Navigating banking transitions: insights into competition and stability from Vietnam’s experience

ABSTRACT

This article investigates the relationship between bank competition and stability within a transitioning economy, focusing on Vietnamese banks from 2007 to 2018. Set against the backdrop of a transformative banking landscape post the collapse of the socialist bloc, the article utilises the Lerner Index to measure bank competition. The Z-score is used as a measure of bank stability. Panel fixed effects and difference GMM estimations suggest that heightened competition correlates positively with bank stability. Notably, these dynamics hold true even after omitting state-owned banks, signifying that government-controlled banks do not exclusively drive the observed relationships. Policymakers, hence, face the crucial task of balancing liberalisation efforts with safeguarding bank stability and emphasising the need for robust risk management and strong corporate governance in transitioning economies dominated by public ownership. This article thus offers valuable insights for policy refinement in similarly poised nations.

KEYWORDS:

• Bank stability
• competition
• transition economies
• Vietnam

1. Introduction

The collapse of the socialist bloc at the end of the 1990s triggered a transformation of former centrally-planned economies into market-based ones. In response, the banking sectors evolved from a one-tier system, where central banks also acted as commercial enterprises, into a two-tier system with separate central and commercial banks (Grigorian & Manole, 2006). This transformation process was accompanied by liberalisation efforts aimed at enhancing efficiency and competition in the banking sector by enabling the participation of domestic private and foreign investors. As a result, state ownership of banks has been significantly reduced (Fang et al., 2014).

However, the transformation into market-based banking systems has diverged into two trends (P. T. Le et al., 2019). First, comprehensive transformation without any restrictions or limitations has been implemented in Europe’s post-communist economies. This approach established fully market-based banking systems in these countries, and as a consequence these systems are primarily controlled by the private sector (Psillaki & Mamatzakis, 2017). Second, a more limited transformation with certain constraints has occurred in China and Vietnam, characterised by the dominance of public ownership of their banking systems. The existence of dominant public ownership aims to ensure that the banking system remains under state control, meaning that banks are still being monitored in order to carry out specific policies, such as providing credit to state-owned enterprises (SOEs) or financially supporting remote areas and infrastructure projects (Stents, 2017).

Due to agency problems, the dominating rationale of public ownership can stifle competition in the banking sector when state-owned commercial banks (SOCBs) are considered to be less competitive and inefficient compared to private sector and foreign rivals (Porta et al., 2002). However, the removal of SOCBs through competition between different ownership types in China and Vietnam is impossible due to the crucial roles these banks play in supporting economic growth and implementing the government’s national development policies. Yet at the same time, these transitioning countries are pursuing global integration strategies to attract foreign investments and boost exports. Commitments to the World Trade Organization (WTO) require both countries to open their banking markets. Concerns over modern banking technology and high-quality management from foreign banks force domestic banks, including private and state-owned ones, to enhance competitiveness once a level playing field is established for all banks regardless of their ownership types.

From the perspective of international integration, SOCBs have to pursue two objectives. First, they must maintain their leading role in the banking sector as the major institutions responsible for both commercial and policy lending. Second, they need to improve their operations to enhance competitiveness against private and foreign rivals when the legal and institutional frameworks are equal for all bank types. To balance the two mentioned objectives, several measures have been implemented to increase competition between public banks, including privatisation, listing on the stock exchange, or permitting the participation of strategic foreign investors who can hold a minority of bank shares.

It is worth noting that stability in the banking system, rather than efficiency, is the top priority of the Chinese and Vietnamese governments (Stents, 2017). Measures to enhance competition that may lead to increased efficiency will not be considered if they jeopardise bank stability. While liberalisation can boost competition in the banking sector, stability remains a substantial concern for the relevant authorities. Consequently, policymakers tend to limit the scope and scale of liberalising measures, only extending them if there is no negative impact on bank stability. Policymakers must consider whether increasing competition as a result of liberal measures can make the banking system more stable. If the answer is ‘yes’, a higher level of liberalisation will be implemented to support both efficiency and competition. In contrast, if the answer is ‘no’, liberalisation will be restricted, and more stringent regulations will be imposed to ensure a more stable banking system.

The objective of this article is to examine the impact of bank competition on stability in the context of financial liberalisation in a transition economy where public ownership still dominates the banking sector. The Vietnamese banking sector has been chosen to investigate the relationship between bank competition and stability under the impact of liberalisation. The rationale behind this choice includes the following. Firstly, Vietnam’s banks are the main source of financing (Le et al., 2019), and stability in the banking sector is crucial for the country’s sustainable growth. Secondly, Vietnam chose a gradual approach to financial liberalisation to ensure the stability of the financial system, with the WTO entry in January 2007 marking a milestone in the liberalisation process. However, the relationship between competition gains as a result of this event and the stability of the banking system in the aftermath of the WTO entry has not been investigated. Thirdly, Vietnam is the most up-to-date country that has experienced different and distinct phases of banking evolution, including liberalisation, distress, and consolidation.

