Diversification strategy, bank risk and performance in Vietnam: Before, during and after financial crisis

Pham Khanh Duy - University of Economics Ho Chi Minh City


The study uses data of 44 commercial banks in Vietnam during the period from 2002 to 2019 to analyze the impacts of diversification and ownership strategy on various indicators of bank risks and performance in Vietnam before, during and post Global financial crisis. The result shows that banks with diversified asset perform better and enjoy lower risk. However, during global financial crisis, these banks have negative return and bear more risk. State-owned banks achieve the lowest profitability ratios in the system, while foreign banks achieve the highest return rates. Last but not least, listed banks are seen to have a better performance than unlisted banks. Research also indicates that big banks with large market share and total assets would have greater diversification in compare to smaller banks.

Keywords: Vietnam, Bank risk, bank performance, diversification, financial crisis, ownership structure.

1. Introduction

Banking industry and financial market are nowadays extremely competitive with changing constantly. In the last decades, dozens of banking crisis occur one by one, worse and worse, namely, Bank of England (1992) and Barings Bank (1995). Recently, Global financial crisis leads to collapse of biggest and oldest banks in the US, i.e. Washington Mutual and Lehman Brothers in 2008. This, again, raise the question about risk. An old and tough question that could never conclude: How to detect and control risk?

Academically, there are many answers for this question. However, during an economic downturn or a financial crisis, the most popular answer for this question is diversification. Previous researches prove that diversification significantly reduces financial risk and posibility of bankruptcy of banks, especially with mergered banks (Boyd and Graham 1988). Diversification strongly effects bank risk in two dimensions: the more diversified, the higher the financial leverage banks enjoy; and diversification through operation restructuring (Berger, Hasan and Zhou, 2010). Research on bank diversification in the case of Vietnam is still limited, including Nguyen and Vo (2015), Batten and Vo (2016), Vo (2017, 2020) and Luu, Nguyen, and Vu (2019). However, these studies only mention the effect of income and funding diversification on either risk or performance, but not both.

After outlining studies related banking and finance on prestigious journals, the author considers conducting a study to assess the impact of asset diversification on bank risk and performance in an emerging market, Vietnam with the following contributions. Firstly, author measure the level of asset diversification for commercial banks in Vietnam, an emerging market. Second, the paper evaluated the effects of diversification strategies on both risk and performance of banks. Third, the study also systematically evaluates the effect of global financial crisis 2007-2009 and ownership strategy on the relationship between bank diversification, risk and performance.

2. Literature review

Currently, there are different theories that analyze the effects of diversification strategy and risk. Boyd and Graham (1988) based on the classical principles  showing the following three theoretical bases. First, the expansion of diversification components in the banking institutional system (shifting from traditional activities to non-traditional service elements) will reduce the bank profit performance fluctuations - also known as risk, derive from the asset adjustment mechanisms on diversification strategy. Second, in the opposite direction, the risk will be motivated when there is an overwhelming effect from the characteristics of profit fluctuations. In other words, the benefits of diversification cannot overcome the inherent volatility risk of bank profits, increasing the risks of the bank. Ultimately, the banking system's risk issue is likely irrelevant to diversification adjustment decisions, as banks have created contingencies as well as built up technical regulatory barriers to may react with risks impairing the value of the risk.

A second group of theories, that Keeton (1991); Templeton and Severiens (1992), with the theory emphasizes the role of "the list" modern. This theory group argues that the variation in shareholder returns can be adjusted by appropriate diversification incremental decisions. This problem is evident in the overall adjustment mechanism across the portfolio weights for the variance of each activity fluctuations old and new activity. From here will form a cycle of reactions increased activity diversification will help adjust decline overall risk of banking operations, but with conditions the portfolio diversification of new unrelated to old banking activities.

Finally, we can consider the arguments of Berger, Hasan, and Zhou (2010), with the following two prominent theoretical background. First, on the basis of classical theory, banks should diversify as much as possible through leverage structure or promote restructuring across all sectors of banking operations. This will help to reduce the probability of bearing risks from financial stress or even bankruptcy. However, based on the view of Winton (1999), Berger et al. (2010) argue that the decision of bank managers to increase their diversification strategy will promote the possibility of creating problem subjected "agent cost", thereby increasing risks for the banking system. This means that, according to the theory of "agent cost", the adjustment to expand diversification will cause negative consequences for the governance system, reducing the "real value" of the bank and creates uncertain risks for the bank's operations in the future.

