INTRODUCTION
Rationale
Throughout history, banks have played a crucial role by accepting deposits and providing financial assistance, but the emergence of credit unions and financial companies has led to significant changes in the financial mediation sector To remain competitive, banks have expanded their networks and diversified their services globally They have ventured into the insurance industry, often requiring borrowers to purchase life and loan insurance, which can be contentious for clients Additionally, banks have increased their offerings of unsecured loans to meet customer demand, although these loans come with higher interest rates and risks of delinquency Furthermore, banks have established a presence in the securities industry through subsidiaries, gaining customer trust, and have also expanded into investment funds and various financial support services.
In the current landscape, banks are diversifying their operations beyond traditional roles by acquiring subsidiaries in various industries, which complicates their operational structure This expansion may lead to increased risks, as issues in one subsidiary could trigger a domino effect across the banking system Additionally, the complexity of managing numerous subsidiaries can negatively impact profitability due to heightened operating costs and potential losses from unsuccessful ventures However, having a diverse range of subsidiaries can enhance a bank's risk allocation capabilities Despite increasing competition from non-banking financial entities in Vietnam, banks can maintain profitability by entering new markets and offering a broader array of financial services, benefiting from economies of scale through diversification Ultimately, the complexity introduced by diversification will influence banks' risk profiles, necessitating a careful evaluation of various factors to assess its overall impact.
Numerous studies globally have explored the relationship between bank complexity and risk; however, Vietnam has limited research on this emerging issue as its banks are only recently diversifying their operations This article aims to examine how the complexity of banks in Vietnam influences their associated risks, utilizing empirical evidence from local financial institutions.
This research report is structured into five distinct chapters The first chapter presents the historical context and poses the central inquiry of the investigation Chapter two reviews existing literature on the subject from both theoretical and empirical perspectives, establishing a foundation for the study framework and justifying the research Chapter three discusses the data collection methods and analytical approaches employed Chapter four presents the empirical results, answering the research questions outlined in the first chapter Finally, chapter five concludes the report.
Research objectives and questions
The objective of the study is to assess how bank complexity affects the risk of selected Vietnamese banks The research question is as follows:
How does bank complexity impact on risk for the case of Vietnamese banks?
Research significance
This study aims to explore the relationship between bank complexity and risk levels, focusing specifically on banks in Vietnam While numerous studies have investigated this topic in various countries and economic regions, there is a notable lack of research on this issue within the Vietnamese context Consequently, this article serves as a foundational reference for future research, encouraging further exploration and development in this important area.
LITERATURE REVIEW
Bank Complexity
The complexity of a bank is primarily determined by the level of its assets, as highlighted by Correa and Goldberg's (2021) research This suggests that a bank's complexity increases in direct relation to the number of assets it holds Furthermore, Marinelli et al contribute to this understanding by exploring additional factors influencing bank complexity.
In 2022, bank complexity was assessed through four key features: exposure to global commercial banking operations, geographical allocation, diversification of income streams, and integration of trading activities Stricter conditions can complicate banking operations, with the ownership structure of subsidiaries also influencing overall complexity Banks play a crucial role in circulating money within the economy, necessitating careful expansion in today's economic climate Over the years, the banking system has evolved in size and complexity, now including a wider range of financial intermediaries and non-financial enterprises Factors such as an expanding branch network and significant growth in non-banking businesses have reshaped the international presence of financial institutions.
In 2014, it was noted that financial institutions may own various entities such as securities firms, insurance companies, or fund management firms Martynova and Vogel (2022) further define bank complexity by the diversity of subsidiaries held by a bank, identifying five main categories: mutual and pension funds, insurance firms, banks, other financial subsidiaries, and non-financial subsidiaries.
To measure bank complexity, a complexity measure was constructed according to the following formula:
In this formula, Counti,j represents the number of affiliates of type j that are owned by bank i Ti denotes the total number of affiliate types associated with bank i, while Totalcounti indicates the overall number of affiliates owned by bank i.
The study highlights the advantage of utilizing widely available supervisory or regulatory data to develop measures that are beneficial for cross-country comparisons and broader conceptual discussions Specifically, it employs Herfindahl concentration indexes and assesses business complexity through a five-category classification to gauge the diversity of affiliates' operations.
In 2014, various financial entities, including banks, insurance companies, mutual and pension funds, as well as financial and non-financial subsidiaries, were analyzed The complexity of their output values ranges from 0, indicating simplicity, to 1, representing the highest level of complexity (Cetorelli & Goldberg, 2014).
The Herfindahl index formula is the preferred method for measuring bank complexity due to its widespread acceptance and comprehensive coverage of various methodologies used by previous researchers By counting the different types and numbers of subsidiaries, banks indicate diverse income sources and reflect their size through affiliate counts This formula has been extensively employed in numerous studies, including Martynova and Vogel (2022) and Cetorelli and Goldberg (2014), both of which assess bank complexity Furthermore, the Herfindahl concentration index is also utilized in Ho et al (2020) to explore the relationship between bank risk and complexity.
Bank Risks
There are many different kinds of risks that a bank could encounter There are primarily four types that are connected to the operation of a bank
Credit risk is a crucial concern for financial institutions, particularly banks, as it arises when borrowers may fail to meet their repayment obligations by the due date outlined in their loan agreements (Shahid et al., 2019) To mitigate potential losses from bad loans, banks are legally required to maintain loan loss reserves proportional to their total loan amounts This risk manifests when a debtor cannot fulfill their financial commitments, leading to economic losses for the bank due to declines in loan portfolio value and perceived loan quality (Shahid et al., 2019) Credit risk encompasses variations in actual and expected losses stemming from loans and reflects the possibility of a borrower's financial situation deteriorating Banks face credit risk not only from individual borrowers but also from transactions involving various financial instruments, including inter-bank dealings and derivatives (Shahid et al., 2019) Additionally, credit risk can be categorized based on the reasons for default, which may include national economic factors or challenges in executing transactions successfully.