While there are several studies investigating bank stability in Vietnam, such as those by T. D. Le and McMillan (2020) on the interrelationship among stability, profitability, and loan growth, and by P. T. Le et al. (2019) on multimarket contacts’ impact on stability, none have explored the relationship between bank stability and competition within the context of financial liberalisation. Additionally, while factors like gender in bank governance and shadow banking have been examined (Hoang et al., 2021; Nguyen et al., 2023), the influence of ownership dynamics on bank stability and competition remains unexplored in the Vietnamese banking sector.

Our article aims to fill this critical gap by investigating the interplay between bank stability and competition among ongoing financial liberalisation efforts in Vietnam. Specifically, we seek to understand how competition, measured by the Lerner Index, influences bank stability, as measured by the Z-score. By addressing this gap, our research contributes to a deeper understanding of banking dynamics in Vietnam and sheds light on the implications of financial liberalisation for bank stability and competition.

The panel fixed effects and difference Generalized Method of Moments (GMM) estimators are applied for estimation, using data from Vietnamese banks for the years 2007 to 2018. The results show that there is a positive and significant association between competition and bank stability in Vietnam.

This article contributes to the literature in two ways. First, it is one of very few research articles that explore the competition-stability relationship in banking sector in transition economies. Second, the competition-stability relationship is considered in the context where public ownership still dominates the banking sector and liberal measures must be implemented if commitments to open the banking market to foreign investors are to be honoured.

The rest of this article is organised as follows. Section 2 describes the theoretical and empirical bases of the relationship between bank competition and stability. Section 3 describes the progress of competition in the Vietnamese banking sector. Section 4 explains the variables and estimation method. Section 5 describes the data. Section 6 presents the empirical findings, and Section 7 provides conclusions.

2. Theoretical and empirical backgrounds

The relevant theories provide contradictory predictions regarding the relationship between bank competition and stability. The ‘charter value’ view suggests there is a negative link between competition and stability, since a higher degree of competition can force bankers to take more excessive risks in pursuit of larger profits, leading to higher fragility and exposure to danger (Keeley, 1990). However, in a banking system with restricted entries and stringent regulations, competition is limited, and banks have fewer incentives to engage in risky activities. Additionally, in a more competitive environment, banks may face difficulties in seeking information from borrowers, reducing their incentives to properly screen borrowers and increasing the probability of fragility (Allen & Gale, 2004).

Contrary to the ‘competition-fragility’ view, the ‘competition-stability’ view, supported by Boyd and de Nicolo (2005), suggests that market power increases bank portfolio risks. Banks with more market power or market share tend to charge higher interest rates, leading borrowers to engage in riskier businesses with higher rates of return. This can increase non-performing loans and destabilise the banking sector. In contrast, a rise in competition can force banks to reduce their interest rates, attracting more potential clients and information, minimising the probability of engaging in moral hazard and adverse selection risks. With this in mind, the ‘Too Big To Fail’ view is a supplement to the ‘competition-stability’ hypothesis. When banks become larger and more influential with policymakers and society, their failures can threaten the stability of the financial system or even expose it to a crisis. Concerns from authorities about possible contagion and systemic distress may lead banks to expect a government bailout in the case of insolvency, encouraging more risk-taking (Mishkin, 1999). The effect of the ‘Too Big To Fail’ view can be compounded when depositors fail to properly scrutinize the banks’ operations, believing their deposits are automatically guaranteed by the government and therefore safe.

Finally, a third approach reconciles the two strands of the literature by theoretically and empirically demonstrating the existence of a U-shaped relationship between competition and risk (Jiménez et al., 2013; Liu et al., 2013; Martinez-miera & Repullo, 2010).

Numerous studies using cross-country data have explored the empirical evidence on the competition-stability and competition-fragility relationship in the banking sector.¹ Uhde and Heimeshoff (2009) employ aggregate balanced bank data from EU-25 between 1997 and 2005 and find that bank concentration compromises financial stability. Furthermore, this concentration-fragility effect is more severe in the Eastern European region, which is less developed. The competition-stability hypothesis is also supported by Clark, Mare, et al. (2018) and P. T. Li (2019). Clark, et al. (2018) utilise data from the Commonwealth of Independent States (CIS) between 2005 and 2013, and show a positive relationship between bank competition and stability. P. T. Li (2019) discovers that competition contributes to bank stability in 22 European and CIS transition countries from 1998 to 2016. However, Berger et al. (2017) analyse bank data from 23 developed countries between 1999 and 2005 and provide empirical support for the competition-fragility hypothesis, as banks with higher market power tended to have less overall risk. For developing countries during the same period, Ariss (2010) finds evidence advocating the competition-fragility hypothesis, stating that greater bank market power enhances bank stability and profit efficiency, although it also deteriorates cost efficiency. Literature based on single-country data also provides mixed results. In the case of Russian banks, Fungáčová and Weill (2013) find that a higher degree of bank competition is associated with an increase in bank failures. Other studies also support the ‘competition-fragility’ hypothesis in Turkey and the UK. Kasman and Kasman (2015) employ unbalanced panel data of Turkish banks between 2002 and 2012 to examine the competition – stability relationship in the context of financial liberalisation, privatisation, and technological changes. They find that increases in bank pricing power can improve bank stability. Additionally, de Ramon et al. (2018) show that competition deteriorates bank stability in the UK, but this effect varies between different levels of bank soundness.