3. Data and methodology

3.1. Data

Dynamic panel data from 44 Vietnamese banks from 2002 to 2019 are used in this research. Zooming in a country figures permits us to analyse the bank behavior in a more coherent environment. The data is collected from several good data sources such as Bankscope, Orbis Bank Focus, Vietstock, and individual bank financial reports.

There are at least two reasons to select this sample period: First, to systematically examine the effects of the crisis of 2008-2009, or global financial crisis, to banking system in Vietnam. After a high rate development from 2003 to 2007 with average annual rate of 8.0%, Vietnam’s GDP growth rate was slow down in the period from 2008 to 2012, before picking up afterward. Especially, in 2008 and 2009, Vietnam had been affected strongly by global financial crisis, with lowest GDP growth rate of only 5.3% (ADB, 2013).

Second, there are too many missing values in Bankscope/OrbisBankFocus database for the sample banks before the year 2002. Too many missing values can create sample selection bias in favor of few banks. After eliminating missing values and outliners, we left with a dynamic panel of 44 banks with approximately 580 bank-year observations.

3.2. Construction of variables

In our empirical models, the key explainatory variables are measures of bank diversification and explained variables are measures of bank risk and return. We also add several control variables. With the aim of assessing the impact of diversification on the performance and risk of banks in Vietnam, this research pursues three phases: (1) measuring the level of asset diversification; (2) examine bank’s risk and return; (3) explain the link between diversification and bank return and risk. Details are presented in the following subsections.

3.2.1. Measures of diversification

Diversification is a form of index that shows the different aspects of an element. This index is shown through the formula of HHI, representing the measurement of diversification for each bank (Curi, Lozano-Vivas, & Zelenyuk, 2015; Elsas, Hackethal, & Holzhäuser, 2010). The diversification index is constructed by subtracting HHI from 1, so that it increases with diversification. We focus on most important categories for commercial banks operating in Vietnam. For asset diversification (ASTDIV), for each bank i at time t, the formula is measured as follows:

Where earning assets (EA) is the sum of the four numerators: interbank loans (IBLOAN), customer loans (CLOAN), securities (SEC) and other earning assets (OTHEREA).

3.2.2. Dependent variables

Following recent literature (Amidu & Wolfe, 2013; Chen & Zeng, 2014; Clark, Mare, & Radić, 2018), this study uses bank income statement return on average assets (ROA) and return on average equity (ROE) computed as the ratio of net income to total assets and equity, respectively.

For a comprehensive analysis and robust results, we employ five inverse bank risk indicators, that have been widely used in banking studies (Amidu & Wolfe, 2013; Vo, 2018). The first two risk proxies are standard deviation of the return on average assets σ(ROA) and equity σ(ROE), measured as the standard deviation of annual return on assets and equity of a bank over the sample period. Higher values of σ(ROA) and σ(ROE) represents higher bank income risk.

We also construct the two risk-adjusted measures of return on average assets (RAROA) and return on average equity (RAROE) by dividing ROA and ROE by their standard deviation (σ) calculated over the entire sample period. The formulas are shown below:


Our last proxy is a measure of default risk of each bank by computing the Z-score (ZScore) as


where ETA is the ratio of total equity to total asset (Boyd & Graham, 1986). Lower values of RAROA, RAROE and ZScore denote respectively a higher risk and higher probability of default.

3.2.3. Control and dummy variables

We include in the regressions a set of variables to control for characteristics that can affect bank risk-taking and performance in addition to diversification. We choose these variables based on existing literature on bank diversification (DeYoung and Roland 2001, DeYoung and Rice 2004, Stiroh and Rumble 2006, Laeven and Levine 2007, Sanya and Wolfe 2011, Lee et al., 2014, Meslier et al., 2014, Edirisuriya et al., 2015, Sissy et al., 2017): the natural logarithm of bank total assets (LOGASSET) to account for bank size; the ratio of equity to total assets (ETA) to control for bank capitalisation; the ratio of loans loss provision (LLP); dummy variables (SOB) to account for the State-owned banks, (LIST) to account for the listed banks and (FOB) to account for Foreign bank’s wholly owned subsidiaries; and finally the inflation rate (INF) and the growth rate of the real gross domestic product (GDP) to take into account diffences in the macroeconomic environment.

