Factors affecting the profitability of commercial banks in Vietnam from 2009 - 2018: Bachelor thesis of Banking and Finance / Nguyen Thi Minh Thu ; Vu Thi Hai Minh supervisor

MINISTRY OF EDUCATION AND TRAINING STATE BANK OF VIETNAM

BANKING UNIVERSITY HO CHI MINH CITY

NGUYEN THI MINH THU

FACTORS AFFECTING THE PROFITABILITY OF

COMMERCIAL BANKS IN VIETNAM

FROM 2009 – 2018

GRADUATE THESIS

MAJOR: FINANCE - BANKING

CODE: 52340201

SUPERVISOR

MSC. VU THI HAI MINH

HO CHI MINH CITY, 2020

MINISTRY OF EDUCATION AND TRAINING STATE BANK OF VIETNAM

BANKING UNIVERSITY HO CHI MINH CITY

NGUYEN THI MINH THU

FACTORS AFFECTING THE PROFITABILITY OF

COMMERCIAL BANKS IN VIETNAM

FROM 2009 – 2018

GRADUATE THESIS

MAJOR: FINANCE - BANKING

CODE: 52340201

SUPERVISOR

MSC. VU THI HAI MINH

HO CHI MINH CITY, 2020

i

ABSTRACT

The purpose of this thesis is to provide an analysis of factors affecting the

profitability of 28 commercial banks in Vietnam from 2009 to 2018 using Pooled

OLS, FEM, REM, FGLS methods. The scope of the study is to aim at finding and

analyzing the degree of impact of specific bank factors, industry-specific factors

and macro-economic factors on the profitability of 28 commercial banks in

Vietnam. The study cannot avoid the limitations because of the selection may not

fully reflect the content and the meaning of all factors affect profitability or may not

fully represent all commercial banks in Vietnam.

Understanding the factors influencing the performance of the bank not only

receives much attention from scholars, but also from shareholders, administrators

and authorities, because it assists policy makers and administrators in formulating

strategies and policies to ensure the stability and sustainability of banking system's

operation, as well as to avoid the risk of banking system crisis. Assessing the

profitability of a commercial bank always raises questions, because of their

diversity. Many researchers consider that profitability can provide a picture of the

bank's business results.

Obtained result showed that ROA of commercial banks in Vietnam was

affected positively and statistically significant by bank size, equity ratio, loan ratio,

deposit ratio, income diversification ratio, credit growth rate and GDP; while non￾performing loan, operating cost and exchange rate have negative impact on ROA. In

contrast, liquidity, credit to GDP and inflation do not affect ROA and are not

statistically significant.

Keywords: commercial banks, profitability, ROA, bank size, non￾performing loan, equity ratio, liquidity ratio, loan ratio, deposit ratio, income

diversification ratio, operating cost, credit to GDP, credit growth rate, GDP,

inflation, exchange rate.

ii

SUMMARY

International economic integration is an indispensable trend and an objective

requirement for any country in the current development stage. However, this

process, in addition to creating certain advantages and opportunities for the country

participating in the integration, also put these countries before significant

difficulties and challenges. The roadmap of international economic integration puts

enterprises of developing countries in general, the commercial banking system in

particular and the new business environment with fierce competition pressures,

unbalanced competitors. Currently, Vietnam is a member of the United Nations, the

World Trade Organization, the International Monetary Fund, the World Bank

Group, the Asian Development Bank and the Joint Forum Economic Cooperation

Asia-Pacific, ASEAN. Vietnam participates in multilateral free trade agreements

with ASEAN countries, South Korea, Japan and China. Vietnam has also signed

with Japan a bilateral economic partnership agreement. For banking and monetary

system, the integration process is associated with the process of liberalizing

financial market, bringing many opportunities, but also many challenges. From that

practice, this thesis will analyze and introduce factors affecting the efficiency in the

operation of Vietnam’s commercial banking system in the current integration

period, especially the factors affecting the profitability of commercial banks in

Vietnam, from which to give practical recommendations to improve the operational

efficiency of commercial banking system, is an urgent issue.

This thesis analyses factors affecting the profitability of 28 commercial

banks in Vietnam from 2009 to 2018 using regression methods such as Pooled OLS,

FEM, REM, FGLS; thereby provide some recommendations to improve the

profitability of commercial banks in Vietnam.