Some studies are ambiguous in identifying the ‘competition-stability’ or ‘competition-fragility’ relationship. For example, Beck et al. (2013) attribute the causes determining the relationship between competition and stability to the regulatory – supervisory framework and the level of institutional and financial development in particular countries. Using the bank data of 68 countries, they document large cross-country variation in the relationship between bank competition and bank stability. In general, an increase in competition will have a larger impact on banks’ fragility in countries with stricter activity restrictions, lower systemic fragility, better developed stock exchanges, more generous deposit insurance, and more effective systems of credit information sharing. In the context of single-country studies, the relationship between competition and stability is heterogeneous in different banking groups. Competition positively impacts stability in mutual saving banks in South Korea (Jeon & Lim, 2013). However, the competition-stability relationship is non-linear in commercial banks, reflecting a trade-off between interest effect and risk-shifting effect.

Overall, the competition-stability relationship is complex and varies from country to country based on institutional quality, financial development, regulatory, and supervisory frameworks. Although numerous studies have been conducted on the competition-stability relationship in the banking sector, research focusing on transition economies where public banks still dominate the banking sector is scarce. This article aims to address this gap in the literature.

3. Competition in the Vietnamese banking sector

During the 1975–1986 period, the Vietnamese government controlled all financial transactions through the State Bank of Vietnam (SBV), which operated as both a central bank and a commercial bank. In 1986, Vietnam changed its banking sector from a one-tier system to a two-tier system, including the SBV and four state-owned commercial banks (SOCB) (Stewart et al., 2016). The participation of private investors was allowed from 1990,² and this significant change led to the establishment of domestic private banks (known as joint stock banks or JSB), joint venture banks (entities established between a SOCB and an overseas bank), and branches of foreign banks. Despite the private sector’s participation, the banking system remained underdeveloped due to the dominance and ineffectiveness of public banks (Leung, 2009).

SOCBs functioned as both commercial and policy banks. Since funding from the state budget to SOEs was eliminated and the stock market had not been established, the SOCBs played an important role in providing financial resources to SOEs. The interference from the central and local government in the lending activities of SOCBs led to adverse selection and moral hazard. Due to the underperformance and inefficiency of SOEs in their operations, the high concentration of SOCBs’ loans to these public companies resulted in a higher share of bad loans (Camen, 2006). In the context where SOEs dominated the economy, their mounting bad loans in SOCBs led to a banking crisis (Leung, 2009). JSBs were undercapitalised and unstable because they had weak financial capabilities and poor-quality management. Engaging in risky lending activities and operating in an unsound regulatory environment, JSBs were affected by influential shareholders who tended to expand credit to related parties (Kovsted et al., 2005).

To avoid the mistakes of East Asian countries which led to the regional financial crisis during 1997–1998, the SBV established a consistent and transparent legal and regulatory environment by promulgating two important laws (Kovsted et al., 2005). First, the Law on the SBV, promulgated in October 1998, defined the roles and functions of the SBV in implementing monetary policy as well as regulating credit institutions. Second, the Law on Credit Institutions, which also took effect at the same time, aimed at building a Vietnamese banking system with provisions on prudential operations and placing credit institutions on a legal basis to facilitate the conduct of monetary policy and stimulate economic growth. The two laws are believed to have set up a legal corridor for banks to operate and encourage competition in the banking system.

Furthermore, the SBV identified a roadmap to improve bank competitiveness prior to Vietnam’s entry into the WTO. Several policies were issued: SOCBs had to be privatised and listed on the stock market; rural banks could be transformed into urban banks, thus expanding their scope and scale; and domestic banks were required to increase their capital. As a condition for WTO membership, Vietnam had to open its banking market to foreign investors, making it possible for them to participate in domestic banks as minority shareholders or establish foreign banks (P. T. Le et al., 2019). Vietnam’s entry into the WTO in 2007 led to an acceleration of credit and assets in the banking sector. In 2001, banking credit accounted for only 39.3% of gross domestic product (GDP), but it rose to 99% in 2007 and then to 125% in 2010. While competition-enhancing measures accelerated the growth of the banking sector in general, the relative number of private banks tended to increase. The share of the credit market owned by JSBs rose from 26% in 2007 to 39% in 2012 (see Table 1).