Operational risk refers to the potential for loss due to failures, disruptions, or damages caused by people, systems, or processes (Lysiak et al., 2022) While operational risk is relatively low in straightforward enterprises like retail banking and wealth management, it is significantly higher in areas such as sales and trading This type of risk is unique, affecting all aspects of a financial institution's operations, yet it is notoriously difficult to identify and analyze (Habachi & Haddad, 2021) Key causes of operational losses include human error and internal fraud, with cybersecurity breaches posing a major threat As technology evolves, new operational hazards emerge, enabling hackers to steal client information and funds, potentially leading to significant financial losses and damage to a bank's reputation Such reputational harm can hinder a bank's ability to attract new customers and retain existing ones Consequently, regulators worldwide are working to enhance the legal and regulatory frameworks that govern commercial banks to better manage operational risks, in line with Basel Committee recommendations (Shaikh et al., 2021).
In the contemporary landscape, risk management is viewed through multiple perspectives, with many businesses dedicated to this field Variations in risk management practices can largely be linked to the differing stages of banking development across countries.
Market risk poses a significant threat to banks, primarily stemming from their activities in the capital markets, where the inherent unpredictability of financial markets—including equity, commodity, interest rates, and credit spreads—can lead to precarious situations (Lysiak et al., 2022) This risk is closely linked to currency risk, which refers to the potential adverse effects of fluctuating exchange rates on a bank's income and capital Banks continuously monitor their open currency positions to mitigate these risks (Lu & Boateng, 2018) Currency risk arises from short-term and long-term exchange rate changes driven by supply and demand dynamics in both national and international money markets (Saidane et al., 2021) It is classified as a speculative risk, influenced by the direction of exchange rates or a bank's net position in foreign currency For instance, a bank with a long currency position will profit from depreciation of its national currency, while an appreciation will lead to losses, and the opposite holds true for a short position (Kulinska-Sadocha et al., 2020).
Liquidity risk refers to a bank's ability to access funds to meet its financial obligations This risk is crucial as it impacts the bank's overall stability and operational efficiency.
In 2022, banks are mandated to ensure the timely withdrawal of customer deposits, which includes obligations related to cash deposits, short-term deposits, credit extensions, and off-balance sheet liabilities such as guarantees (Alawattegama, 2018) Failure to meet these obligations can lead to significant repercussions; for example, a one-day delay in cash delivery could erode customer trust, prompting a rush of withdrawals from other depositors This scenario not only disrupts efficient banking practices but also contributes to a contraction in the bank's money supply.
This paper focuses on credit risk, a critical concern for banks due to their primary activity of lending money The inability of borrowers to repay loans can significantly impact a bank's operations and potentially destabilize the entire economy, as evidenced by the 2008 financial crisis During this period, banks relaxed lending standards and increased credit availability, enabling many previously ineligible individuals to secure loans However, when the housing and stock markets collapsed, a vast number of borrowers defaulted on their loans, leading to a global financial crisis with lasting repercussions.
Measuring credit risk involves various methods, including credit scoring systems and international credit management techniques, as well as leveraging the expertise of senior credit department staff (Shahid et al., 2019) However, these measures do not guarantee the complete elimination of credit risk or the enhancement of financial performance and profitability Additionally, a universal credit risk model may not be effective across different regions due to demographic and cultural differences Therefore, it is essential to implement policies that consider the unique norms and values of specific cultures (Shahid et al., 2019) By adapting credit risk rules to meet regional requirements, commercial banks can effectively reduce credit risk, thereby enhancing their reputation and profitability within the banking sector.
This study utilizes the z-score as a measure of credit risk in banking, owing to its popularity and simplicity, as it relies solely on publicly available accounting data (Li et al., 2017) The z-score serves as a complementary tool to stock market-based approaches, making it a crucial risk metric in markets lacking accessible stock prices (Li et al., 2017; Lepetit & Strobel, 2015) Essentially, the z-score connects a bank's capital level to its profitability fluctuations, indicating how much profit variation can be supported by the bank's capital The change in returns is typically assessed through the standard deviation of return on assets (ROA), which forms the denominator of the z-score, while the numerator comprises the capital assets ratio plus ROA, reflecting the bank's capacity to sustain operations or adjust capital in the event of losses.
In this paper, the relationship between bank complexity and the risk of bankruptcy in Vietnam’s banks will be investigated by using the inverse coefficient of the Z-score
The z-score is a crucial measure for assessing bank risk, calculated by the formula that combines return-on-assets (ROA) and capital-assets-ratio (CAR) divided by the standard deviation of ROA This score indicates how many standard deviations a bank's ROA can drop below its expected value before reaching insolvency To align with existing literature and interpret increases in the z-score as heightened bank risk, the natural logarithm of the inverse of the z-score will be utilized (Laeven & Levine, 2009; Berger et al., 2017).
Impact of bank complexity on its risk
Numerous studies indicate a link between bank complexity and the risks they encounter However, the impact of this complexity on banking risk remains a topic of considerable debate, primarily due to the presence of certain paradoxes.
Recent theories suggest that complexity in banks can mitigate risks and enhance stability Ho et al (2020) highlight that multiple revenue streams can bolster institutional resilience against economic shocks, allowing banks to remain relevant in the financial system Diversification helps banks manage challenges related to bad debts and ensures operational funding Correa and Goldberg (2021) found that complex banks often derive a significant portion of their income from non-financial sources, which can lower overall risk through diversified activities (Goetz et al., 2016; Cetorelli et al., 2017) Laeven and Levine (2007) also noted that diversified banks benefit from a broader range of income sources Cetorelli and Goldberg (2016) demonstrated that complex institutions effectively manage liquidity through internal capital markets, reducing exposure to liquidity risks For instance, Vietnamese banks have diversified into non-traditional services to stabilize income, as noted by Boyd and Graham (1988), who argued that expanding operations can mitigate risks and performance fluctuations The COVID-19 pandemic illustrated this, as banks that invested in insurance profited from increased demand Additionally, the pandemic spurred a surge in new securities accounts, providing banks with capital to support customers through extended repayment periods and lower interest rates, ultimately reducing bad debts Keeton (1991) emphasized that strategic diversification can lower banking risk volatility, while DeYoung et al (2009) warned that excessive complexity might lead to "too-big-to-fail" scenarios Berger et al (2010) concluded that increased bank complexity enhances financial leverage and risk management, allowing for better income oversight and risk control across diverse operational fields.
