories, the findings are robust to a series of endogenous checks using an alternative regression method, such as pooled ordinary least squares, and the generalized method of the moments regression model.

Keywords:

• commercial bank
• COVD-19
• Foreign ownership
• Liquidity risk
• National governance quality
• Vietnam

JEL classification:

• G21
• G32

PUBLIC INTEREST STATEMENT

The Bank is also a particular business category in which shareholders invest in the banks to maximize profits or increase the earnings per share. The paper aim to examine the impact of foreign ownership and national governance quality on liquidity risk in Vietnam’s listed commercial banks from 2010 to 2021, covering the COVID-19 outbreak. Using the two-step generalized method of moments regression model (GMM), the findings indicate that foreign ownership deteriorates liquidity risk, while national governance quality has a positive effect on liquidity risk. In particular, the study highlights the positive impact of COVID-19 on liquidity risk because of the unpredicted demand for cash withdrawals by customers, which led to a shortage of cash holdings in commercial banks. Consistent with agency and neo-institutional theories, the findings are robust to a series of endogenous checks using an alternative regression method, such as pooled ordinary least squares, and the GMM regression model.

1. Introduction

The development of the commercial banking system has created many positive changes to the growth and development of the economy. In particular, commercial banks also take a main function of financial intermediaries, connecting relevant parties in the market (P. T. H. C. Nguyen, 2022). In the context of international integration, commercial banks must expand their operation scales, equip modern technology, and enhance banking risk management to improve performance. In detail,

According to Ross (2021), risk management is defined as processes that are integrated with specific areas or risks, development of a comprehensive plan, integrated plan, and conducted ongoing evaluation. In the bank sector, risk management relates to implementing operation and manual processes operated by employees who take responses for risk assessments for the banks’ operations and services (Bojinov, 2016). So, under risk management in the bank, the stakeholders can evaluate the performance and allocate their capital appropriately to invest. One of the priority strategies is to maintain and stabilize their substantial capital resource for their operations and development.

According to Hoang Ton Thanh and Tatay (2019), attracting foreign capital will create many opportunities to help commercial banks build more assets and plan business strategies with higher profitability in the future, especially meeting the requirements of increasing the capital of the State Bank according to the prescribed schedule and improving their capital adequacy ratio. Furthermore, Detragiache and Gupta (2004) argue that foreign banks, which have substantial financial resources, have a stable influence on domestic banks during crises. Besides, foreign banks really take a priority of liquidity risk management.

Since the collapse of several banks worldwide, liquidity risk has become a serious problem that needs to monitor and control risks in the banking system. Besides, in the banking industry, the increased foreign ownership ratio is being considered and examined because the previous studies found different results on the relationship between foreign ownership and liquidity risk as (Al-Harbi, 2020; Dinger, 2009; T. V. T. Nguyen & Mai, 2018; Siaw, 2013; Sopan & Dutta, 2018; Vazquez & Federico, 2015). In addition, some studies found some drivers of liquidity risk, such as national and institution-specific characteristics (Bonner et al., 2015; Chen et al., 2018; Tammenga & Haarman, 2020).

Especially, in Vietnam, the attraction of capital contributions from foreign investors is bound by the government’s regulation through Decree No. 01/2014/ND-CP regulations on the percentage of foreign shareholdings. The foreign shareholding percentage of the Vietnamese credit institution charter capital, in particular, cannot be greater than 20%. The regulation of the maximum ownership ratio by the government of Vietnam is intended to limit the control of foreign investors on bank activities because the banking industry is one of the sensitive industries directly affecting national economic security. Eventhough there is the foreign ownership limitation in commercial banks, some studies conducted the effect of foreign ownership on liquidity risk in Vietnamese banking system. Specifically, P. T. Nguyen (2022), Vinh and Duc (2016) conclude that foreign banks can control liquidity risk better than domestic banks. Besides, foreign banks support Vietnamese commercial banks in liquidity shortage cases. T. V. T. Nguyen and Mai (2018) explore the relationship between the ownership structure and liquidity risk of Vietnamese commercial banks. From these studies, they provide policy implications to improve and manage liquidity risk at commercial banks. H. C. Nguyen (2022) investigates factors affecting the liquidity risks of joint-stock commercial banks in Vietnam from 2008 to 2018. The study’s findings show customer deposit to total assets ratio, loan to assets ratio, liquidity ratio, credit growth, external funding ratio, and loan loss provision significantly impact liquidity risk. Especially, the study also confirms that these factors play the most crucial role compared to macro-economic variables. Dinger (2009), Siaw (2013), Vazquez and Federico (2015) investigate the effect of foreign ownership on liquidity risk in the banking sector. Because, foreign-owned banks can manage liquidity better than domestic banks due to the opportunity to receive external funds from foreign investors (Siaw, 2013). While the empirical studies by Bonner et al. (2015), Chen et al. (2018), Sopan and Dutta (2018), Tammenga and Haarman (2020), Al‐Homaidi et al., (2019) explore the determinants of liquidity risk in banks which including the bank-specific characteristics and macro-economic factors. These authors also demonstrate that national factors are another factor that affects liquidity risk.