Table 1. Vietnamese bank shares in different measures in selected years (2007, 2012 and 2018).

	Assets			Loans			Deposit
	2007	2012	2018	2007	2012	2018	2007	2012	2018
Private banks	0.27	0.42	0.44	0.26	0.39	0.40	0.33	0.47	0.41
Public banks	0.73	0.58	0.56	0.74	0.61	0.60	0.67	0.53	0.59
Total in Billion VND	848,436	1,385,301	8,616,769	663,324	1,094,045	5,728,467	943,752	1,456,162	6,333,223

Source: The State Bank of Vietnam (2007–2018)

However, rapid credit growth over an extended period under a weak regulatory and supervisory framework led to banking distress towards the end of the 2010s (see Table 1). A report by the World Bank (2014) declared that the rate of non-performing loans (NPLs) during 2010 − 2012 was between 10 to 12%. Restructuring measures were issued to consolidate the banking sector, including forcing weaker banks to merge with stronger ones, establishing the Vietnam Asset Management Corporation (VAMC), and nationalising insolvent banks. All banks had to increase their capital to be commensurate with their risks. Subsequently, the number of banks decreased, and the average bank size rose. Accordingly, the level of market power has risen since 2014 as shown in Table 2 and Figure 1. Market power is measured by the Lerner Index and will be discussed in detail in Section 4.

Figure 1. Lerner Index by ownership types in Vietnam during the 2007–2018 period.

The Vietnamese bank market’ competition is measured by Lerner Index. The index is measured from 0 to 1 with the higher value indicating a lower level of competition.

Table 2. Lerner Index of the Vietnamese banks from 2007 to 2018.

Year	Private banks	Public banks	Banking Industry
2007	0.3483	0.2726	0.3105
2008	0.2801	0.2809	0.2805
2009	0.3201	0.3130	0.3165
2010	0.3101	0.2820	0.2961
2011	0.2786	0.2792	0.2789
2012	0.2494	0.2411	0.2453
2013	0.2491	0.3173	0.2832
2014	0.2541	0.3111	0.2826
2015	0.2105	0.2440	0.2272
2016	0.2684	0.2738	0.2711
2017	0.2881	0.2767	0.2824
2018	0.3018	0.2885	0.2952

Source: Authors’ calculation.

The aftermath of WTO entry has witnessed fluctuations in bank market power, with both declines and increases as a result of liberalisation and bank restructuring efforts. Concurrently, instances of bank fragility have been observed, prompting the implementation of stabilising measures during this period. A critical research question arises: does increased competition, stemming from financial liberalisation, lead to bank fragility or stability? If competition, as a by-product of liberalisation, contributes to bank fragility, policymakers must re-evaluate their market-liberalising strategies. However, if competition makes a positive contribution to bank stability, then efforts to promote liberalising measures should be intensified. This article aims to address this crucial question.

4. Methodology

Competition measurement

There are two indicators that can be used to measure market power at the bank level, these being the Lerner Index and market share. According to Beck et al. (2013), the latter criterion is subject to measurement error that is similar to Tobin’s Q. On the other hand, market share and market concentration indicators require a definition of the geographical market, which is not the case for the Lerner Index. Furthermore, the Lerner Index can capture both pricing power on banks’ assets as well as on bank funding side. Hence, in this article, the Lerner Index is utilised to measure the magnitude of competition across different banks.

The Lerner Index is identified as the mark-up of price over marginal cost and is a ‘level’ indicator of the proportion by which price exceeds marginal cost (Berger et al., 2017). If we have calculated the price and marginal cost, the Lerner Index for each bank and each year can be identified as follows:

(1) $\mathit{Lerne}{r}_{\mathit{st}}=\frac{{\mathit{p}}_{\mathit{st}}-\mathit{m}{c}_{\mathit{st}}}{p_{\mathit{st}}}$(1)

where the price of the bank s at the year t ($p_{\mathit{st}}$) is the average price of bank production and measured by the ratio of total revenues to total assets (Carbo et al., 2009; Fungáčová et al., 2013). Following Berger et al. (2017), marginal cost ($\mathit{m}{c}_{\mathit{st}}$) is derived via estimating a translog cost function with one output (total cost) and three input prices (price of fund, price of labour and price of physical capital). We construct the cost function as below:

(2) $\ln \mathit{T}{C}_{\mathit{s},\mathit{t}}={\alpha }_0+{\alpha }_1\ln \mathit{T}{A}_{\mathit{s},\mathit{t}}+{\alpha }_2{\left(\mathit{lnT}{\mathit{A}}_{\mathit{s},\mathit{t}}\right)}^2+{\sum }_{j=1}^3{\beta }_j\mathit{ln}{w}_{s,t}^j+{\sum }_{j=1}^3{\sum }_{k=1}^3{\beta }_{\mathit{j},\mathit{k}}\mathit{ln}{w}_{s,t}^j\mathit{ln}{w}_{s,t}^k+{\sum }_{j=1}^3{\gamma }_j\mathit{ln}{w}_{s,t}^j\times \mathit{lnT}{A}_{\mathit{s},\mathit{t}}+{\epsilon }_{\mathit{s},\mathit{t}}$(2)

where $\mathit{T}{C}_{\mathit{s},\mathit{t}}$ presents the total cost (interest expenses, personnel and other operating costs), $\mathit{T}{A}_{\mathit{s},\mathit{t}}$ measures the total assets for the bank s at the time t. The three input prices $w_{s,t}^j$ capture the price of funds _ $w_{s,t}^1$ (interest expenses/total funding), the price of labour _ $w_{s,t}^2$ (personnel expenses/total assets), and the price of physical capital _ $w_{s,t}^3$ (other noninterest expenses/fixed assets). Under the restrictions on symmetry and linear homogeneity in input prices, marginal cost can be computed using the equation written below:

(3) $\mathit{M}{C}_{\mathit{s},\mathit{t}}=\frac{\mathit{T}{\mathit{C}}_{\mathit{s},\mathit{t}}}{\mathit{T}{A}_{\mathit{s},\mathit{t}}}\left({\hat{\alpha }}_1+2{\hat{\alpha }}_2\mathit{lnT}{A}_{\mathit{s},\mathit{t}}+{\sum }_{j=1}^3{\hat{\gamma }}_{\mathit{j}}\mathit{ln}{w}_{s,t}^j\right)$(3)

The range of Lerner Index is between 0 and 1. If the index value is closer to 0, the market is more competitive. In contrast, the market is more monopolistic or oligopolistic if the value is reaching to 1.

Fragility measurement

To measure bank fragility, we employ the Z-score. The Z-score measures the distance from insolvency and is the most common criterion of bank stability in the literature (World Bank, 2016; Zigraiova & Havranek, 2016). The Z-score is calculated as follows:

(4) $Z_{\mathit{s},\mathit{t}}=\frac{\left(\mathit{RO}{\mathit{A}}_{\mathit{s},\mathit{t}}+E_{\mathit{s},\mathit{t}}/E_{\mathit{s},\mathit{t}}\mathit{T}{A}_{\mathit{s},\mathit{t}}-0.5\mathit{muT}{A}_{\mathit{s},\mathit{t}}\right)}{{\sigma }_{\mathit{ROA}}}$(4)

where $\mathit{RO}{A}_{\mathit{s},\mathit{t}}$ is the return on assets, $E_{\mathit{s},\mathit{t}}/E_{\mathit{s},\mathit{t}}\mathit{T}{A}_{\mathit{s},\mathit{t}}-0.5\mathit{muT}{A}_{\mathit{s},\mathit{t}}$ denotes the equity to total assets ratio, and ${\sigma }_{\mathit{ROA}}$ stands for the standard deviation of return on assets. The Z-score is inversely related to the probability of a bank’s insolvency. A higher Z-score implies there is less probability of insolvency. Because a bank becomes insolvent when its asset value drops below its debt, the Z-score can be interpreted as the number of standard deviations that a bank’s return must fall below its expected value to wipe out all equity in the bank and render it insolvent (Boyd & Runkle, 1993). This article opts for the approach used by Beck et al. (2013), which consists of using a three-year rolling time window to compute the standard deviation of return on assets (ROA) rather than the full sample period, whereas the ROA and the equity to total assets ratio are contemporaneous. As argued by Beck et al. (2013), this approach has two main advantages. First, it avoids the variation in Z-scores within banks that is exclusively driven over time by variation in the levels of capital and profitability. Second, given the unbalanced nature of panel dataset, it avoids the computation of the denominator at different window lengths for different banks.

Empirical model

The baseline empirical model (5) is presented below:

(5) $\mathit{Z}-\mathit{scor}{e}_{\mathit{st}}={\beta }_1\mathit{Lerner}\mathit{\hspace{0.17em}}\mathit{Inde}{x}_{\mathit{st}}+{\beta }_2{X}_{\mathit{st}}+v_s+{\mu }_t+u_{\mathit{st}}$(5)

where Z-score is the dependent variable and stands for bank stability, Lerner Index is the main variable of interest and measures banking competition. Control variables (X) include loan-to-assets ratio, bank size and total cost to total assets ratio (Clark, Mare, et al., 2018; de Ramon et al., 2018; Fungáčová & Weill, 2013). The loan-to-asset ratio demonstrates how bank assets are structured, with which the higher value indicates the preference of traditional banking activities. The bank size is measured by the total assets while the overhead criterion is measured by the total cost to total assets ratio. v_s captures the bank-specific effects; µ_t takes into account the year effects; s = 1, … , n denotes the bank; t = 1, … , t denotes the time period; β₁ and β₂ are coefficients; and u is an error term. All variables are converted into logarithmic form for the empirical estimation.