However, at the bank level, complexity may raise the risk (Martynova & Vogel,
In 2022, the expansion of legal entities, broader commercial activities, and increased geographic presence of bank subsidiaries have heightened risks within banking institutions The ownership of multiple subsidiaries engaged in non-banking or non-financial activities complicates management and control for banks This diversification often necessitates hiring specialized personnel, which can create agency problems due to potential conflicts among different entities In Vietnam, the rising number of non-banking organizations is diminishing the significance of financial intermediaries in the economic landscape (Pham et al.).
In the evolving landscape of banking, institutions are expanding beyond traditional functions to maintain their relevance, which exposes them to increased risks from subsidiaries involved in non-banking activities (Kwan et al., 2019; Martynova & Vogel, 2022) The complexity of managing multiple subsidiaries complicates risk control, as banks must navigate diverse economic sectors while facing heightened operational challenges This diversification, while essential for sustaining their economic position, leads to greater interconnectedness and potential managerial failures, which can diminish strategic risk-taking benefits (Chernobai et al., 2021) Historical studies highlight the agency costs and monitoring challenges that arise from this complexity, as communication between parent companies and affiliates deteriorates, exacerbating management conflicts Consequently, these dynamics may incentivize subsidiaries to engage in riskier behaviors, resulting in suboptimal risk-taking and inefficient investments (Scharfstein & Stein, 2000).
Following the 2008 financial crisis and the collapse of Lehman Brothers, concerns about bank complexity have prompted experts and policymakers to implement new regulations aimed at improving the operations of large banks Vietnam's banking system is also affected, particularly with recent bank mergers raising questions about the effectiveness of private banks In response, many banks are establishing risk management departments to mitigate business risks and avoid bankruptcy Previous studies indicate that a bank's complexity can either heighten its risk exposure or facilitate risk distribution, with findings varying based on the research context This article will explore the interplay between bank complexity and risk within the Vietnamese banking market.
Research gap
Recent research on bank complexity and its associated risks has predominantly emerged from various countries, highlighting a relatively new challenge for Vietnam Over the past decade, Vietnam's financial system has undergone significant transformation, emphasizing growth in specific sectors and economic activities Investigating the effects of bank complexity on risk is crucial for Vietnamese banks, as it will aid in evaluating the implications of sector expansion and inform the development of effective growth strategies.
DATA AND METHODOLOGY
Model Methodology
The paper follows Cetorelli and Goldberg (2014) and Aldasoro et al (2021) to compute a Herfindahl-Hirschman index as a proxy for bank complexity:
The index Counti, j quantifies the number of affiliate types j owned by bank i, with Ti representing the total number of affiliate types for that bank Totalcounti indicates the overall number of affiliates under bank i's ownership A higher index value signifies greater complexity, while a uniformity in affiliate types results in a minimum value of 0 Conversely, a diversity of affiliate types yields a maximum index value of 1.
When calculating the bank complexity of each bank over the years, then it is necessary to average their values to find their complexity over the observed time period
As mentioned in the previous chapter on bank risk measures, in this study, the index will be calculated as follows
First, it is essential to compute standard deviation of banks' ROA over the years Then use the following formula to calculate the z-score
After that, inverse z - score to compute bank risk
Finally, calculate the average risk of each bank
In this study, the variables used in the bank specification group are presented as follows
The loans to assets ratio, calculated by dividing total loans outstanding by total assets, is indicative of a bank's credit risk Research by Parvin et al (2019) reveals a positive correlation between this ratio and credit risks, suggesting that a higher ratio indicates increased lending and reduced liquidity, resulting in a greater risk of default Furthermore, Bhowmik and Sarker (2021) highlight that a rapidly rising bank debt ratio correlates with a surge in bad loans and elevated bank risks Additional studies, including those by Baron & Xiong (2017), Jordà et al (2013), and Schularick & Taylor (2012), demonstrate that high debt adversely impacts financial structure and performance Consequently, this index is deemed appropriate for risk assessment in banking models.
The deposit-to-asset ratio, as outlined by Mwangi et al (2015), gauges how much of a bank's assets are financed through public deposits, which are considered a cost-effective funding source This reliance on deposits not only lowers operating costs but also enhances profitability and sustainability for banks Furthermore, this ratio significantly impacts a bank's risk profile, as noted by Ekinci & Poyraz (2019) In scenarios such as banking panics, a higher deposit-to-asset ratio enables banks to manage mass withdrawals more effectively Additionally, during periods of increased bad debt, this ratio provides crucial support, alleviating financial pressure Consequently, the deposit-to-asset ratio has been a focal point in various studies on bank risk, including those by Mwangi et al (2015) and Farkasdi et al.
(2021) or the paper by Ekinci & Poyraz (2019)
The cost to income ratio is a key financial metric that reflects a bank's operating expenses relative to its total revenue It is calculated by dividing the bank's expenses by its income According to Kingu et al (2018), an increase in bad loans due to adverse selection prompts banks to allocate more resources to managing and supervising these loans, ultimately raising operating costs and the cost to income ratio A higher ratio signals poor management in underwriting and loan portfolio control, indicating low cost efficiency and a correlation with an increasing bad debt ratio (Tripe).