On the crisis period, Karim et al. (2021) find out a deterioration in the liquidity position of the listed banks after the emergence of the COVID-19 pandemic in Bangladesh. It means liquidity risk has increased after the pandemic. However, the findings by Just and Echaust (2020) point out that the COVID-19 has no impact on the market liquidity in the US. In addition, during the coronavirus spreading period, liquidity is shortage significantly in the emerging stock markets (Haroon & Rizvi, 2020), and bond markets (Gubareva, 2021).

Besides, some researchers proved the impacts of foreign ownership on liquidity risk, but they have not been concerned about COVID-19 and national governance quality factors in the model. However, during the pandemic, many governments adjusted the interest rate to a low level to adapt to the fact. This scenario reduces the profitability of banks. In reality, banks cannot completely predict the needs of customers to deposit or withdraw their capital, which leads to liquidity risk in banks. So, the COVID-19 effect is a crucial element that creates a risk to the bank liquidity, especially in the Vietnam context. Hence, the primary purpose of this paper is to examine the effect of foreign ownership and national governance quality on liquidity risk from 2010 to 2021, in Vietnamese commercial banks under agency theory, signalling theory and neo-institutional theory. To achieve the research goals, the following question needs to be clarified: “To what extent, do foreign ownership and national governance quality affect the liquidity risk at commercial banks in Vietnam from 2010 to 2021?”. From the foreign ownership and liquidity risk in commercial banks’ literature review, the author will discuss them briefly to build a foundation model. The next section presents the research hypotheses. To solve the endogeneity, the author uses the GMM model to find out the results which are interpreted and discussed in the following section. Some limitations and future research suggestions are listed in the final part.

2. Literature review and hypothesis development

2.1. Literature review

The agency theory derives from separating ownership and management (Jensen & Meckling, 1976). As a result, to eliminate the conflict between two parties, the managers must align the shareholders’ different interests. Therefore, to prevent managers from violating ethics issues and complying with rules, enterprises, including commercial banks, must enforce internal control rigidly (Quoc Trung & Abdul Wahab, 2021).

From an agency theory perspective, the existence of foreign ownership increases the level of information asymmetry (Bousnina et al., 2022; Lesmond, 2005; Rhee & Wang, 2009). According to Ferreira and Matos (2008), foreign investors typically advocate for greater corporate governance more actively than local investors. Increasing foreign ownership leads to improving corporate governance, which impacts to risk-taking activities of managers (John et al., 2008). Thus, appropriate corporate governance are considered the useful mechanisms to reduce conflicts within the agency problem (Eisenhardt, 1989). In addition, Fuentelsaz et al. (2000) argues that the risks can increase if the principal and agent have a different attitude toward the risks because of asymmetric information. A good corporate governance also contributes to reducing information asymmetry between management and company stakeholders (Pernamasari, 2018; Shubita & Shubita, 2019; Utami et al., 2020). So, the company with high liquidity ratio disclosures more information to create the signal for the investor’s decision-making (Hassan et al., 2006).

Under Neo-institutional theory, improving national governance quality can reduce information asymmetries in the bank system. So, national governance quality engages more effectively in risk management. Hence, banks can reduce the risks, including liquidity risk (Aguilera et al., 2008; Barakat & Hussainey, 2013; Beltratti & Stulz, 2012).

2.2. Liquidity risk and its determinants at commercial banks

2.2.1. Liquidity risk

According to the Basel Committee on Banking Supervision (2008), liquidity is the ability of a bank to finance an increase in assets and meet its due obligations without suffering unacceptable losses. According to Duttweiler (2011), liquidity shows the ability to convert from an asset that is converted into cash. Bank liquidity can be divided into two categories: natural liquidity and artificial liquidity. A bank’s assets create natural liquidity with a specified maturity. By being able to turn assets into cash before they mature, artificial liquidity is created.

Banks’ liquidity ensures enough cash reserves to meet all cash demands. From the perspective of commercial banks, liquidity is the bank’s ability can handle financial obligations fully and promptly, such as payment of the customer deposit, lending, and other financial transactions (V. T. Nguyen et al., 2012; T. D. Pham, 2018). In their activities, commercial banks must be able to meet all cash needs, including the demand for money transfers between other banks in the system; and banks, in general, must be able to supply any demand made to them to withdraw money from the system, whether overseas or for use in domestic circulation (Whittlesey, 1945). liquidity shows the bank can cover all payment obligations with a particular currency. Since it is done in cash, liquidity is only concerned with cash flows. In this case, banks’ inability to meet their payment obligations will lead to a lack of liquidity (B. H. Nguyen, 2015) and create liquidity risks in banking operations.