3.3. Measurement of relationship between diversification, risk and return

Using quantitative method, specifically a general panel data with fixed effects model (FEM) and System Generalized Method of Moments estimation method (System-GMM), we estimate the following equation to test the effect of asset, income and funding diversification on bank risk and performance:

where Y denotes dependent variables measure bank risk and performance, which can be either ROA, ROE, σROA, σROE, RAROA, RAROE or ZScore; i and t identify individual banks in the sample and time variable in years; Y_(i,t-1) are their lagged value; α: coefficient; βs are the parameters to be estimated; Control is matrix of bank specific control variables and macroeconomic indicators; Dummy is matrix of dummy variables.

4. Results and discussion

Table 1, descriptive statistics, shows that asset diversification variables are in the range of 0.04 to 1, with average value of 0.45 and large standard deviation of 0.16. This suggests that there is a big difference in asset diversification across the sample banks.

Table 1. Descriptive statistics

Notes: ADIVCRIS, interation terms between asset diversification index and financial crisis dummy variable, represent the level of asset diversification during crisis period 2007 - 2009.

Source: Bankscope and author’s own calculation. This table reports the summary statistics of annual data for 44 sample banks in Vietnam, from 2002 to 2019.

During the crisis, the average asset diversification (ASTCRI) decline to around 0.09, while the standard deviation increases to about 0.19. The lowest and highest values slightly decrease. It is are shown that banks restrict their diversification during crisis.

Table 2. Empirical results

Note: ***,**,* stand for statistically significant at 1&, 5% andd 10%, respectively

Source: Author’s calculations

In table 2, the result shows that asset diversification has a positive impact on bank’s performance and risk-adjusted returns in the normal period. Meanwhile, in times of crisis, bank profitability are adversely affected by diversification, i.e. the trading and investment in external industry and holding non-core assets; reflected through the bank behaviors when they try to restrict diversification activities in difficult economic times. This is also related to the agency cost when management uses bank capital to finance projects that involve potential risks.

In regards to ownership structure, state-owned banks achieve the lowest profitability ratios in the system, while foreign banks achieve the highest return rates. Last but not least, listed banks are seen to have a better performance than private banks.

Regarding the risk of bank profitability, asset diversification reduces the volatility in ROA in the non-crisis period. Banks, with diversified asset, enjoy lower volatility in return on total assets. We also find evidence of an inverse relationship between asset diversification and SDROA,SDROE during the crisis period. In other words, during the crisis period, the diversification of assets increased the risks in ROA and ROE fluctuations and was statistically significant. This suggests that banks need to minimize immediately diversification of assets during the crisis to ensure profit stability.

As for the impact of diversification on bank's Zscore bankruptcy coefficient, asset diversification once again increases the operational safety of banks in the non-crisis period. Meanwhile, we found no evidence on the impact of asset diversification on bankruptcy risk during the crisis.

5. Conclusion

The diversification of assets has a positive impact on reducing the bank's risk in the pre- and post-crisis stages but increases the risk in the crisis period. This indicates that bankers need to have flexible and appropriate adjustments to the concentration or diversification of assets in different stages. Yet, the transfer of assets structure is not simple. Hence, banks need to make accurate forecasts on long-term macroeconomic situation in order to make proper provisions and decisions in managing their asset portfolios.