1998) Therefore, it can be seen that this ratio also plays a role in influencing the bank's risk
Assets imply the size of bank, taken in logs (Altunbas et al., 2007; Laeven et al.,
A study by Laeven et al (2014) reveals that larger banks tend to create higher individual and systemic risks compared to smaller banks, particularly when they are undercapitalized or rely on unstable capital sources Additionally, as these larger banks engage more in market-based or complex organizational activities, their exposure to risks increases significantly.
In 2014, it was noted that banks with high complexity tend to possess a large number of assets, which can complicate management and control Consequently, it is essential to incorporate this variable into the model, as it significantly influences the bank's risk profile.
The Loan Loss Reserve (LLR) represents the provisions banks set aside for potential bad debts, specifically related to groups 3-4-5, while excluding group 2 debts This reserve, found on the income statement, is crucial for accounting for potential loan defaults and expenses, ensuring a clear picture of the bank's financial health (Cho & Chung, 2016) A significant provision for loan losses can greatly influence banking risk (Ozili & Outa, 2017), as a higher LLR reduces the threat posed by bad debts, enabling banks to better manage credit risk (Jasman & Murwaningsari, 2022) However, since LLR is classified as an expense, excessively high reserves can negatively impact the bank's profitability This study will focus on bank risk rotation issues, incorporating LLR as a key variable in the analysis.
Non-Performing Loans (NPL) are defined as the ratio of bad debts to a customer's outstanding loan balance, representing the total borrowed amount for which payments have not been made within a designated timeframe As highlighted by Foglia (2022), NPLs play a crucial role in assessing credit risk.
A bank's high ratio of non-performing loans (NPLs) to total loans significantly increases its risk of failure, highlighting the declining credit quality as a major contributor to financial system fragility and potential banking crises (Foglia, 2022) This aligns with the findings of Khan et al (2020), who emphasized that the level of bad debt directly influences bank performance, thereby affecting both financial stability and the broader economy Consequently, incorporating non-performing loans as a control variable in financial models is essential due to their substantial impact on banking operations.
The group of macro variables also contributes to the bank's credit risk In the study, two main indicators were selected which are GDP and Inflation
Indicators of the general health of an economy can be found in the country's macroeconomic policy, gross domestic product, inflation rate, and interest rate (Musau et al.,
Macroeconomic conditions significantly influence financial industry participation, with improved conditions expected to enhance financial inclusion Research by Musau et al (2018) highlights that Gross Domestic Product (GDP) reflects economic development and its impact on bank stability, indicating that higher GDP correlates with increased profitability and stability due to rising average incomes Additionally, Mpofu and Nikolaidou (2018) found that GDP growth effectively reduces banks' bad debt ratios, as a thriving economy fosters a favorable environment for borrowers, enhancing their loan repayment capabilities and minimizing credit risks Therefore, GDP is a crucial variable for analysis in the econometric model of this study.
Inflation (INF) is a crucial macroeconomic factor linked to bank risk, as it interacts with interest rates in a cyclical manner When the Federal Reserve lowers the prime interest rate, borrowing becomes more attractive, leading to increased money circulation and consumption However, this influx of low-cost money can devalue the national currency, potentially accelerating inflation Consequently, a decline in interest rates often results in rising inflation rates While inflation can diminish the real value of existing debt, it may also hinder borrowers' ability to repay loans if their incomes do not rise in line with increasing living costs Thus, understanding the interplay between inflation and interest rates is essential for assessing credit risk in banking.
The econometric model is stated as below
The model for assessing bank risk is represented by the equation: \(BankRisk_{it} = \beta_0 + \beta_1 BC_{i,t-1} + \beta_2 CONTROL_{i,t-1} + \phi_t + \omega_i + \epsilon_{it}\), where \(i\) denotes individual banks and \(t\) indicates the year The terms \(\phi_t\) and \(\omega_i\) account for year and bank-specific fixed effects, respectively The variable \(BankRisk_{it}\) measures the likelihood of bank default, derived from the natural logarithm of the inverse z-score for bank \(i\) in year \(t\) Additionally, \(BC_{it}\) reflects the complexity of bank \(i\) during year \(t\), while control variables are included to address potential omitted variable bias.
𝐶𝑂𝑁𝑇𝑅𝑂𝐿 𝑖,𝑡 , recommended by existing studies on financial fragility.
Data description
This study analyzes a panel dataset comprising over 278 observations from 28 Vietnamese banks, collected between 2010 and 2020 The selected financial institutions are listed on various stock exchanges in Vietnam, including the Ho Chi Minh Stock Exchange, Hanoi Stock Exchange, and UPCOM, and are identified by their respective stock tickers.
The dataset's frequency is categorized by year, focusing on a specific time period chosen for observation due to its significance in the analysis.
2010 that the data provided by banks started to become completely detailed and updated In addition, the time frame is sufficient for carrying out observations
The model evaluates several factors, including bank complexity, bank risk, and a control variable group comprising bank specifications and macroeconomic indicators Bank complexity is assessed by analyzing the number of different types of subsidiaries reported in each bank's annual reports, which can be accessed on Vietstock, a reliable source for stock market information Additionally, essential data for calculating the z-score and variables like Loan to Assets, Deposit to Assets, Cost to Income, Assets, Loan Loss Reserve (LLR), and Non-Performing Loans (NPL) are also available on this platform Furthermore, the World Bank provides data on various macroeconomic indicators, including GDP and inflation.
RESULT AND COMMENTS
Bank Complexity
Figure 1 Distribution of Bank Complexity Over Banks
The bar chart illustrating bank complexity in Vietnam reveals that Vetcombank (VCB), BIDV (BID), and Sacombank (STB) rank as the most complex banks, with complexity levels nearing one This indicates their significant involvement in various non-banking sectors examined in the study In contrast, other banks such as BacA Bank (BAB), Eximbank (EIB), HDBank (HDB), Kien Long Bank (KLB), and LienViet Post Bank exhibit lower complexity levels.