Based on the terminology of liquidity, liquidity risk is defined as “Liquidity risk represents the danger of not being able to fulfil payment obligations, whereby the failure to perform is followed by undesirable consequences” (Duttweiler, 2011). Besides, liquidity risk occurs when the maturities of short-term deposits and long-term loans do not match (Tammenga & Haarman, 2020). Therefore, the liquidity risk appears when the bank cannot handle its obligations for full and prompt payment (Jenkinson, 2008), especially customers withdraw their deposits before maturity. Hence, commercial banks can not promptly balance their capital to repay the depositors’ claims (Diamond & Rajan, 2005).

In the Vietnam context, liquidity risk occurs when the bank has inadequate money for immediate payment needs or sufficient money to cover payment at a high cost. This all begins with the bank’s insolvency to meet the requirements of payment contracts (Phan, 2009). Hence, a cash deficit can be understood as a liquidity risk, which occurs because the bank falls into a situation of lack or inability to meet regular financial obligations (B. H. Nguyen, 2015; Tran et al., 2019). In this paper, liquidity risk is calculated by the differential bank’s loans with customers’ deposits divided by total assets (Anthony & Millon, 2018; Chen et al., 2018).

2.2.2. Determinants of liquidity risk

2.2.2.1. Foreign ownership

There are some categories of owners: a government or state, a family, a non-financial company, a bank, and an institutional investor (Barry et al., 2011). Ownership structure is also classified as family control, government ownership, and foreign ownership (Ika & Widagdo, 2021). Where foreign ownership is defined as “the owner or control of a company is individuals who are not citizens of that country or by firms or institutions whose headquarters are outside that country. It can occur when a domestic property or shares is acquired or bought by a foreign individual” (Ika & Widagdo, 2021).

In Vietnam, foreign ownership is the percentage (%) of shares that total foreign investors are allowed to own. They are only allowed to buy the maximum number of issued shares according to the specified maximum percentage (B. H. Nguyen, 2015). The Vietnamese government limits the percentage of foreign ownership for the public companies in conditional business sectors in accordance with the new Investment Law (Investment Law No. 61/2020/QH14 passed by the National Assembly on 17 June 2020).

In summary, foreign ownership is the capital contribution of foreign shareholders in banks. Two issues are of major interest in these studies: the number of banks with foreign ownership and the percentage of shares held by foreign shareholders in a bank (Le, 2017). Foreign shareholders play an essential role in improving operational efficiency, system development, and profitability improvement of the banking system. Increasing the foreign ownership ratio in developing countries can assist banks in minimizing the risks (Micco et al., 2007). A study by (Vo & Mai, 2017) has demonstrated the influence of foreign ownership on the liquidity risk of Vietnamese commercial banks. The results show that the higher the foreign ownership, the lower the liquidity risk of commercial banks and vice versa.

In this paper, foreign ownership is the ratio of shares held by foreign shareholders to the total number of shares issued by commercial banks (ElBannan, 2015). From the above argument, the hypothesis is proposed as follows: foreign ownership negatively influences liquidity risk at commercial banks in Vietnam.

2.2.2.2. COVID-19

The global financial crisis of 2007–2008 has proven that liquidity is one of the crucial determinants of a bank’s survival (Rahman et al., 2020). In addition, macro-economic factors affect banking activities and economic risks (Bloom et al., 2018). About ten years later, over a million people have been infected by COVID-19 since its global outbreak began at the end of 2019, and more than 100,000 people have died as a result (De Vito & Gómez, 2020). Its consequence is that most countries are in lockdown, and consumer demand for products and services has plummeted. This pandemic has taken a heavy toll on the world economy and caused countries’ financial markets to again fall into severe disruption. Countries’ banking systems face liquidity risks because of the significant operational disruption for many companies.

In that context, some studies have analyzed the impact of macro-economic factors on liquidity risk of banks, especially COVID-19 (De Vito & Gómez, 2020; Dev & Sengupta, 2020; Karim et al., 2021; Li et al., 2020). They have shown the evidence about the liquidity of enterprises, including banks, is badly broken due to the COVID-19 pandemic.

The previous study has highlighted the concern that crises often expose banks to risks due to declining confidence levels in the financial sector (Cecchetti & Schoenholtz, 2020). Practically, from 2020, the effects of a pandemic outbreak on global economies will be adverse, which means a collapse in the activities of businesses, individuals, and households. As a result, there has been a decline in their income, net profits, ability to pay back their loans, and long-term and short-term interest deposits (M. U. Ali et al., 2020). Under further pressure as interest income continued to decline, the depositors might not favor banks that offer lower deposit rates. This would mean they withdraw money from their savings accounts and hold cash as a store of value rather than for making payments (M. U. Ali et al., 2020; Kosse & Szemere, 2021). Besides, during the pandemic, cash in circulation reached a decade high due to a surge in demand (M. U. Ali et al., 2020), suggesting that banks’ liquidity positions are threatened.