5. Data

This research utilises the data of Vietnamese banks covering 12 years from 2007 to 2018 and this period witnessed the liberalisation of the country’s banking sector under the impact of WTO entry requirements and the restructuring process of the industry (P. T. Le et al., 2019). Bank-level data are collected from annual financial statements including balance sheets and income statements. These statements can be found and retrieved from banks’ websites. The balanced dataset includes 28 banks, of which there are four state-owned commercial banks and 24 joint stock banks (private banks).

To estimate Z-score, the data of total assets (presenting bank output), total cost, price of funds (interest expenses/total funding), price of labour (personnel expenses/total assets) and price of physical capital (other noninterest expenses/fixed assets) are collected. As can be seen in Table 3, the mean value of total assets during the 2007–2018 period is 119,901,351 million Vietnamese dong (VND). A higher value of Z-score indicates a better level of bank stability.

Table 3. Definitions of variables and descriptive statistics.

Variables	Definition	Mean	Standard Deviation	Max	Min
Z_score	The Z_score is an indicator of bank stability using bank-level data. This indicator measures the number of standard deviation by which returns would have to fall from the mean to wipe out all of a bank’s equity. Higher value of Z_score suggest there is less probability of insolvency.	123.75	429.92	846.05	28.94
Lerner Index	The Lerner Index is a kind of competition indicator at the bank level and is calculated as the proportion of excess of price over marginal cost. Higher value of Lerner Index indicates there is more market power or less competition in the banking system. The range of Lerner Index is between 0 and 1.	0.28	0.10	0.6148	0.08
Total assets	Total assets (in Billion VND) are measured in millions of Vietnamese dong, a commonly used proxy for the bank size.	119,901	219,326	1,313,038	3,023
Overhead	The ratio of operating expenses to total assets	0.08	0.03	0.25	0.006
Loans/assets ratio	The ratio between bank loans and assets refers to the preference of traditional banking activities, while higher level of this ratio means that the bank is lending more money.	0.70	0.13	0.9580	0.31

Bank competition is measured by the Lerner index, which indicates a higher value in an environment with lower levels of competition. Control variables include: (1) loans to assets ratio, a measure of portfolio mix identifying the proportion of assets penetrated by traditional activities; (2) bank size, this measures the scale effects on bank stability; and (3) Overhead, the ratio of operating expenses to total assets identifies how cost saving impact bank stability. The descriptions of control variables are summarised in Table 3.

6. Empirical analysis

Bank competition and stability

We apply a panel fixed effects model for estimations that accounts for the unobserved factors of banks and years. The bank fixed effects capture all time-invariant differences across banks, such as management style and organisational culture. The year fixed effects control for factors that vary over time but are constant across Vietnamese banks, such as the Global Financial Crisis (2008–2009) and 2010s oil prices shocks.

Table 4 presents the results of panel fixed effect estimations. Columns (1) and (2) show a positive relationship between competition and stability in Vietnam’s banking sector for the whole 2007–2018 period where Z-score is presented for bank stability. The coefficient of Lerner Index is negative and significant suggesting that 1% increase in Lerner Index leads to −0.64% rises in Z-score. Note that higher value of Lerner Index indicates there is more market power or less competition in the banking system. This positive relationship still remains even we exclude the public banks from our sample (see column 2 of Table 4).³ While this result supports the competition-stability hypothesis, it is in line with other studies such as those conducted by Clark, Mare, et al. (2018), Goetz (2018), and Ibrahim et al. (2019). The positive link between bank competition and stability in Vietnam can be explained by two points. First, following the work done by Caminal and Matutes (2002), more competition can alleviate agency problems, and banks have fewer incentives to take aggregate risks. On the other hand, less competition can result in less credit rationing, larger loans, and subsequently, a higher probability of failure. In the Vietnamese banking sector, where the system is dominated by the big four SOCBs, the principal-agency problem is persistent and significant. In the 2000s, a substantial portion of SOCBs’ lending was allocated to state-owned enterprises, known for their inefficiency in utilising capital resources. The outcome was mounting bad loans generated by these public entities. Hence, the expansion of competition and alleviation of agency problems in Vietnam’s banking sector, especially in SOCBs, can improve bank stability.

Table 4. Bank competition and stability: panel fixed effect regressions.