LPB, VietBank (VBB), and VPBank (VPB) are categorized as banks with low complexity, as indicated by a score of 0 in the bar chart, signifying their exclusive focus on the banking sector without any subsidiaries Most of the remaining banks exhibit moderate complexity levels ranging from 0.6 to 0.8, typically owning three to four subsidiaries Banks such as MaritimeBank (MSB), PVcomBank (PVCOM), An Binh Bank (ABB), Petrolimex Group Bank (PGB), and Vietnam International Bank (VIB) fall within a complexity range of 0.4 to 0.6, indicating they possess about two to three subsidiaries Meanwhile, Saigon Hanoi Bank (SHB), Tien Phong Bank (TPB), Orient Commercial Bank (OCB), and Nam Viet Bank (NVB) show complexity levels between 0.2 and 0.4, owning one or two subsidiaries The data reveals that most Vietnamese banks have at least one subsidiary outside the banking sector, highlighting a trend towards diversification similar to that seen in developed banking systems, despite few reaching near-absolute complexity as defined by the Herfindahl-Hirschman index.
Bank Risk
Figure 2 Distribution of bank risk over banks
The bar chart highlights the impact of various bank risks on financial institutions in Vietnam, revealing that TienPhong Bank (TPB) has the highest risk index at over 0.18, indicating it is the most dangerous bank In contrast, Techcombank (TCB) and VietBank (VBB) have moderate risk levels ranging from 0.06 to 0.08, with TCB exhibiting greater complexity due to its three subsidiaries, while VBB has no subsidiaries, resulting in a complexity score of 0 Most banks fall within a risk index of 0.02 to 0.06, showcasing diverse operational complexities Only BacA Bank (BAB), BIDV (BID), and SaiGon Bank (SCB) present very low risks below 0.02, with BAB being straightforward and BID and SCB being more complex Overall, the data indicates that while many banks share similar risk levels, the relationship between bank risk and complexity remains inconclusive.
Figure 3 The correlation between bank complexity and bank risk
The scatter plot demonstrates a positive correlation between the risk level and complexity of banks, suggesting that higher complexity is associated with increased risk However, the scatterplots in the histogram reveal a lack of uniformity and a weak correlation between these variables To achieve more reliable results, it is essential to focus on controlling variables and refining the econometric model, which will be discussed in the following section.
Control Variables
Mean Maximum Minimum Std.Dev Observations
This article presents a summary of descriptive statistics based on a dataset comprising 278 observations The primary focus is on Bank Risk as the dependent variable, while Bank Complexity serves as the independent variable Additionally, several quantitative control variables are included, namely Cost_to_Income, Deposit_to_Asset, Loan_to_Asset, LLR, NPL, Assets, GDP, and INF.
It is seen that the average degree of risk of those banks is about 0.039 from 2010 to
2020 Besides, they also have their complexity about 0.46 on average It means that banks in Vietnam own at least about two subsidiaries About average cost to income, there is about
The observed financial index indicates that Vietnamese banks operate at a moderate efficiency level, with an average expense-to-income ratio of 54.99% The deposit-to-asset ratio stands at 0.65, suggesting that banks are not very effective in mobilizing deposits and often rely on more expensive external funding sources Additionally, the average loan-to-asset ratio of 56.17 reflects a relatively safe capital structure, while the Loan Loss Reserve (LLR) averages 72.26, indicating prudent provisioning against bad debts However, the significance of these figures requires trend analysis and peer comparison for a meaningful assessment The average Non-Performing Loan (NPL) ratio of 2.25% highlights potential operational challenges due to high bad debt levels Key economic indicators, including average assets, GDP, and inflation rates, are also essential for comparative analysis to evaluate bank performance objectively.
Standard deviation measures the dispersion of statistical values from the mean, indicating volatility; a high standard deviation signifies significant fluctuations In this study, the standard deviation of cost to income, loan to asset, and loan loss reserves (LLR) is notably high, suggesting that these variables often deviate substantially from their average values This elevated standard deviation also reflects a higher level of risk, indicating considerable variability in these metrics In contrast, other variables with lower standard deviations demonstrate more stable volatility.
Empirical Results
The following table presents the results of data series regression of 28 commercial banks in the period 2010 - 2020 through two methods: fixed effect regression (1) and random effect regression (2)
The research findings indicate that bank complexity positively influences bank risk at a 5% significance level, yet the overall outcome suggests a negative correlation, revealing that increased bank complexity is associated with reduced risk This contradicts existing studies from other countries, such as Ho (2020), which argue that greater complexity can elevate a bank's risk due to the challenges in controlling numerous subsidiaries engaged in non-banking or non-financial activities This perspective aligns with the views of Kwan et al.
In 2019, it was observed that banks can absorb various risks from their subsidiaries This finding aligns with perspectives from other research studies, including those by Ho et al., particularly in the context of Vietnamese banks.
A study conducted in 2020 revealed that bank complexity and risk positively influence each other, suggesting that certain levels of complexity can benefit an organization By diversifying its activities, a bank can engage in multiple sectors, thereby enhancing its significance in the economic operations of countries.
A 2021 study highlights that enhanced complexity in banking operations can mitigate risks associated with bad debt, offering a potential income boost for financial institutions Supporting this notion, research conducted by Goetz et al further explores the relationship between operational complexity and risk management in the banking sector.
Research by Cetorelli et al (2017) indicates that complex banks, similar to those in Vietnam, can effectively manage liquidity through internal capital markets This approach allows banks and their subsidiaries to share risks, thereby reducing overall exposure.