In this paper, COVID-19 is dummy variable, which is measured by 1 if the year had COVID-19, otherwise it receives the value of 0. Hypothesis 2: COVID-19 positively affects the liquidity risk of commercial banks in Vietnam.

2.2.2.3. National governance quality

National governance is measured by the Worldwide Governance Indicators (WGI) based on the study by (Kaufmann et al., 2011). WGI includes six indicators meant to capture different concepts, such as “Voice and Accountability; Political Stability and Absence of Violence/Terrorism; Government Effectiveness; Regulatory Quality; Rule of Law; Control of Corruption”. Furthermore, S. Ali et al. (2022) applied the Institutional Quality index created via PCA comprised of six world governance indicators as complementary assets at the country level that will moderate the relationship between corporate governance and stock liquidity. Studies by Krishnan and Teo (2012), Elamer et al. (2020) examine the relationships between national governance quality and risk management in banking system. Their findings confirm that national governance quality is positively related to risk management, that are demonstrated by (Alon & Dwyer, 2014; Barakat & Hussainey, 2013; Elamer et al., 2020; Pfeffer & Salancik, 2003). They agree the effective national governance will create motivations to provide greater management of risk at banks. As a result, they have more chance to access the resources for their operation. Besides, acording to Matemilola et al. (2019), legal enforcement measure the legal framework quality which is an important indicators to control liquidity risk in the developing countries. So, the national governance is measured by regulatory quality to examine its effect on liquidity risk in Vietnamese commercial banks.

Hypothesis 3

(H3): national governance positively affects liquidity risk at commercial banks in Vietnam.

3. Research methodology

3.1. Sample

The data is collected from the available resources from annual reports and websites of Vietnamese joint-stock and state-owned commercial banks from 2010 to 2021. The research time are chosen because the bank have recovered after the financial crisis 2008 – 2009.Thus, the sample size is 35 joint-stock commercial banks for the years 2010 to 2021, the research estimates 35 × 12 = 420 observations. However, after eliminating unobtained data, the research just collected 376 observations which are available and transparent information.

3.2. Research model

Based on the studies by S. Ali et al. (2022), De Vito and Gómez (2020), Dev and Sengupta (2020), Elamer et al. (2020), ElBannan (2015), Karim et al. (2021), Li et al. (2020), Vo and Mai (2016, 2017), the research model is proposed as follows:

$\mathit{\text{liq}}{r}_{i,t}={\alpha }_0+{\alpha }_1\mathit{\text{fow}}{n}_{i,t}+{\alpha }_2\mathit{\text{covi}}{d}_t+{\alpha }_3\mathit{\text{ng}}{o}_t+{\sum }_4^n{\alpha }_j\mathit{\text{control}}\mathrm{\_}\mathit{\text{factor}}{s}_{i,t}+\epsilon [1]$

The proposed model [1] is modified into model [2], which includes the instrument variable to solve the endogeneity (Bonfim & Kim, 2012). Table 1 provides a clearer explanation of all the variables.

$\mathit{\text{liq}}{r}_{\mathit{\text{it}}}={\alpha }_0+{\alpha }_1\mathit{\text{liq}}{r}_{i,t-1}+{\alpha }_2\mathit{\text{fow}}{n}_{i,t}+{\alpha }_3\mathit{\text{covi}}{d}_t+{\alpha }_4\mathit{\text{ng}}{o}_t+{\alpha }_5\mathit{\text{npl}}{r}_{i,t}+{\alpha }_6\mathit{\text{siz}}{e}_{i,t}+{\alpha }_7\mathit{\text{pr}}{o}_{i,t}+{\alpha }_8\mathit{\text{ca}}{p}_{i,t}+{\alpha }_9\mathit{\text{gd}}{p}_t+{\alpha }_{10}\mathit{\text{in}}{f}_t+\epsilon [2]$

