Bank income structure and risk an empirical analysis of vietnam commercial banks

Employing a sample of Vietnam commercial banks for the period 2008–2011, the results indicate that banks expanding into non-interest income activities do not present higher risk than ban

1

MINISTRY OF EDUCATION AND TRAINING

UNIVERSITY OF ECONOMICS HOCHIMINH CITY

-o0o -

TRẦN ĐÌNH KHẢI

BANK INCOME STRUCTURE AND RISK: AN EMPIRICAL ANALYSIS OF VIETNAM COMMERCIAL BANKS

MASTER OF BUSINESS ADMINISTRATION

HO CHI MINH CITY, 2012

2

MINISTRY OF EDUCATION AND TRAINING

UNIVERSITY OF ECONOMICS HOCHIMINH CITY

-o0o - E Â1

TRẦN ĐÌNH KHẢI

BANK INCOME STRUCTURE AND RISK: AN EMPIRICAL ANALYSIS OF VIETNAM COMMERCIAL BANKS

MAJOR: BUSINESS ADMINISTRATION

MAJOR CODE: 60.34.05

MASTER THESIS SUPERVISOR: Dr VÕ XUÂN VINH

HO CHI MINH CITY, 2012

ACKNOWLEDGEMENT

This thesis would not have been possible without the guidance and the help of several individuals who in one way or another contributed and extended their valuable assistance in the preparation and completion of this study

First and foremost, my utmost gratitude to Dr Vo Xuan Vinh, whose sincerity and encouragement I will never forget Dr Vinh has been my inspiration as I hurdle all the obstacles in the completion this research Dr Vinh has spent plenty

of time in many weeks to support me with his fully enthusiasm

I also express my warm gratitude to all professors at Faculty of Business Administration, University of Economics in Hochiminh City for their teaching and sharing during my MBA course

I have many thanks to my classmates who show their sharing and encouragements when I do this thesis

Last but not the least, I want to thank so much to my beloved wife who always understands and encourages me to complete my thesis especially in the hardly time of taking care of my newborn baby

ABSTRACT

This paper investigates the relationship between bank risks and product diversification of the Vietnam banking industry Employing a sample of Vietnam commercial banks for the period 2008–2011, the results indicate that banks expanding into non-interest income activities do not present higher risk than banks that mainly supply loans However, considering size effects and splitting non-interest activities into both trading activities and commission and fee activities, we show that the link with risk is mostly accurate for small banks and essentially driven by commission and fee activities Whereas, for both of small and large banks, a higher share of trading activities is never associated with higher risk

Keywords: Bank risk; Interest income; Non-interest income; Diversification

CONTENTS

ACKNOWLEDGEMENT 1

ABSTRACT 2

LIST OF TABLES 5

ABBREVIATIONS 6

CHAPTER 1: INTRODUCTION 7

1.1 BACKGROUND 7

1.2 RESEARCH PROBLEM 9

1.3 RESEARCH OBJECTIVE 9

1.4 RESEARCH METHODOLOGY AND SCOPE 10

1.5 STRUCTURE OF RESEARCH 10

CHAPTER 2: BASIC THEORETICAL BACKGROUND AND LITERATURE REVIEW 11

2.1 BANK RISK 11

2.2 BANK INCOME 13

2.3 DIVERSIFICATION 13

2.4 BANK RISK AND DIVERSIFICATION 15

2.5 HYPOTHESIS DEVELOPMENT 18

CHAPTER 3: DATA AND RESEARCH METHODS 19

3.1 MODEL 19

3.2 VARIABLES 20

3.2.1 DIVERSIFICATION VARIABLES 20

3.2.2 BANK RISK MEASURES 20

3.2.3 CONTROL VARIABLES 21

3.3 DATA 22

CHAPTER 4: RESULTS AND DISCUSSION 27

4.1 UNIVARIATE MEAN TESTS 27

4.1.1 BANK RISK AND INCOME STRUCTURE 27

4.1.2 INCOME STRUCTURE AND BANK CHARACTERISTICS 30

4.2 MULTIVARIATE REGRESSION ANALYSIS 32

CHAPTER 5: CONCLUSION 44

5.1 RESEARCH FINDINGS: 44

5.2 RESEARCH CONTRIBUTIONS: 45

5.3 RECOMMENDATIONS FOR FUTURE RESEARCH: 45

REFERENCES 46

APPENDIX 49

LIST OF TABLES

Table 1: Descriptive statistics for Vietnam commercial banks, on average over

the period 2008-2011 (%)