	Dependent variable: Log (Z-score)
	Full sample	Only private banks
	(1)	(2)
Log (Lerner Index)	−0.647*** (0.293)	−0.777** (0.291)
Log (Asset)	−0.171 (0.249)	−0.041 (0.233)
Log (Overhead)	0.366** (0.155)	0.438*** (0.142)
Log (Loan/Asset)	−1.072*** (0.367)	−1.192*** (0.407)
Bank fixed effect	Included	Included
Year fixed effect	Included	Included
Number of obs.	321	273
Within R-sq	0.26	0.28

Log (Z-score) is the logarithm form of Z score. Log (Lerner Index) is the logarithm form of Lerner index. Log (Asset) is the logarithm form of total assets. Log (Overhead) is the logarithm form of total operating expenses to total assets ratio. Log (Loan/Asset) is the logarithm form of loans to assets ratio. ** and *** represent significance at the 5% and 1% levels, respectively. Robust standard errors are in parentheses.

Second, the SBV has implemented a ‘no banks to be bankrupt’ policy aiming to consolidate Vietnamese depositors’ confidence in the banking system. They argue that, in the context of asymmetric information and an unclear business environment like Vietnam’s, the cost of bailing out insolvent banks is lower than the cost of losing depositors’ confidence in the banking sector, which could lead to a systematic bank run and a collapse of the entire system. This ‘too important to fail’ policy can intensify risk-taking incentives, thus resulting in an increase in bank fragility (Mishkin, 1999; Molyneux et al., 2014). According to this perspective, competition-enhancing policies can minimise the probability of bank distress by forcing banks to disclose their information, permitting the participation of foreign investors as strategic shareholders in domestic banks, and privatising public banks.

Robustness check: alternative estimation method⁴

The previous findings are based on static panel fixed effects estimation method. As a robustness check, we re-estimate the relationship between bank competition and banking stability in Vietnam using the difference GMM developed by Arellano and Bond (1991). The GMM is useful for our dataset because the dynamic panel model is designed for panels with a larger number of cross-sections (N = 28) and a shorter time series (T = 12). In addition, it is likely that there is persistence in the dynamics of banking stability, such that the previous level of stability wields an influence on the current level (Clark, Mare, et al., 2018). The other reason for using the GMM method for estimation is to minimise the potential problem of endogenous explanatory variables in Equation 5(5) $\mathit{Z}-\mathit{scor}{e}_{\mathit{st}}={\beta }_1\mathit{Lerner}\mathit{\hspace{0.17em}}\mathit{Inde}{x}_{\mathit{st}}+{\beta }_2{X}_{\mathit{st}}+v_s+{\mu }_t+u_{\mathit{st}}$(5) . In our article, for example, we assume that higher levels of banking competition affect banking stability; however, it has been shown that increases in instability can equally affect banking competition (e.g. Carlson et al., 2022).

The results of the difference GMM estimator are presented in Table 5. In both columns, Lerner Index maintains their expected negative signs and significance levels. In general, these results reaffirm our earlier findings obtained from the fixed effects regressions.

Table 5. Bank competition and stability: GMM regressions.

	Dependent variable: Log (Z-score)
	Full sample	Only private banks
	(1)	(2)
Log (Z-score (−1))	−.141 (.126)	−.117 (.130)
Log (Lerner Index)	−1.463*** (.529)	−1.527** (.604)
Log (Asset)	.076 (.132)	−.010 (.144)
Log (Overhead)	−.591* (.342)	−.549 (.357)
Log (Loan/Asset)	2.273* (1.179)	1.961 (1.137)
Number of obs.	261	221
Number of instruments	21	21
Hansen test of overid. Restrictions (p value)	0.273	0.487
Arellano-Bond test for AR(2) in first differences (p value)	0.204	0.070

Log (Z-score) is the logarithm form of Z score. Log (Lerner Index) is the logarithm form of Lerner index. Log (Asset) is the logarithm form of total assets. Log (Overhead) is the logarithm form of total operating expenses to total assets ratio. Log (Loan/Asset) is the logarithm form of loans to assets ratio. *, ** and *** represent significance at the 10%, 5% and 1% levels, respectively. Robust standard errors are in parentheses. We use 2 to 3 lags of explanatory variables as instruments in the GMM estimations.

At the bottom of Table 5, we show results of two diagnostic tests that are carried out on the GMM estimations. The Hansen test (Hansen, 1982) of over-identification restrictions indicate the specification is correct and the instruments are valid (the p-value of this test is not significant). Second, the Arellano-Bond test for second order correlation (AR(2)) in the first differenced residuals show that there is no serial correlation, as the p-value of this test is non-significant.

7. Conclusion

This article examines the impact of competition on bank stability using data from Vietnamese banks from 2007 to 2018. During this period, various liberal measures were implemented due to Vietnam’s commitments to the WTO, and the distress of the banking system led to a long-term restructuring process. Despite the fluctuations in competition during this period, a positive relationship between competition and stability is observed.

The findings from this article have several policy implications for the Vietnamese banking sector. First, additional liberal measures should be implemented to increase bank stability. For instance, the SBV could raise the maximum proportion of bank shares owned by foreign investors, which would potentially increase competition and improve the overall health of the banking sector. Moreover, policymakers may consider easing restrictions on entry for new domestic and foreign banks, which could further promote competition and diversification within the sector.