Table 1. Variables in model

	Variables	Index	Formula/Source data	Related studies
Dependent variables
$\text{liquidityris}{\mathrm{k}}_{\mathrm{i},\mathrm{t}}$	liqr	$\mathit{\text{liq}}{r}_{i,t}$	$\mathit{\text{liqr}}=\frac{\mathit{\text{bank}}{\hspace{0.17em}}^{\mathrm{\prime }}\mathit{\text{sloans}}-\mathit{\text{customerdeposits}}}{\text{Totalassets}}$	Anthony and Millon (2018), Chen et al. (2018)
Independent variables and control variables
Independent variables	Foreign ownership	$\mathit{\text{fow}}{n}_{i,t}$	$\text{fown}=\frac{\mathit{\text{sharesheldbyforeignshareholders}}}{\mathit{\text{Totalnumberofsharesissuedbycommercialbanks}}}$	ElBannan (2015), Goodhart and Schoenmaker (2006), Schinasi and Teixeira (2006), Schoenmaker and Oosterloo (2007), Vo and Mai (2016), (2017)
	COVID-19	$\mathit{\text{AccessisdeniedAccessisdenied}}$	$\mathit{\text{AccessisdeniedAccessisdenied}}$ = 1 if year t has COVID-19 $\mathit{\text{AccessisdeniedAccessisdenied}}$ = 0 if year t does not have COVID-19	De Vito & Gómez, (2020), Dev and Sengupta (2020), Karim et al. (2021), Li et al. (2020)
	National governance quality	$\mathit{\text{ng}}{o}_{\mathrm{t}}$	Regulatory quality (WGI indicators)	S. Ali et al. (2022), Elamer et al. (2020), Kaufmann et al. (2011)
Control variables	non-performing loan ratio	$\mathit{\text{npl}}{r}_{i,t}$	$\text{nplr}=\frac{\text{non}-\text{performing loans}}{\text{Totalloans}}$	Cucinelli (2013), Dietrich and Wanzenried (2010), Mohammad et al. (2020), Vodova (2011), Vong and Chan (2009)
	bank size	$\mathit{\text{siz}}{e}_{i,t}$	$\text{Ln}(\text{totalassets})$	Delechat et al. (2012), El Khoury (2015), Pham Q. V. and Nguyen (), Tammenga and Haarman (2020), Vodova (2011)
	profitability	$\mathit{\text{pr}}{o}_{i,t}$	$\text{profitability}=\frac{\text{Earnings before interest}\mathrm{\&}\text{tax}}{\text{Onwe}{\mathrm{r}}^{\prime }\text{sEquity}}$	Ayoush et al. (2020), Delechat et al. (2012), Singh and Sharma (2016), Tammenga and Haarman (2020), Vodova (2011), Vodová (2013)
	capital to assets	$\mathit{\text{ca}}{p}_{i,t}$	$\mathit{\text{ca}}{p}_{i,t}=\frac{\mathit{\text{Owner}}{\hspace{0.17em}}^{\mathrm{\prime }}\mathit{\text{sequity}}}{\mathit{\text{totalassets}}}$	Chagwiza (2014), Chen et al. (2018), Vodova (2011), Vodová (2013)
	Inflation	$\text{in}{\mathrm{f}}_{\mathrm{t}}$	World Bank	Ahamed (2021), Chagwiza (2014), Chen et al. (2018), Vodova (2011)
	Economic growth	$\text{gd}{\mathrm{p}}_{\mathrm{t}}$		Ahamed (2021), Chagwiza (2014), Chen et al. (2018), Vodova (2011)
$\text{Instrument variabl}{\mathrm{e}}_{\text{it}}$	lqdr	$\mathit{\text{lqd}}{r}_{\mathrm{i},\mathrm{t}-1}$	$\mathit{\text{liq}}{r}_{i,t-1}=\frac{\mathit{\text{bank}}{\hspace{0.17em}}^{\mathrm{\prime }}\mathit{\text{sloan}}{s}_{i,t-1}-\mathit{\text{customerdeposit}}{s}_{i,t-1}}{\mathit{\text{Totalasset}}{s}_{i,t-1}}$	Bonfim and Kim (2012), H. C. Nguyen (2022)

Where: $\mathit{\text{liq}}{r}_{i,t}$: liquidity risk of bank i at time t.

$\mathit{\text{liq}}{r}_{i,t-1}$: lag of liquidity risk of bank i at time t-1.

$\mathit{\text{fow}}{n}_{i,t}$: foreign ownership ratio in bank i at time t.

$\mathit{\text{covi}}{d}_t$: COVID-19 occurred at time t.

$\mathit{\text{ng}}{o}_t$: national governance quality at time t.

$\mathit{\text{npl}}{r}_{i,t}$: non-performing loan ratio of bank i at time t.

$\mathit{\text{siz}}{e}_{i,t}$: the size of bank i at time t.

$\mathit{\text{pr}}{o}_{i,t}$: profitability of bank i at time t.

$\mathit{\text{ca}}{p}_{i,t}$: the capital to total assets ratio of bank i at time t.

$\mathit{\text{gd}}{p}_t$: gross domestic products at time t.

$\mathit{\text{in}}{f}_t$: inflation rate at time t.

3.3. Research method

The endogeneity occurs in the model, so the GMM method with instrument variables has been implemented to solve the issue and check the robustness of the findings (Bonfim & Kim, 2012; H. C. Nguyen, 2022). The GMM model (Arellano & Bond, 1991; Liu et al., 2018) is used significantly at 5 percent. To ensure the results’ reliability, Sargan, and Hansen tests are applied. Besides, according to Roodman (2009), the number of instruments should not be larger than the number of groups in order for adequate instrument variables to eliminate endogeneity.

Further, the authors conduct the Arellano and Bond tests to check whether the idiosyncratic error term is serially correlated. The AR(2) probability must be insignificant at 5% to confirm the absence of serial autocorrelation in the errors. As a result, we can conclude that the serial autocorrelation in the errors is absent in the model.