Table 2: Descriptive statistics for small and large Vietnam commercial banks,

on average over the period 2008-2011 (%)

Table 3: Income structure and accounting indicators of risk for Vietnam

Table 6: OLS estimations for the first cohort of 13 small banks with M_NNII as

the independent variable

Table 7: OLS estimations for the second cohort of 22 large banks with M_NNII

as the independent variable

Table 8: OLS estimations for the sample of 35 banks with M_COM and

M_TRAD as the independent variables

Table 9: OLS estimations for the first cohort of 13 small banks with M_COM

and M_TRAD as the independent variables

Table 10: OLS estimations for the second cohort of 22 large banks with

M_COM and M_TRAD as the independent variables

ABBREVIATIONS

NII _ Net interest income to net operating income

NNII _ Net non interest income to net operating income

COM _ Net commission and fee income to net operating income

TRAD _ Net trading income to net operating income

Q75 _ Third quartile

Q25 _ First quartile

LOANS _ loans to total assets;

DEP _ deposits to total assets;

EQUITY_ equity to total assets;

EXPENSES _personnel expenses to total assets;

ROA_ return on average assets;

ROE_ return on average equity;

TA _ total assets

SDROA_ standard deviation of the return on average assets;

SDROE_ standard deviation of the return on average equity;

LLP_ ratio of loan loss provisions to net loans;

ADZ_ Z-score;

ADZP_ ‘‘ZP-score”;

ADZP1_ measure of bank portfolio risk;

ADZP2_ measure of leverage risk

OLS _ Ordinary least squares

HOSE_Hochiminh Securities Exchange

HNX_Hanoi Securities Exchange

OTC_ Over The Counter

CHAPTER 1: INTRODUCTION

1.1 BACKGROUND

Following more definitions of bank income structure in the world, bank income structures has two main components: interest income that stems from traditional activities such as giving loan and non-interest income that stems from nontraditional activities such as supplying services, trading services

In the context of financial deregulation that took place in the seventies and in the eighties, western banking systems faced major changes in the form of increased competition, concentration and restructuring Banks have reacted to the new environment by adopting a proactive strategy widening the range of products they offer to their clients These changes mainly implied an increasing share of non-interest income in profits Non-interest income stems from traditional service charges (checking, cash management, letters of credit .) but also from new sources With the decline in interest margins induced by higher competition banks were incited to charge higher fees on existing or new services (cash withdrawal, bank account management, data processing .) As a result, structure of bank income experienced a dramatic change in the world In the eighties, non-interest income represented 19% of US commercial banks’ total income This share had grown to 43% of total income in 2001 (Stiroh, 2004) and increased from 10% in 1984 to 31% in 2011 In Europe, non-interest income has increased from 26% to 41% between 1989 and 1998 (European Central Bank, 2000) In Vietnam, non-interest income represented 20% of commercial banks’

total income and its share had increased 9% for period of 2008-2011 (State Bank

of Vietnam, 2012)

With the adoption of the new universal banking principle, commercial banks can compete on a wider range of market segments (investment banking, market trading ) Numerous studies questioned the implications of this new environment on bank risk The issue is of importance for the safety and soundness of the banking system and a major challenge for supervisory authorities