  1. Amidu, M., & Wolfe, S. (2013). Does bank competition and diversification lead to greater stability? Evidence from emerging markets. Review of Development Finance, 3(3), 152-166.
  2. Batten, J. A., & Vo, X. V. (2016). Bank risk shifting and diversification in an emerging market. Risk Management, 18(4), 217-235.
  3. Berger, A. N., Hasan, I., & Zhou, M. (2010). The effects of focus versus diversification on bank performance: Evidence from Chinese banks. Journal of Banking & Finance, 34(7), 1417-1435.
  4. Boyd, J. H., & Graham, S. L. (1986). Risk, regulation, and bank holding company expansion into nonbanking. Quarterly Review, Spring, 2-17.
  5. Boyd, J. H., & 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), 3.
  6. Chen, P.-F., & Zeng, J.-H. (2014). Asymmetric effects of households’ financial participation on banking diversification. Journal of Financial Stability, 13, 18-29.
  7. Clark, E., Mare, D. S., & Radić, N. (2018). Cooperative banks: What do we know about competition and risk preferences? Journal of International Financial Markets, Institutions and Money, 52, 90-101.
  8. Curi, C., Lozano-Vivas, A., & Zelenyuk, V. (2015). Foreign bank diversification and efficiency prior to and during the financial crisis: Does one business model fit all? Journal of Banking & Finance, 61, S22-S35.
  9. Elsas, R., Hackethal, A., & Holzhäuser, M. (2010). The anatomy of bank diversification. Journal of Banking & Finance, 34(6), 1274-1287.
  10. Keeton, W. R. (1991). The Treasury plan for banking reform. Economic Review, May/June, 5-2.
  11. Luu, H. N., Nguyen, L. Q. T., & Vu, Q. H. (2019). Income diversification and financial performance of commercial banks in Vietnam. Review of Behavioral Finance.
  12. Nguyen, T. C., & Vo, D. V. (2015). Risk and income diversification in the Vietnamese banking system. Journal of Applied Finance and banking, 5(1), 93.
  13. Templeton, W. K., & Severiens, J. T. (1992). The effect of nonbank diversification on bank holding company risk. Quarterly Journal of Business and Economics, 3-17.
  14. Vo, X. V. (2017). How does the stock market value bank diversification? Evidence from Vietnam. Finance Research Letters, 22, 101-104.
  15. Vo, X. V. (2018). The role of bank funding diversity: Evidence from Vietnam. International Review of Finance.
  16. Vo, X. V. (2020). The role of bank funding diversity: Evidence from Vietnam. International Review of Finance, 20(2), 529-536.
  17. Winton, A. (1999). Don't put all your eggs in one basket? Diversification and specialization in lending. Diversification and Specialization in Lending. 

Chiến lược đa dạng hóa, rủi ro và kết quả kinh doanh của các ngân hàng tại Việt Nam: Trước, trong và sau khủng hoảng tài chính toàn cầu

Phạm Khánh Duy - Đại học Kinh tế TP. Hồ Chí Minh

Tóm tắt:

Thông qua dữ liệu được thu thập từ 44 ngân hàng thương mại tại Việt Nam trong giai đoạn từ năm 2002 đến năm 2019, nghiên cứu này phân tích tác động của chiến lược đa dạng hóa và chiến lược cấu trúc sở hữu đối với các chỉ số khác nhau đánh giá rủi ro và kết quả kinh doanh của các ngân hàng tại Việt Nam trước, trong và sau cuộc khủng hoảng tài chính toàn cầu. Kết quả nghiên cứu cho thấy các ngân hàng có tài sản đa dạng hoạt động tốt hơn và đối mặt với rủi ro thấp hơn. Tuy nhiên, trong cuộc khủng hoảng tài chính toàn cầu, các ngân hàng này có lợi nhuận âm và chịu nhiều rủi ro hơn. Các ngân hàng quốc doanh đạt tỷ suất sinh lời thấp nhất trong hệ thống ngân hàng, trong khi đó, các ngân hàng nước ngoài ghi nhận mức tỷ suất sinh lời cao nhất. Bên cạnh đó, các ngân hàng niêm yết có kết quả kinh doanh tốt hơn các ngân hàng chưa niêm yết. Nghiên cứu cũng cho thấy các ngân hàng lớn có thị phần và tổng tài sản lớn sẽ có sự đa dạng hóa hơn so với các ngân hàng nhỏ hơn.

Từ khoá: Việt Nam, rủi ro ngân hàng, kết quả kinh doanh của ngân hàng, đa dạng hóa, khủng hoảng tài chính, cấu trúc sở hữu.