4. Research results and discussions

4.1. Research results

Table 2 presents basic characteristics of the collected data, including the mean, minimum and maximum values. The mean value of liquidity risk is −0.103. The minimum value is −0.574 which belongs to Global Petro Sole Member Limited Commercial Bank because the bank loans are lower than deposits from 2012 to 2015 to control the lowest liquidity risks. In contrast, the highest value of liquidity risk is 0.443 which belongs to Petro Vietnam Finance Joint Stock Corporation falling into the period of financial crisis from 2010–2011.

Table 2. Descriptive statistics

Variable	Obs	Mean	Min	Max
liqr	420	−0.103	−0.574	0.443
fown	420	0.079	0.000	0.300
covid	420	0.250	0.000	1.000
ngo	420	53.884	44.08	63.46
nplr	420	0.021	0.000	0.825
size	420	32.246	29.738	35.105
pro	420	0.143	−0.563	16.092
cap	420	0.087	−0.361	0.256
gdp	420	0.059	0.026	0.072
inf	420	0.055	0.006	0.187

In the ownership structure, there are two categories which involve foreign ownership and domestic ownership in the Vietnamese comercial banks. In detail, descriptive statistics results show that the mean value of foreign ownership factor is 0.080. The lowest bound belongs to Bac A Commercial Joint Stock Bank, Bao Viet Joint Stock Commercial Bank, Southeast Asia Commercial Joint Stock Bank, Ho Chi Minh City Development Joint Stock Commercial Bank, Kien Long Commercial Joint Stock Bank have no foreign ownership. While the highest proportion of foreign ownership is at An Binh Commercial Joint Stock Bank (ABB) with 30%.

For the firm specific factors, they involve non-performing loan ratio, profitability, capital, bank size. The mean for the control variables from highest to the lowest are 0.087, 0.143, 0.021, for capital, profitability and non-performaning loan ratio. Besides, bank size is measured by ln total assets with the range is from 29.738 to 35.105. Especially the average value of the bank size is 32.246, which shows these commercial banks try to increase the total assets to express the potential resources for competition.

Macro factors include GDP, inflation, covid and national government quality. Regarding COVID-19 factor, its maximum and minimum values are 1 and 0, respectively. Its value equals 1 when the year occurred COVID-19, and otherwise. For national governance quality factor, it is indexed by regulatory quality, that is one of the six world governance indicators (WGI). National governance quality takes the maximum and minimum values 63.46, and 44.08, respectively. Hence, its mean value is 53.88. It means the Vietnamese government has increased the protection of liquidity for banking system via the WGI indicators from 2015. The average value of GDP is 0.059 and inflation is 0.55. The government has issued policies to control the inflation from 2014 to 2019 while there are many policies to improve the GDP from 2014 to 2019 aim to economic development.

The following section shows the results of the examination of multicollinearity by using the correlation matrix and variance inflation factor (VIF). In addition, the Wooldridge test for autocorrelation in panel data and the Breusch-Pagan test for heteroskedasticity are also conducted to ensure the regression results are BLUE.

Table 3 confirms that multicollinearity does not exist in the model because two rules of thumb are satisfied. First, all VIF values are well below the required threshold value of 10 (Hair & Anderson, 1995; Montgomery et al., 2021). Second, there are no correlation coefficients greater than 0.8.

Table 3. Correlation matrix and VIF

	liqr	fown	covid	ngo	size	nplr	roe	cap	gdp	inf	VIF
liqr	1.000
fown	0.105	1.000									2.23
covid	0.019	−0.155	1.000								1.10
ngo	0.204	0.461	−0.105	1.000							1.21
size	−0.148	−0.628	0.157	0.314	1.000						2.01
nplr	0.075	0.042	−0.028	0.063	0.022	1.000					1.04
roe	0.328	0.039	0.198	−0.091	0.006	.004	1.000				1.12
cap	0.007	−0.284	0.004	0.268	0.217	.122	0.047	1.000			1.17
gdp	−0.081	−0.193	0.026	0.125	0.246	.075	−0.109	.032	1.000		1.13
inf	0.126	0.613	−0.180	−0.362	−0.594	.037	−0.169	−.316	−.016	1.000	2.18

With the OLS method, the error terms are correlated, and the variance is not constant, which causes the research results to be biased and inconsistent. Therefore, to overcome these phenomena, the study uses the GMM method (Bonfim & Kim, 2012; H. C. Nguyen, 2022). In this paper, the p-values of the Wooldridge test and the Breusch-Pagan/Cook-Weisberg test are less than 5%; thus, there is sufficient evidence to conclude that the model still has autocorrelation and heteroskedasticity.

The findings of Sargan and Hansen tests show in Table 4 have confirmed an overidentifying restriction problem related to the heterogeneity is reliable for both Sargan, and Hansen tests because their p-values are higher than 5%. Moreover, the number of instruments (28) is less than the number of groups (35), which is satisfied the condition by Almarzoqi et al. (2015); Roodman (2009). So, instrument variables are valid to solve endogeneity.