The existing literature, mostly based on US banks, either focused on portfolio diversification effects risk return profile (Boyd et al., 1980; DeYoung & Roland, 2001; Kwan, 1998) or on incentives approaches (Boyd et al., 1998; John et al., 1994; Puri, 1996; Rajan, 1991) Few studies were able to show that the combination of lending and non-interest income activities allows for diversification benefits and therefore risk reduction Conversely, some papers find a significant positive impact of diversification on earnings volatility (DeYoung & Roland, 2001; Stiroh, 2004; Stiroh & Rumble, 2006) As noted by DeYoung & Roland (2001), three main reasons may explain this increase in risk Firstly, income from lending activities is likely to be relatively stable over time because switching and information costs make it costly for either borrowers or lenders to walk away from a lending relationship In contrast, income from non-interest income activities may suffer from larger fluctuations as it might be easier

to switch banks for this type of activities than for lending Secondly, expanding non-interest income activities may imply a rise in fixed costs (for example, additional staff may be required), which increases the operational leverage of banks Conversely, once a lending relationship is established, the marginal cost

induced by the supply of additional loans is limited to interest expenses Thirdly, because bank regulators do not require banks to hold capital against non-interest income activities, earnings volatility may increase because of a higher degree of financial leverage Moreover, as mentioned by Stiroh (2004), cross-selling of different products to a core customer does not imply diversification benefits (more products are sold to the same customer) which may explain why interest income growth and non-interest income growth are highly correlated in his study

1.2 RESEARCH PROBLEM

An increasing number of papers based on the US context have focus on the effect of earning diversifications of banks on earning volatility (DeYoung & Roland, 2001; Stiroh, 2004; Stiroh & Rumble, 2006) DeYoung and Roland (2001) concludes that fee-based activities increase the volatility of banks revenue

in USA Lepetit et al (2008) find that European banks expanding into interest income activities present higher risk and higher insolvency risk than banks mainly supply loans However, no paper has tried to study the effect of the diversification of Vietnam commercial banks’ earning on risks This paper aims

non-to fill the gap of lacking the empirical analysis of commercial banks in Vietnam

1.3 RESEARCH OBJECTIVE

The aim of this paper is to assess the risk implications of the changing structure

of the Vietnam banking industry which has shifted away from traditionalactivities (deposit funded loans) towards activities generating non-interest income Using individual bank data from 2008 to 2011 for 35 Vietnam commercial banks, we begin by analyzing the link between bank risk and the degree of output diversification measured by three indicators, the income share

of: (i) non-interest income, (ii) trading income and (iii) commission and fee income We hence start by comparing the risk level of banks which haveexpanded into nontraditional activities with the risk level of banks which have invested less in those activities Our results show that higher reliance on non-interest activities is not associated with higher risk, however, higher risk is correlated with commission and fee income for small banks with total assets smaller than 25.500 billion VND (equivalent to 1 billion €)1 Conversely, for banks with a larger share in trading income are not associated with both of risk exposure and insolvency (default) risk

1.4 RESEARCH METHODOLOGY AND SCOPE

The subject of this research is all listed and non-listed banks in Vietnam for the period from 2008 to 2011 The sample size is 35 This paper uses quantitative research based on Lepetit et al (2008) model to investigate the link between bank risk and the degree of output diversification We use various data analysis methods in conducting the research such as descriptive statistics, T-test, correlation test, and OLS regression with Eviews 7 for Windows

1.5 STRUCTURE OF RESEARCH

Section 2 reviews the basic theoretical background and prior work of previous researches Chapter 3 describes model for this paper, data collection and analysis methodology Chapter 4 contains the result and discussion of empirical tests while Chapter 5 concludes

1

Our criterion for distinguishing large and small banks is similar to Bank scope’s and is frequently used in the literature to categorize banks (Carter & McNulty, 2005)

CHAPTER 2: BASIC THEORETICAL BACKGROUND

AND LITERATURE REVIEW

Chapter 2 is to review the basic theoretical background and literature on bank risk, bank income, diversification and shows the relationship between bank risk and diversification

2.1 BANK RISK

There are multiple definitions of risk Banks face several types of risk All the following are examples of the various risks banks encounter:

 Borrowers may submit payments late or fail altogether to make payments

 Depositors may demand the return of their money at a faster rate than the bank has reserved

 Market interest rates may change and hurt the value of a bank’s loans

 Investments made by the bank in securities or private companies may lose value

 Human input errors or fraud in computer systems can lead to losses

To monitor, manage, and measure these risks, banks are actively engaged in risk management In a bank, the risk management function contributes to the management of the risks that a bank faces by continuously measuring the risk of its current portfolio of assets and other exposures

From a regulatory perspective, the size and risk of a bank’s assets are the most important determinants of how much regulatory reserve capital the bank is required to hold A bank with high-risk assets faces the possibility that those

assets could quickly lose value If the market -depositors- perceives that the bank

is unstable and deposits are in peril, then nervous depositors may withdraw their funds from the bank If too many depositors want to withdraw their funds at the same time, then fear that the bank will run out of money could break out In addition, when there is a widespread withdrawal of money from a bank, the bank may be forced to sell its assets under pressure To avoid this, regulators would want a bank with high-risk assets to have more reserves available (Apostolik et al., 2011)

There are many methods to measure bank risk; in this paper, some main measurements are presented

According to Lepetit et al (2008), there are three standard measures of risk, based on annual accounting data and determined for each bank throughout the period: (i) the standard deviation of the return on average assets (SDROA); (ii) the standard deviation of the return on average equity (SDROE); (iii) the ratio of loan loss provisions to net loans (LLP) And he also computes insolvency risk measures: (i) the‘‘Z-score” (ADZ) (Boyd & Graham, 1986) which indicates the probability of failure of a given bank; (ii) the ‘‘ZP-score” (ADZP) as in (Goyeau

& Tarazi, 1992) and its two additive components which we call ADZP1 and ADZP2 ADZP1 is a measure of bank portfolio risk whereas ADZP2 is a measure of leverage risk Whereas, ADZ = (100 + average ROE)/SDROE and ADZP = ADZP1 + ADZP2 = average ROA/SDROA + average (Total equities/Total assets)/ SDROA

Following his study, a large number of reseaches have focused on it Among of them, Barry et al (2009), Grassa (2009), Dhouibi & Mamoghli (2009) are few well-known studies

2.2 BANK INCOME

There are many papers researched on structure of bank income such as Davis & Tuori (1998), European Central Bank ( 2000), Stiroh (2004), Lepetit et al.(2008), Orlik & Blancas (2012) With Davis and Tuori (1998), bank income has two sources which are interest income from traditional activities and non-interest income from new activities Interest income is the difference between the interest which the bank has to pay for customers’ deposits and interest which the bank earns from lending money to customers While non-interest income is calculated as the sum of net fees and commissions (fees and commissions receivable less fees and commissions payable), income from securities and the net profit (loss) on financial operations and other operating income (European Central Bank, 2000)

In the paper of Lepetit et al (2008), the author considers bank income with ratio

of net interest income or net non-interest income to total operating income In addition, non-interest income is distinguished two components:commission and fee income and trading income Whereas, trading income that the author considers is total of income from securities, the net profit (loss) on financial operations and other operating income

2.3 DIVERSIFICATION

In the paper of Stiroh (2004), the U.S banking industry is steadily shifting away from traditional sources of revenue like loan making toward nontraditional

activities that generate fee income, service charges, trading revenue, and other types of non-interest income Non-interest income has always played an important role in banking revenue This shift toward non-interest income has contributed to higher levels of bank revenue in recent years, but there is also a sense that it can lower the volatility of bank profit and revenue, and reduce risk One potential channel is that non-interest income may be less dependent on overall business conditions than traditional interest income, so that an increased reliance on non-interest income reduces the cyclical variation in bank profits and revenue Alternatively, expanded product lines and cross-selling opportunities associated with growing non-interest income may offer diversification benefits for a bank’s revenue portfolio