The Cost to Income ratio exhibits a positive correlation with bank risk at a 1% significance level, indicating that higher ratios are associated with increased bank risk This relationship is supported by empirical findings from Vietnamese banks, aligning with previous studies, such as Kingu et al (2018), which also identified a positive link between the Cost to Income ratio and credit risk A decreasing Cost to Income ratio suggests that banks are effectively managing operational costs to enhance efficiency Experts note that a high Cost to Income ratio reflects substantial operational expenditures, particularly in managing and monitoring bad debts, corroborated by Tripe (1998), who highlighted the positive correlation between operational expenses and bad debt ratios.
The Assets variable demonstrates a significant correlation with bank risk, aligning with the Cost to Income variable at a 1% significance level, indicating it can explain bank risk with 99% confidence This finding is consistent with previous research, such as Laeven et al (2014), which highlights the positive relationship between bank assets and risk, suggesting that larger banks may encounter management challenges due to their complexity Similarly, Terraza (2015) found that medium and large banks face greater risks compared to smaller banks, which can more effectively manage assets and control costs, thereby minimizing the risks associated with bad debt and liquidity.
The Loan Loss Reserve (LLR) is statistically significant at the 95% confidence level in relation to bank risk, exhibiting an inverse correlation; as LLR increases, bank risk decreases This relationship highlights the importance of LLR as a reserve for covering estimated losses in a bank's loan portfolio An increase in LLR indicates that the bank has adequate financial resources to manage bad debts, thereby reducing credit risk This finding aligns with previous studies, including Ozili and Outa (2017), which identified a negative effect of LLR on bank risk, and research by Jasman and Murwaningsari (2022), which supports the notion that higher provisions for bad debts correlate with lower credit risk.
The variables Loan to Assets, GDP, Inflation, NPL, and Deposit to Assets are not statistically significant in Vietnam's banking context at a 95% confidence level Unlike previous studies that indicate a positive correlation between loans to assets and bank risks, this study finds no significant relationship in Vietnam Similarly, while prior research suggests that a high Deposit to Asset ratio enhances financial stability, this study indicates different results for Vietnamese banks Non-Performing Loans (NPL) also show no significant impact on bank risk in Vietnam, contradicting findings from other regions where NPLs negatively affect lending quality and increase risk Additionally, neither GDP nor inflation demonstrated a significant effect on bank risk in Vietnam, diverging from other studies that link GDP growth and inflation to financial stability Overall, the findings suggest that macroeconomic variables do not significantly explain bank risks in Vietnam.
Robustness Check with Alternative Measure
In this section, Totalcount serves as an alternative measure for assessing bank risk, replacing the concept of bank complexity This allows individuals to evaluate the robustness of the results presented in the two tables above As previously noted, Totalcount quantifies bank complexity by counting the number of affiliates owned by a bank.
Table 3.1 Alternative Measure (with Totalcount)
Table 3.1 presents the empirical findings of the econometric model utilizing the Totalcount variable Previous studies have sparked a debate among experts regarding the impact of the number of affiliates on bank risk, which can be either negative or positive A negative impact may arise due to challenges in control (Scharfstein & Stein, 2000), while a positive effect may result from diverse income sources that enhance liquidity management (Laeven & Levine, 2007; Goetz et al., 2016; Cetorelli et al., 2017).
The analysis reveals that Totalcount and bank risk are statistically significant at a 95% confidence interval in both fixed and random effect regression models The findings indicate a negative correlation, suggesting that an increase in affiliates corresponds to a reduction in bank risk This aligns with previous observations regarding bank complexity, reinforcing the robustness of these results Specifically, banks with higher complexity or more subsidiaries are better positioned to mitigate their risk.
The Cost to Income ratio exhibits a positive correlation with bank risk at a 1% significance level, consistent with previous findings that included the bank complexity variable Additionally, assets are statistically significant in relation to bank risk Consequently, both the Cost to Income ratio and assets provide stable explanatory power regarding the impact of bank risk.
On the contrary, when the totalcount variable is included in the model instead of bank complexity, the LLR is no longer statistically significant Hence, this variable has no robustness
In this analysis, the fixed effect model reveals that Deposit to Assets significantly impacts bank risk at a 1% level, while the random effect model shows significance at 5% This indicates a positive correlation, suggesting that an increase in bank deposits correlates with heightened risk Conversely, the random effect model identifies a significant inverse relationship with the Non-Performing Loans (NPL) variable at a 95% confidence level, indicating that higher bad debt levels may lead to reduced bank risk These findings contrast sharply with existing literature, which typically reports the opposite effects.
Bank risk is primarily influenced by Totalcount, Cost to Income, and Assets, demonstrating robustness in the model When analyzing bank complexity as an independent variable, similar results are observed, reinforcing the notion that bank complexity negatively impacts bank risk Conversely, other variables lack robustness due to their insignificance and illogical outcomes.
Table 3.1 utilizes Totalcount as a measure of bank complexity, while Table 3.2 employs the T index The T index reflects the total number of affiliate types that own banks, as indicated by the bank complexity formula.
The analysis indicates that T does not influence bank risk, suggesting that the sectors in which a bank operates do not affect its risk levels Furthermore, existing studies have not identified a correlation between these factors Instead, variations in bank risk are primarily attributed to bank complexity and total count Empirical findings support the thesis that increased complexity, characterized by a greater number of subsidiaries, is associated with reduced risks for banks.
In the new model, both the Cost to Income ratio and Assets demonstrate statistical significance for bank risk at a 99% confidence level, indicating a positive relationship with bank risk These findings align with previous sections, confirming the robustness of these two variables.
In the model utilizing the independent variable T, the fixed effect analysis reveals that the ratio of Deposit to Assets is statistically significant at a 95% confidence level; however, it exhibits a positive relationship with bank risk, which is inconsistent with the findings from the model based on the independent variable T.
The findings in this section reinforce previous conclusions, indicating that bank risk is effectively explained by Cost to Income and Assets, which are robust variables in the model In contrast, other variables lack robustness due to their insignificance and illogical outcomes.