Table 4. Results of robustness check

	OLS		GMM
liqr	Coef.	P>z	Coef.	P>z
liqr L1.			0.624	0.000*
fown	−0.037	0.041*	−0.051	0.039*
covid	0.082	0.001*	0.094	0.000*
ngo	0.097	0.001*	0.195	0.000*
size	0.005	0.515	0.001	0.320
nplr	0.209	0.368	0.266	0.575
pro	0.218	0.018*	0.385	0.006*
cap	−0.776	0.040*	−0.851	0.030*
gdp	−0.273	0.309	−0.280	0.462
inf	0.154	0.386	0.180	0.448
Arellano–Bond test for AR(1) in first differences: z = −3.69 Pr > z = 0.000
Arellano–Bond test for AR(2) in first differences: z = 0.63 Pr > z = 0.532
Sargan test of overid. restrictions chi2(18) = 21.50 Prob > chi2 = 0.255
Hansen test of overid. restrictions: chi2(18) = 24.15 Prob > chi2 = 0.150
Number of instruments = 28
Number of groups = 35
variables are statistically significant at 5%.

Notes: $\mathit{\text{liqr}}$is liquidity risk of bank i at time t; liqr L1. is lag of liquidity risk of bank i at time t-1; $\mathit{\text{fown}}$ is foreign ownership ratio in bank i at time t; $\mathit{\text{covid}}$ is COVID-19 occurred at time t; $\mathit{\text{ngo}}$ is the national governance quality at time t; is the size of bank i at time t; $\mathit{\text{nplr}}$is non-performing loan ratio of bank i at time t; is profitability of bank i at time t; $\mathit{\text{cap}}$ is the capital to total assets ratio of bank i at time t; $\mathit{\text{gdp}}$ is gross domestic products at time t; $\mathit{\text{inf}}$ is inflation rate at time t.

Given that AR(2) is reported its p-value is 0.532, the null hypothesis of “no serial autocorrelation” is rejected at 5% significance level. The adoption of the Arellano-Bond test for AR(2) errors demonstrate there is no serial autocorrelation in the errors.

In summary, the results of the model estimation by the systematic GMM method are reliable and unbiased. At the same time, the model is meaningful in explaining the impact of bank-specific factors and macroeconomic factors on liquidity risk of commercial banks in Vietnam. In addition, to check the robustness of the research results, we perform the regression results of the OLS model is compared to the GMM model. For the purposes of the test, the authors replace and compare the regression results of two models with the explanatory variables in the research model. The findings confirm the reliability and certainty of alternative model specifications.

4.2. Discussions and implications

The regression results in Table 4 show five statistically significant factors at 5%, including the lagged liquidity risk, foreign ownership, COVID-19, national governance quality, profitability, and capital. Among these factors, the lagged liquidity risk and profitability positively affect the liquidity risk, while the following factors, such as foreign ownership, capital, and GDP, have a negative effect on the liquidity risk.

First, the findings show that foreign ownership affects the liquidity risk significantly and negatively. The estimated coefficient of foreign ownership is −0.051, which implies that a one-standard-deviation increase in the foreign ownership level will lead to a 0.051 decrease in liquidity risk. The findings are confirmed by Dinger (2009), Siaw (2013), Vo and Mai (2016, 2017). The results confirm that banks with higher foreign ownership percentages are better able to manage their liquidity risk. Taking advantage from capital sources, technology, management skills from the foreign investors, the Vietnamese banks can improve capital, human, technology and increased competition and risk management especially liquidity risk. When foreign investors act as owners (shareholders), they will require monitoring activities to limit problems related to liquidity risk when investing in projects. Under agency theory, the managers are authorized to use corporate resources, including cash for corporate investments. So, in many cases, the over usage of cash may damage company performance. To safeguard the liquidity of the company, some mechanisms are applied to monitor and control the managers’ activities (Wijaya & Noviany, 2017). In addition, good corporate cash management is a positive signal conveyed by managers to outside parties (investors) to tell whether the company has good liquidity management to foreign investors, which is based on signalling theory.

Second, COVID-19 factor has a positive effect on the liquidity risk because the coefficient is 0.094 (greater than 0). The findings are confirmed by De Vito and Gómez (2020), Dev and Sengupta (2020), Karim et al. (2021), Li et al. (2020). Put in another, liquidity risk has increased because of the consequences of the pandemic. As discussed above, at the end of 2019, COVID-19’s appearance turned the world upside down in many fields, especially banking system of many countries must face many challenges and risks. During the pandemic period, customers and related parties have faced many difficulties in balancing the cost and benefits in finance. Especially, their demand’s financial usage, and their private consumption have changed significantly, so the banks cannot completely control the needs of customers in deposit or withdraw capital. This leads to liquidity risk that is increasing in the pandemic, in the bank’s operation. To avoid liquidity risk, a minimum capital adequacy ratio is required by Basel and the Vietnam State Bank, which aims to ensure their activities and absorb losses. So, during the pandemic crisis, the government issues an initials policy such as Decree No 31/2022/ND-CP “Interest subsidies provided by state budget for loans of enterprises, cooperatives, and household businesses” that support the borrowers. Hence, these actions reduce the pressure about liquidity for the commercial banks when abnormal behavior of customers.