In order to further analyze the impact of Vietnam's macro environment and

banking industry on the profitability of Vietnamese commercial banks, the author

decided to consider three more independent variables, namely exchange rate, credit

iii

to GDP and credit growth rate along with two familiar macro factors which are

inflation and GDP growth.

The table below summarizes related previous studies and theoretical

framework of this thesis.

Table 0.0.1: Related previous studies and theoretical framework

Theory/Related previous

studies Subjects are mentioned Previous studies and

related law, decree

Commercial banks Definition of commercial

banks

- Law: Consolidated

documents No.

20/2018/VBHN-NHNN,

the State Bank of Vietnam

- Previous studies: Getter,

(2016), Tariq et al. (2014),

Tuyishime et al. (2015),

Kalpana & Rao (2017),

Mongid et al. (2012),

Erina & Lace (2013)

Profitability

Definition of bank

profitability

Wang & Wang (2015),

Bikker & Vervliet (2017),

Iacobelli (2017), Yuksel

et al. (2018), Muya &

Gathogo (2016), Niresh &

Velnampy (2014),

Sehrish, Irshad & Khalid

(2010), Pastory &

Marobhe (2015), Kohler

(2013)

Measurement

Book: Principals of

finance, Scott, B. &

Eugene, F. B. (2015)

Market power theory

It explains the profit of

firms to be increased from

the achievement of more

market power

Bain (1951)

Efficiency theory

It explains other aspect of

bank profitability which

can be generated from

better management and

business efficiency

Demsetz (1973)

Agency cost theory This theory focuses on the

use of financing structure Jensen & Meckling

iv

of the firms to drive the

managers and the

investors to solve free

cash flow issues

Signaling theory

This theory emphasizes

that the firms with

profitable condition and

high business

performance can supply

the market and the

customers with positive

information

Arrow (1972) and Spence

(1973)

Foreign studies

Almaqtari et al. (2018), Yao et al. (2018), Yuksel et al.

(2018), Kawshala & Panditharathna (2017), Nuhiu et

al. (2017), Pastory & Marobhe (2015), Sehrish Gul,

Faiza Irshad and Khalid Zaman (2011), Abuzar (2013),

Usman, D. (2014), Khaled, M. et.al (2016), Sufian, F.

& Noor-Mohamad-Noor, A. (2012), Al-Omar, H. &

Al-Mutairi, A. (2008)

National studies

Nguyen (2018), Nguyen et al. (2018), Le et al. (2017),

Nguyen, D. (2017), Phan & Phan (2014), Vo & Batten

(2014), Vu & Nahm (2013), Nguyen Thanh Phong

(2015), Phan Thi Hang Nga (2011)

Source: Author’s synthesis

The research model proposed specifically as follows:

Where: βo : constant term

β1, ..., β14: Regression coefficients for the independent variables

i: bank; t: year of observation; μi

: error term/ residual term

Dependent variables:

ROA: return on assets (%)

Independent variables (bank-specific factors):

SIZE: bank size (billion VND)

NPL: non-performing loan (%)

LIQ: liquidity ratio (%)

v

EQR: equity ratio (%)

LOAN: loan ratio (%)

DEP: deposits ratio (%)

IDR: income diversification ratio (%)

OPR: operating cost ratio (%)

Independent variables (industry-specific factors):

CRD: credit to GDP

CGR: credit growth rate

Independent variables (macro-economic specific factors):

INF: inflation

GDP: gross domestic product

ER: exchange rate between VND and USD

The thesis performs: (1) Descriptive statistics; (2) Correlation analysis; (3)

Model estimation (Pooled OLS, FEM, REM); (4) Model choice (Between Pooled

OLS and FEM, between FEM and REM); (5) Model diagnostics (Autocorrelation

diagnostics, multicollinearity diagnostics, heteroskedasticity diagnostics); (6) Model

fix; (7) The prediction of residuals; (8) Normality test of residuals.