Kwast (1989) finds limited diversification benefits from expanded bank securities powers from 1976 to 1985 Similarly, Kwan (1998) reports that bank Section 20 subsidiaries typically posted more volatile accounting returns, although not necessarily higher returns DeYoung and Roland (2001) examine the link between bank profitability, volatility, and different revenue shares for

472 large commercial banks from 1988 to 1995 They conclude that increased fee-based activities (revenue from all sources except loans, investment, deposit, and trading activities) increases the volatility of bank revenue and bank earnings Taken together, there is little evidence of large diversification benefits from these papers

One way to capture the degree of diversification of bank activities in the literature (Stiroh, 2004) is to consider the structure of income statements that is the shares of net interest income generated by traditional activities and non-interest income produced by nontraditional activities

With Lepetit et al (2008), he considers bank diversification as the changing structure of the banking industry that has shifted away from traditional activities (deposit funded loans) towards activities generating non-interest income He starts by analyzing the link between bank risk and the degree of output diversification measured by three indicators, the income share of: (i) non-interest income, (ii) trading income and (iii) commission and fee income They split their samples into different panels of banks based on the value of the ratio of net non-interest income to net operating income (NNII) In addition, they consider as diversified, banks for which the value of the NNII ratio is higher than the third quartile (Q75) and as non-diversified, banks with a NNII ratio lower than the first quartile (Q25) For deeper insights, they compare the level of risk of banks which are characterized by high levels of fee-based activities that is banks with a ratio of net commission income to net operating income (COM) higher than the third quartile Q75, with banks with the same ratio not exceeding the value of the first quartile (COM lower than Q25) In the next step, they undertake the same comparison based on the degree of reliance on trading activities (ratio of net trading income to net operating income (TRAD) higher than Q75 versus TRAD lower than Q25

2.4 BANK RISK AND DIVERSIFICATION

There is a variety of studies that analyze diversification of income sources, more specifically interest and non-interest income, has attracted increasing attention in academic research Generally, it is believed that diversification of income sources should reduce total risk, as diversification should stabilize operating income For instance, Boyd et al (1980), who simulated portfolios of banking and non-bank subsidiaries during the seventies, find a potential for risk reduction

at relatively low levels of non-bank activities The results obtained by Kwast (1989) to determine an optimal risk-minimizing combination of banking and nonbanking activities for the period 1976–1985 show only a slight potential for risk reduction Saunders & Walter (1994) perform a simulation exercise and conclude that there are potential gains in the reduction of risk from bank expansion into new activities They find that property and casualty insurance is a particularly attractive area for money center bank expansion Gallo et al (1996) find over the 1987–1994 period, that combining traditional banking activities and securities and/or insurance activities allows for some diversification benefits increasing profitability for moderated risk levels

Another strand of the literature reports an increase in risk when combining traditional and non-interest income activities According to Boyd & Graham (1986), expansion by bank holding companies (BHCs) into nonbank activities during the seventies tended to increase the risk of failure of banks during the less stringent policy period Demsetz & Strahan (1997) who study the stock returns

of BHCs between 1980 and 1993 find that although banks extended their product mixes, no risk reduction could be observed as banks tended to move to riskier activities and to lower their capital ratio Kwan (1998), who investigated bank section 20 subsidiaries during the 1990–1997 periods, underlines the increased volatility of accounting returns despite a non-increase in bank profitability DeYoung & Roland (2001) examine the link between bank profitability, volatility, and different revenue shares for 472 large commercial banks from

1988 to 1995 They conclude that increased fee-based activities (revenue from all sources except loans, investment, deposit, and trading activities) increases the volatility of bank revenue and bank earnings DeYoung & Roland (2001)

provide three reasons why non-interest income may increase volatility First, revenues from fee-based activities might be more volatile than interest income because the customer-bank relationship is stronger in the traditional lending business, i.e for many of the new fee-based activities it is easier for customers to switch to another bank Second, expanding into fee-based services can considerably increase fixed costs (e.g by investments in technology and human resources) whereas, if a lending relationship is already established, the only cost

of an additional loan are the bank’s interest expenses Third, in contrast to the lending business, fee-based activities require less regulatory capital, which suggests a higher degree of financial leverage and therefore leads to a higher earnings volatility