Third, it has been found that a positively and statistically significant 5% relations have been observed between national governance quality and liquidity risk because the coefficient greater than 0. The findings are similar to the studies by Bhati et al. (2015), Bhati and Zoysa (2012), Choon et al. (2013), Delechat et al. (2012), and Moussa (2015). The research results match neo-institutional theory that confirm national governance quality can mitigate information asymmetry and make the banks effort in risk management (Aguilera et al., 2008; Barakat & Hussainey, 2013; Beltratti & Stulz, 2012).

Forth, profitability has a positive effect on liquidity risk at Vietnamese commercial banks. The findings are consistent with the studies by Iqbal (2012), Roman and Şargu (2014), Widyarti et al. (2022). They agrue that the main activities in banks involve taking in funds, which are called deposits, lending to customers, and making other investments. Although bank loans and investments increase a bank’s profitability, these activities should still be considered risky assets, which leads to increased liquidity risk. So, the profits positively impact the liquidity risk.

Finally, the negative coefficient of capital demonstrates that capital has a negative impact on liquidity risk in the Vietnamese banking sector. According to Acosta Smith et al. (2019), Shah and Lahiani (2018), capital serves as a loss-absorbing buffer, making banks with higher capital ratios less susceptible to bank runs. The findings are supported by Berger and Bouwman (2009), Lei and Song (2013), Repullo (2004). They point out the negative effect of capital on liquidity risk for the different banks’ sizes.

5. Conclusions and limitations

The findings infer that factors are significant predictors of liquidity risk in Vietnamese banking system, such as the foreign ownership, COVID-19, and national governance quality. The regression and analysis are based on an unbalanced panel data set for Vietnamese commercial banks for the period 2010–2021. It is found that the liquidity risk is negatively correlated with foreign ownership and capital while COVID-19 and national governance quality affect the risk positively. The findings are confirmed based on agency theory, and neo-institutional theory. Besides, the robustness of the research results is confirmed by using OLS and GMM methods, hence the findings are reliable and un-biased.

Liquidity affects the financial health and development of commercial banks. A liquidity surplus will assist the bank in its operations, such as the safe usage and allocation of assets for investment activities and maintain a reasonable capital structure by increasing the ratio of customer deposits. On the contrary, liquidity shortage will make banks face liquidity risk. Commercial banks need to accurately calculate their liquidity needs to make reasonable reserves to prevent this problem. Second, they need to build a reasonable portfolio, with a reasonable proportion of investments in securities, capable of quickly converting to cash with minimal or zero conversion costs. Finally, regarding effective asset management, the banks must forecast the withdrawal needs of customers in each period so that they can proactively prepare capital for timely payment.

In addition, the study suggests the implications of reducing liquidity risk. First, there is a proposal of up to 49% foreign bank ownership ratio at Vietnamese credit institutions to supplement the bank’s operating capital. Second, regarding banking activities, the State needs to build an appropriate legal system to support credit institutions in managing and recovering debts more conveniently. Finally, the study has emphasized the importance of foreign investors in commercial banks taking advantage of investment capital from abroad to increase charter capital that aims to meet Basel’s regulations. Therefore, the research results are an antecedent for the government agencies to refer to because the government plans to amend and supplement a number of articles of Government Decree No. 01/2014/ND-CP on government affairs to allow foreign investors to buy shares of Vietnamese credit institutions.

Although the results of this study documented evidence that some significant predictors of liquidity risk are highlighted, some limitations need to be explicitly acknowledged. First, the COVID-19 factor could be measured by another proxy, such as number of people deaths, instead of dummy variable. Second, this study attempts to study bank liquidity risk that focuses only on state-owned and joint-stock commercial banks without considering full-foreign owned banks and joint-venture banks. Third, national governance quality could be measured by six world governance indicators as country-level complementary assets instead of regulatory quality. Finally, the study should be conducted broader by using cross section of countries. To sum up, further research needs to concern the different COVID-19 measurements and increase sample sizes that cover all bank types in Vietnam to create a comprehensive picture. In addition, it is considered a moderator or mediator for the effect of foreign ownership and national governance quality on liquidity risk.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Pham Tien Dat

Pham Tien Dat is Rector of University of Finance – Marketing, Vietnam. His fields of research and teaching are Banking, Finance, and Valuation - Real estate business.

Kim Quoc Trung Nguyen

Kim Quoc Trung Nguyen is currently a lecturer of the Faculty of Accounting - Auditing; the University of Finance – Marketing, Vietnam. He is interested in researching the banking sector and finance and accounting.