POLICY IMPLICATION AND RECOMMENDATION
Recent concerns regarding banking complexity and its implications have prompted an investigation into the impact of bank complexity on bank risk This study is structured into five parts, beginning with an introduction that outlines the research background and questions It includes a review of existing literature to establish a theoretical and empirical foundation for the research framework, focusing on the interplay between bank complexity and risk The methodology section details data collection and analysis, employing the Herfindahl index to measure bank complexity and using inverse z-score through panel data analysis to assess bank risk amidst ongoing financial system changes The findings reveal a significant relationship between bank complexity and risk, indicating that increased complexity can reduce risk due to its negative correlation This diversification allows banks to engage in multiple sectors, enhancing their economic relevance Supporting literature from Correa & Goldberg (2021) and similar studies by Goetz et al (2016) and Cetorelli et al (2017) further underscores that complex banks can effectively manage liquidity through internal capital markets To ensure robustness, the study adjusts the model using Totalcount and T, revealing that while a higher number of affiliates (Totalcount) correlates with reduced risk, the type of affiliates (T) does not affect bank risk The paper concludes with recommendations for further research and policy directions.
The study revealed a negative correlation between bank complexity and risk, primarily due to the small size of banks in Vietnam, which facilitates easier management However, as banks grow, management efficiency may decline, necessitating an annual review and update of governance strategies to align with changing contexts Emphasizing the importance of government regulations, the research indicates that banks with strong governance can better navigate complex structures, diversify income streams, and reduce exposure to specific risks and liquidity issues While most developed countries have adopted BASEL III, many Vietnamese banks are still implementing BASEL II's three pillars Transitioning to BASEL III, which introduces significant innovations, is crucial for ensuring comparable capital adequacy across banks and requires capital adjustments based on risk levels, ultimately enhancing the stability of Vietnam's banking system.
Aldasoro, I., Hardy, B., Jager, M (2021) The Janus face of bank geographic complexity
Alawattegama, K (2018) The impact of enterprise risk management on firm performance: Evidence from Sri Lankan banking and finance industry International Journal of
Boyd, J H and Graham, S L (1988) The profitability and risk effects of allowing bank holding Federal Reserve Bank of Minneapolis Quarterly Review-Federal Reserve Bank of Minneapolis 12(2), pp.3
Berger, A., Hasan, I., Zhou, M (2010) The effects of focus versus diversification on bank performance: Evidence from Chinese banks Journal of Banking & Finance 34(7), pp.1417-1435
Baron, M and Xiong, W (2017) Credit expansion and neglected crash risk The Quarterly Journal of Economics 132(2), pp.713-764
Berger, A., El Ghoul, S., Guedhami, O., Roman, R (2017) Internationalization and Bank Risk Management Science 63(7), pp.2283-2301
Bhowmik, K., P and Sarker, N (2021) Loan growth and bank risk: empirical evidence from SAARC countries National Library of Medicine 7(5)
Cetorelli, N and Goldberg, L (2014) Measures of Global Bank Complexity Economic Policy
Cetorelli, N., McAndrews, J., Traina, J (2014) Evolution in bank complexity Economic
Cucinelli, D (2015) The impact of non-performing loans on bank lending behavior: Evidence from the Italian banking sector Eurasian Journal of Business and Economics 8(16), pp 59-71
Cetorelli, N., Jacobides, M., Stern, S (2017) Transformation of corporate scope in U.S banks: patterns and performance implications Federal Reserve Bank of New York
Chernobai, A., Ozdagli, A., Wang, J (2021) Business complexity and risk management: Evidence from operational risk events in U.S bank holding companies Journal of
Correa, R and Goldberg, L (2021) Bank complexity, governance, and risk Journal of
DeYoung, R., Evanoff, D., Molyneux, P (2009) Mergers and Acquisitions of Financial Institutions: A Review of the Post-2000 Literature Journal of Financial Services
Ekinci, R and Poyraz, G (2019) The Effect of Credit Risk on Financial Performance of
Deposit Banks In Turkey Available from: < https://www.researchgate.net/publication/336816182_The_Effect_of_Credit_Risk_on_Fi nancial_Performance_of_Deposit_Banks_In_Turkey> [Accessed 09 September 2022]
Farkasdi, S., Septiawan, B., Syawal Alghifari, E (2021) DETERMINANTS OF
COMMERCIAL BANKS PROFITABILITY: EVIDENCE FROM GERMANY Jurnal
Foglia, M (2022) Non-Performing Loans and Macroeconomics Factors: The Italian Case Risks 2022 10, pp.21
Goetz, M., Laeven, L., Levine, R (2016) Does the geographic expansion of banks reduce risk? Journal of Financial Economics 120(2), pp.346-362
Ho, K., Wong, E., Tan, E (2020) Complexity of global banks and the implications for bank risk: Evidence from foreign banks in Hong Kong Journal of Banking & Finance 134
Habachi, M and Haddad, S (2021) Impact of Covid-19 on SME portfolios in Morocco: Evaluation of banking risk costs and the effectiveness of state support measures
Investment Management and Financial Innovations 18, pp.260–76
Igan, D and Pinheiro, M (2011) Credit Growth and Bank Soundness: Fast and Furious? IMF Working Paper
Jordà, Ò., Schularick, M., Taylor, A M (2013) When Credit Bites Back Journal of Money, Credit and Banking 45(2), pp.3-28