A similar result is obtained by Stiroh (2004)who assesses the potential benefit of diversification for US banks engaging in non-interest activities for the period 1984–2001 He finds that empirical evidence that reliance on non-interest activities increases the volatility of large U.S banks Lepetit et al (2008) find that European banks expanding into non-interest income activities present higher risk and higher insolvency risk than banks mainly supply loans

However, no paper has tried to study the effect of the diversification of Vietnam commercial banks’ earning on risks Hence, to our knowledge, this is the first study to explore the impact of the diversification of banks’ earning on risk in the case of Vietnam Firstly, this study considers a large set of risk and insolvency risk measures based on accounting data at the bank individual level Secondly, to assess the risk implications of non-interest generating activities are split into two components: trading activities and commission and fee activities Third, we

conduct a regression analysis which enables to capture the major changes in our period of study and we focus on risk implications both for large and small banks

2.5 HYPOTHESIS DEVELOPMENT

The literature cited above highlights, with regards to US, EU banks, that activity diversification does not necessarily imply lower risk, and may on the contrary increase bank risk As a first step, we check if similar results can also be obtained for Vietnam banks by simply conducting univariate mean tests We therefore go to test some hypothesis such as:

H1: “Risk/Insolvency risk is not different for high and low degree of diversification”

H2: ‘‘Bank characteristics are not different for high and low degree of diversification”

As the last step, we do multivariate regression analysis to investigate the issue on risk and diversification It whether supports the hypothesis as below:

H3: “Diversification has positive effect on bank risk”

H4: “Fee and commission income has positive effect on bank risk”

H5: “Trading income has not effect on bank risk”

H6: “Effectiveness of diversification on bank risk is not different for large and small banks”

CHAPTER 3: DATA AND RESEARCH METHODS

Based on the results in the literature review, this Chapter presents model and the variables for the model Furthermore, it also describes the data for this research

- M_DIVi is the mean value, for bank i, for the period 2008-2011, of each product diversification variable (NNII, COM and TRAD)

- M_Xhi is the mean value, for bank i, for the period 2008-2011, for a set of control variables Xh (ROA, EQUITY, LOANS and DTA)

3.2 VARIABLES

3.2.1 Diversification variables

Following the literature reviews, in our paper, we consider the structure of income statements that is the shares of net interest income and non-interest income produced We therefore define several variables Firstly, we consider the ratio of net non-interest income to net operating income NNII Net non-interest income is defined as the difference between non-interest income and non-interest expenses; net operating income is the sum of net interest income and net non-interest income Secondly, for more precisely, we distinguish two components of non-interest income: commission and fee income and trading income We hence define a ratio of net commission and fee income to net operating income (COM) and a ratio of net trading income to net operating income (TRAD) Net commission and fee income is equal to commission income minus commission and fee expense and net trading income is equal to trading income minus trading expense

3.2.2 Bank risk measures

In this study, we used three standard measures of risk, based on annual accounting data and determined for each bank throughout the period, are used in our study: (i) the standard deviation of the return on average assets (SDROA); (ii) the standard deviation of the return on average equity (SDROE); (iii) the ratio of loan loss provisions to net loans (LLP)

We also compute insolvency risk measures: (i) the ‘‘Z-score” (ADZ)2 which indicates the probability of failure of a given bank; (ii) the ‘‘ZP-score” (ADZP)3

as in Goyeau and Tarazi (1992) and its two additive components which we call ADZP1 and ADZP2 ADZP1 is a measure of bank portfolio risk whereas ADZP2 is a measure of leverage risk

3.2.3 Control variables

We chose large set of control variables which was initially considered to account for size differences (total assets (TA)), the annual growth rate of total assets (DTA) , profitability differences (ROA and ROE), business differences (deposits

to total assets (DEP)), loans to total assets (LOANS), personnel expenses to total assets (EXPENSES) and leverage differences (EQUITY)