The result of model choice analysis showed that both FEM and REM were

more appropriate than Pooled OLS. Furthermore, FEM was proven as more

appropriate than REM. However, the author identified that there was

autocorrelation issue in the dataset. The result of heteroskedasticity check revealed

that FEM model did face up with this issue. To fix the model problems, the author

conducted FGLS regression technique and here is the result:

vi

However, while some independent variables are in line with the expectations

of the author, there are some independent variables that contrary to the author's

initial expectations. Detailed is shown as below:

Table 0.0.2: Expectations and research results of independent variables

Code Description Expected

signs Results

SIZE Bank size + +

NPL Non-performing loan - -

LIQ Liquidity ratio - No impact

EQR Equity ratio + +

LOAN Loan ratio + +

DEP Deposit ratio + +

IDR Income diversification ratio + +

OPR Operating cost ratio - -

CRD Credit to GDP + No impact

CGR Credit growth rate + +

GDP Gross domestic product growth + +

INF Inflation rate - No impact

ER Exchange rate between VND and USD - -

(Source: Author’s synthesis)

Based on the research results, the author will provide some recommendations

in chapter 5 to improve commercial banks’ profitability, thereby helping policy

makers to easily control and optimize the profitability of commercial banks.

vii

DECLARATION

I hereby declare that the graduate thesis is my original work, written by me

under the guidance of Ms. Vu Thi Hai Minh and it has not been presented for

examination in anywhere else. The data and sources cited in the thesis have been

duly acknowledged.

I take full responsibility for my pledge.

Author

Nguyen Thi Minh Thu

viii

ACKNOWLEDGEMENT

It is my radiant sentiment to place on record my best regards, deepest sense

of gratitude to my supervisor at Banking University in Ho Chi Minh City - Ms. Vu

Thi Hai Minh, for her patience, motivation, enthusiasm, and immense knowledge.

She has wholeheartedly helped and guided me in the process of writing this thesis. I

choose this moment to acknowledge her contribution as she has been supportive of

my study and has actively provided me with enormous information to achieve the

goal.

Also, I would like to thank all of my friends, who always support and help

me through the study together with the inspiration they gave me.

Author

Nguyen Thi Minh Thu

ix

ABBREVIATION LIST

Abbreviation Meaning

FEM Fixed effect model

FGLS Feasible Generalized Least Squares

GDP Gross domestic product

INF Inflation

MARG Net financing margin

NIM Net Interest Margin

NOVX Non-interest expenses to average total assets

Pooled OLS Pooled Ordinary Least Squares

REM Random effect model

ROA Return on asset

ROCE Return on capital employed

ROE Return on equity

Std. Dev. Standard Deviation

VIF Variance Inflation Factors

x

LIST OF TABLES

Table 0.0.1: Related previous studies and theoretical framework ............................ iii

Table 0.0.2: Expectations and research results of independent variables................. vi

Table 2.1: Summary of studies on the impact of factors affecting bank profitability

...................................................................................................................................22

Table 2.2: Limitation of empirical research..............................................................25

Table 3.1: Summary of debt groups..........................................................................34

Table 3.2: Variables used in the model.....................................................................39

Table 4.1: Descriptive statistics ................................................................................46

Table 4.2: Pearson correlation...................................................................................49

Table 4.3: Pooled OLS output...................................................................................50

Table 4.4: FEM output ..............................................................................................52

Table 4.5: REM output..............................................................................................54

Table 4.6: Model choice between Pooled OLS and FEM.........................................56

Table 4.7: Model choice between FEM and REM....................................................56

Table 4.8: Autocorrelation diagnostics.....................................................................57

Table 4.9: Multicollinearity diagnostics ...................................................................57

Table 4.10: Heteroskedasticity diagnostics...............................................................58

Table 4.11: Cross-sectional time-series FGLS regression........................................58

Table 4.12: The prediction of residuals ....................................................................59

Table 4.13: Normality test of residuals.....................................................................60

Table 0.1: List of commercial banks selected for the research........................... xxxiii

LIST OF FIGURES

Figure 3.1: Research process.....................................................................................31

Figure 3.2: ROA trends of Vietnamese commercial banks during 2009-2018.........33

xi

LIST OF PICTURES

Picture 1: Descriptive statistics............................................................................... xlv

Picture 2: Correlation analysis............................................................................... xlvi

Picture 3: Pooled OLS output ............................................................................... xlvii

Picture 4: FEM output.......................................................................................... xlviii

Picture 5: REM output ........................................................................................... xlix

Picture 6: Model choice between Pooled OLS and FEM ...................................... xlix

Picture 7: Model choice between FEM and REM ...................................................... l

Picture 8: Autocorrelation diagnostics....................................................................... li

Picture 9: Collinearity diagnostics............................................................................. li

Picture 10: Heteroskedasticity diagnostics ............................................................... lii

Picture 11: FGLS output ........................................................................................... lii

Picture 12: The prediction of residuals..................................................................... lii

Picture 13: Normality test of residuals...................................................................... lii