Second, the development of a strong corporate governance framework is crucial for promoting stability in the banking sector. Policymakers should focus on enhancing transparency, accountability, and internal controls within banks. This can be achieved by implementing clear and enforceable regulations regarding the roles and responsibilities of bank management and boards of directors, as well as by promoting effective communication between banks, regulators, and the public. Lastly, fostering a culture of financial literacy and consumer protection can contribute to the stability of the banking sector. By ensuring that consumers are well-informed about financial products and services, and have access to appropriate channels for resolving disputes and complaints, policymakers can help to reduce the risk of financial instability caused by widespread consumer dissatisfaction or mistrust. In conclusion, by implementing these policy recommendations, policymakers can work towards creating a more stable and competitive banking industry in Vietnam, which can better withstand future economic challenges and contribute to the nation’s long-term growth and development.

This article is subject to several limitations. Firstly, the small scale of state-owned banks impedes the analysis of the relationship between bank competition and stability within this group. Secondly, the association between bank stability and competition can be explored across various stages of the evolution of the Vietnamese banking system, which includes liberalisation, distress, and consolidation. Nonetheless, this approach would necessitate dividing the overall bank sample into numerous smaller sub-samples, potentially yielding unreliable regression results. Subsequent research could examine the correlation between bank stability and competition in environments where the banking sector is under administrative control by the government and public ownership of banks is prevalent, such as China, Russia, or other transitioning and emerging nations.

Disclosure statement

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

Additional information

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Notes

1. A summary of previous studies on bank competition and stability is provided in Appendix 1.

2. To reform the former centrally-planned economy to a market-based one, the Vietnamese government chose the financial sector as the one of the first sectors to liberalise. Consequently, both foreign and domestic private investors can attend the banking market (Kovsted et al., 2005).

3. We also checked if one-year lag of Lerner Index has any effect on Z-score. However, we did not find evidence of lag effect.

4. In addition, we applied Robust regression, and the results show that there is a negative relationship between Log (Lerner Index) and Log (Z-score).

Appendix 1.

A summary on previous studies on the bank competition - stability relation

Table

Authors	Context	Data	Measures of competition and stability	Regression Method	Findings
FungacovaFungacova & Weill (2009)	Institutional reform and persistent distress	Russian banks 2001–2007	Lerner index Z_scores	Logit estimation	Greater bank competition is detrimental for financial stability.
Uhde and Heimeshoff (2009)	Cross-border banking consolidation in Europe.	25 European countries over the period from 1997 to 2005.	Concentration ratio HHI Z_score	Random effects 2SLS instrumental variable estimation	National banking market concentration has a negative impact on European banks’ financial soundness.
Ariss (2010)	Bank competition in developing countries	60 developing countries, from 1999 to 2005.	Conventional and efficiency-adjusted Lerner index Z_score	OLS Tobit	An increase in the degree of market power leads to greater bank stability and enhances profit efficiency, despite significant cost efficiency losses.
Beck et al. (2013)	Bank competition and stability in different regions	17055 banks in different regions from 1994 to 2009	Conventional Lerner index Z_score Concentration ratio HHI	OLS Fixed effects	The extent of competition’s impact on bank fragility depends on the countries’ s institutional characters.
Jeon and Lim (2013)	Competition-stability relationship in different types of banks	Korean banks 1999–2011	Boone index Z-score HHI Concentration ratio	OLS Fixed effects Random effects	The relationship between bank stability and competition is non-linear in case of commercial banks while it is linear and positive with mutual savings banks.
Kasman and Kasman (2015)	Financial liberalisation, privatisation, economic integration and technological changes.	Unbalanced panel data 28 Turkish banks 2002–2012	Conventional Lerner Index Efficiency-adjusted Lerner Index Boone Indicator HHI Concentration ratio	Two-step system GMM	Increases in bank pricing power improve bank stability. Banks are more stable in less competitive and low concentrated environment.
de Ramon et al. (2018)	Different level of financial soundness in banks	UK banks 1994–2013	Boone indicator Lerner index Z-score HHI	Fixed effects	The negative link between bank competition and stability is recorded and depends on the underlying financial health of the institution.
Clark, Mare, et al. (2018)	Privatisation and liberalisation	the Commonwealth of Independent States (CIS) 2005–2013	Lerner index Z_Score	Dynamic GMM	Competition contributes to financial stability.
Li (2019)	Reform in the transition banking systems has been accelerated in the past two decades	22 European and CIS transition countries from 1998 to 2016	Z-score Lerner index H-Statistic developed by Panzar and Rosse	Fixed effects 2SLS with IV	The positive relationship between market power and bank fragility is recorded. The structural reform improves bank stability.