Because of frequent collinearity among the variables both in the large sample and the smaller sample of banks (see Appendix A.2), control variables are restricted to ROA, EQUITY, LOANS and DTA In Appendix A.2, we used Spearman correlation test to consider degree of correlation of dependent and independent variables with the sample of all banks, the first cohort of small banks and the second cohort of large banks With high correlation coefficent (over from 0.7) of independent variables, we left these one out of the model In Appendix A.2, we saw that TA has high correlation with ROE, DTA, EQUITY; ROE has high correlation with TA, DTA; EXPENSES has high correlation with

LOANS; DEP has high correlation with EQUITY; NII has high correlation with NNII

3.3 DATA

We use a sample consisting of an unbalanced panel of annual report data from

2008 to 2011 for a set of 35 commercial banks in Vietnam (see Appendix A.1)

We collect data for our research from the annual reports of commercial banks in Vietnam Since the number of banks has been increasing every year, we select banks existing for at least three years (since 2009) Our sample is constituted by

of 35 banks observed during the period 2008-2011, whereas, 9 banks listed in Vietnam stock market (HOSE, HNX), 26 banks unlisted but in OTC market

Source of annual reports of commercial bank is on website: www.sanotc.com, www.cafef.vn Most of financial statements of banks in our research are audited

by famous auditing companies in the world such as: PricewaterhouseCoopers, KPMG, Ernst &Young Therefore, our data in this paper is reliable at a certain level

Table 1

Descriptive statistics for Vietnam commercial banks, on average over the period 2008-2011 (%)

LOANS DEP EQUITY LLP EXPENSES ROA ROE NII NNII COM TRAD TA

Sample : non-listed and listed banks (35 banks)

1) all variables are expressed in percentage except TA

2) Variable definitions: LOANS = loans/total assets; DEP = deposits/total assets; EQUITY = equity/total assets; LLP = loan loss provisions/net loans; EXPENSES = personnel expenses/total assets; ROA = return on average assets; ROE = return on average equity; NII = net interest income/net operating income; COM = net commission income/net operating income; TRAD = net trading income/net operating income; NNII = net non-interest income/net operating income; TA = total assets (million VND)

Based on the sample of 35 banks, we firstly consider two cohorts of listed banks and non-listed banks Descriptive statistics of our sample and two cohorts are presented in Table 1 Both of two cohorts show sufficient heterogeneity in different types of banking activities, enabling us to analyze the behavior of banks depending on their degree of product diversification

In the average, net interest income of all banks takes account for 78.56% of total net operating income Whereas, the ratio of net interest income on net operating income (NII) of non-listed banks is higher than of listed banks And of course,

(NNII=COM+TRAD) of non-listed banks is lower than of listed banks Listed banks with higher NII get higher total assets (TA) However, return on assets and personnel expenses for both of listed and non-listed banks are just the same

In another different point between two cohorts is the different leverage level (EQUITY) It is lower with listed banks which have higher NNII and higher with non-listed banks which have lower NNII With listed bank, loans supplying (LOANS) and deposits (DEP) are higher, but with lower net interest income (NII) because the interest expenses are higher from deposit activities And higher loan loss provision (LLP) of listed banks with higher LOANS is as appropriated regulation

LOANS DEP EQUITY LLP EXPENSES ROA ROE NII NNII COM TRAD TA

Sample of large banks: non-listed and listed banks (22 banks)

Small banks with TA <25.500 billion VND

Sample of small banks: non-listed and listed banks (13 banks)

1) all variables are expressed in percentage except TA

2) Variable definitions: LOANS = loans/total assets; DEP = deposits/total assets; EQUITY = equity/total assets; LLP = loan loss provisions/net loans; EXPENSES = personnel expenses/total assets; ROA = return on average assets; ROE = return on average equity; NII = net interest income/net operating income; COM = net commission income/net operating income; TRAD = net trading income/net operating income; NNII = net non-interest income/net operating income; TA = total assets (million VND)