Jasman, J and Murwaningsari, E (2022) Loan loss provision index and bank risk: An empirical study in Indonesia Banks and Bank Systems 17(2), pp.27-36
Keeton, W (1991) The Treasury Plan for Banking Reform Economic Review - Federal
Reserve Bank of Kansas City 76(3)
Kingu, S P., Macha, S., Gwahula, R (2018) Impact of Non-Performing Loans on Bank’s Profitability: Empirical Evidence from Commercial Banks in Tanzania International Journal of Scientific Research and Management 6(1)
Kwan, S., Ho, K., Tan, E (2019) Complexity of Global Banks and Their Foreign Operation in Hong Kong SSRN Electronic Journal,
Khan, M A., Siddique, A., Sarwar, Z (2020) Determinants of non-performing loans in the banking sector in developing state Asian Journal of Accounting Research 5, pp.135–
Kulinska-Sadłocha, E., Marcinkowska, M., Szambelanczyk, J (2020) The impact of pandemic risk on the activity of banks based on the Polish banking sector in the face of COVID 19 Bezpieczny Bank 2, pp 31–59
Laeven, L and Levine, R (2007) Is there a diversification discount in financial conglomerates? Journal of Financial Economics 85(2), pp.331-367
Laeven, L and Levine, R (2009) Bank governance, regulation and risk taking Journal of
Laeven, L., Rainovski, L., Tong, H (2014) Bank Size and Systemic Risk International Monetary Fund
Lepetit, L., and Strobel, F (2015) Bank insolvency risk and Z-score measures: A refinement Finance Research Letters 13 pp 214-224
Li, X., Tripe., D., Malone., C (2017) Measuring bank risk: An exploration of z-score
Available from: < https://papers.ssrn.com/sol3/papers.cfm?abstract_id(23946> [Accessed 09 September 2022]
Lu, J and Boateng, A (2018) Board composition, monitoring and credit risk: Evidence from the UK banking industry Review of Quantitative Finance and Accounting 51, pp 1107–
Lestari, D (2018) Corporate Governance, Capital Reserve, Non-Performing Loan, and Bank Risk Taking International Journal of Economics and Financial Issues 8(2), pp 25-32
Louati, S and Boujelbene, Y (2020) Inflation targeting and bank risk: The interacting effect of institutional quality Cogent Business & Management 7(1)
Lysiak, L., Masiuk, I., Chynchyk, A., Yudina, O., Olshanskiy, O., Shevchenko, V (2022) Banking Risks in the Asset and Liability Management System Journal of Risk and Financial Management 15
Mwangi, M., Muturi, W., Ombuki, C (2015) THE EFFECTS OF DEPOSIT TO ASSET RATIO ON THE FINANCIAL SUSTAINABILITY OF DEPOSIT TAKING MICRO FINANCE INSTITUTIONS IN KENYA International Journal of Economics,
Musau, S., Muathe, S., Mwangi, L (2018) Financial Inclusion, GDP and Credit Risk of Commercial Banks in Kenya International Journal of Economics and Finance 10(3)
Mpofu, T R and Nikolaidou, E (2018) Determinants of credit risk in the banking system in Sub-Saharan Africa Review of development finance 8(2), pp 141–153
Mosk, B and Welz, P (2022) Financial stability implications of higher than expected inflation Financial Stability Review
Marinelli, G., Nobili, A., Palazzo, F (2022) The multiple dimensions of bank complexity: Effects on credit risk-taking Journal of Banking & Finance 134
Martynova, N and Vogel, U (2022) Banks’ complexity-risk nexus and the role of regulation
Ozili, K P and Outa, E (2017) Bank loan loss provisions research: A review Borsa Istanbul Review 17(3), pp 144-163
Parvin, S., Chowdhury, A., Siddiqua, A., Ferdous, J (2019) Effect of Liquidity and Bank Size on the Profitability of Commercial Banks in Bangladesh Asian Business Review 9(1), pp.7-10
Pham, H., Pham, Anh., Truong, H., Dao, N (2020) Could Non-bank Financial Institutions be a Threat to Financial Stability? Evidence from Vietnam Available from:
[Accessed 09 September 2022]
Scharfstein, D and Stein, J (2000) The Dark Side of Internal Capital Markets: Divisional Rent-Seeking and Inefficient Investment The Journal of Finance 55(6)
Schularick, M and Taylor, M., A (2012) Credit Booms Gone Bust: Monetary Policy,
Leverage Cycles, and Financial Crises, 1870-2008 AMERICAN ECONOMIC
Shahid, S M., Gul, F., Naheed, K (2019) Credit Risk and Financial Performance of Banks: Evidence from Pakistan International Journal of Business & Management 14(1)
Shaikh, A A., Glavee-Geo, R., Karjaluoto, H (2021) How relevant are risk perceptions, effort, and performance expectancy in mobile banking adoption? In Research Anthology on Securing Mobile Technologies and Applications pp 692–716
Saidane, D., Sène, B., Kanga, D., K (2021) Pan-African banks, banking interconnectivity: A new systemic risk measure in the WAEMU Journal of International Financial Markets Institutions and Money 74, pp.101405
Templeton, W and Severiens, J (1992) The Effect of Nonbank Diversification on Bank Holding Company Risk Business and Economics 31(4), pp.3-17
Tripe, D (1998) Cost to income ratios in australasian banking Available from: < https://www.researchgate.net/publication/228423135_Cost_to_income_ratios_in_australa sian_banking> [Accessed 09 September 2022]
Terraza, V (2015) The effect of bank size on risk ratios: Implications of banks’ performance Procedia Economics and Finance 30, pp 903-909
Torre, O., Begoủa, M C S., Sergio, S A (2021) Sustainable Banking, Market Power, and Efficiency: Effects on Banks’ Profitability and Risk Sustainability 13, pp 1298
Yurdakul, F (2014) Macroeconomic Modelling Of Credit Risk For Banks Procedia - Social and Behavioral Sciences 109
Zhang, D., Cai, J., Kutan, A M., Dickinson, D G (2016) Non-performing loans, moral Hazard and regulation of the Chinese Journal of Banking and Finance 63, pp 48-60
Appendix 3: Fixed Effect Regression (with Totalcount)
Appendix 4: Random Effect Regression (with Totalcount)
Appendix 5: Fixed Effect Regression (with T)