(TIỂU LUẬN) TOPIC FACTORS AFFECTING NON PERFORMING LOANS OF COMMERCIAL BANKS IN VIETNAM IN THE PERIOD 2010 2019

FOREIGN TRADE UNIVERSITY
FACULTY OF INTERNATIONAL ECONOMICS
------------------------------------------------

ECONOMETRICS REPORT
TOPIC: FACTORS AFFECTING NON-PERFORMING LOANS
OF COMMERCIAL BANKS IN VIETNAM IN THE PERIOD
2010-2019
CLASS ID: KTEE310.1
Lecturer: Dr. Vũ Thị Phương Mai
Members:
1. Nguyễn Hà Vy - 1234567654323456
2014340215
1 2 3 4 5 6 7
2. Nguyễn Thị Phương Thảo - 2345432123456787654
2014340209
3. Đỗ Thị Thu Hương - 234567uy5456u
2014340204
1234567876543
4. Nguyễn Thị Thu Hằng -2014340213

Hanoi, 2021



TABLE OF CONTENTS
ABSTRACT……………………………………………………………………...…… 1
INTRODUCTION………………………………..……………………………..…… 2
1.
2.
3.

4.
5.

Rationale of the study……………………..………………………..……......… 2
Research methodology…...………………..………………………..………..… 3
Goals and Purpose…...………………..………………………..……...…...…... 3
Research subject and scope…………..………………………....….…...……… 3
The structure of report…………..………………….……....….…….....……… 3

SECTION 1: OVERVIEW OF THE TOPIC ………….…………………………… 5
1.1. Overview about commercial bank………….………………………...……… 5
1.2. Overview about non-performing loans………….…………………………… 6
1.3. The determinants affecting non-performing loans………….……………… 8
1.3.1. Bank-specific determinants………….……..…………………….......……… 8
1.3.2. Macroeconomics factors………….………………….............……..……… 11
SECTION 2: MODEL SPECIFICATION…….....………………………...……… 15
2.1. Methodology in the study………….…….…………...…..………...……… 15
2.2. Methodology to collect and analyze the data………...……………........… 15
2.2.1. Collect the data………….…...........……………...……………….………… 15
2.2.2. Analyze the data………….…...........……………...……………………….… 15
2.3. Building the research model…………..…….……………………….……… 20
2.3.1. Population Regression Model………….…...………………...…….……… 20
2.3.2 Sample Regression Model………….………….………….....………...…… 21
2.4. Description of the data………….………………………...................……… 21
2.4.1. Statistical description of the variables ………….………..…………......… 21
2.4.2. Correlation matrix between variables……………………………...……… 23
SECTION 3: ESTIMATED RESULTS AND STATISTICAL INFERENCES.... 25
3.1. OLS regression and conclude the model……..….……………………….… 25
3.1.1. Testing the significance of an
individual regression coefficient…………………………………………..……...... 25

3.1.2. Testing specification error………..………………..……...………………… 26
3.1.3. Sample regression model and explain the results …….……………..…… 27
3.1.3.1. Explain general results…………………………………..…….. 27
3.3.3.2. Testing the significance of an individual
regression coefficient and explain the estimated coefficient……...……. 28


3.2. Testing the model’s defect……………………………………………….…… 29
3.2.1 Testing multicollinearity error………………………………………….……. 29
3.2.2 Testing heteroskedasticity error……………………………………….…….. 30
3.2.3 Testing autocorrelation………………………………………………….……. 30
3.2.4 Testing disturbance’s distribution………………………………………..….. 31
3.3 Comparison to the literature and interpretation of result………………… 32
3.3.1 Bank Specific Determinants………………..……………………………..….. 32
3.3.2 Macroeconomic determinants………………..…………………………..….. 33
SECTION 4. CONCLUSION AND RECOMMENDATION…………………….. 34
4.1. Conclusion…………………………………………………………………….. 34
4.2. Recommendation……………………………………………………………... 34
4.2.1 Recommendations for Vietnamese Commercial Banks ……………………. 34
4.2.2 Recommendations for Governments and the State Bank of Vietnam……. 36
REFERENCES…………………………………………………………………..…. 38
APPENDIX………………………………………………………………………..… 41
1. List of commercial banks …..………………………………………………….. 41


TABLE OF FIGURES
Exhibit 1.1: Research model......................................................................................... 8
Exhibit 2.1: The distribution of some variables........................................................... 16
Exhibit 2.2: The relationship between and the dependent variable (NPL) and some
independent variables................................................................................................... 17

Exhibit 2.3: Statistical description of the variables...................................................... 22
Exhibit 2.4: Correlation matrix between variables....................................................... 23
Exhibit 3.1: OLS regression results.............................................................................. 25
Exhibit 3.2: Testing specification error........................................................................ 27
Exhibit 3.3:Testing multicollinearity.................................................................................. 30
Exhibit 3.4:Testing heteroskedasticity.............................................................................. 30
Exhibit 3.5: Testing disturbance’s distribution.................................................................. 31

TABLE OF TABLES
Table 2_1: Variable description................................................................................... 18
Table 3_1: Testing significance of variables................................................................ 25
Table 3_2: Testing significance of variables of new model and explain the estimated
coefficient..................................................................................................................... 28
Table 3_3: Hypothesis conclusion................................................................................ 41


ABSTRACT
The term “Non-performing Loans” (NPLs) has become more common . Since
the 2008 global crisis, NPLs have been monitored worldwide and became systemically
important. NPLs are important variables on a macro scale for the financial stability as
well as microscale for banks profitability itself in Vietnam.In the research, we aim to
examine the determinants of Non-performing Loans (NPLs) in the Vietnamese banking
system by using regression models. To address the research problem, data of
commercial banks in Vietnam from 2010 to 2019 were collected. Macroeconomics
factors such as: GDP, Inflation,Unemployment rate and bank specific factors such as:
Return on equity, equity-to-asset ratio,credit growth, and previous non-performing loans
are studied to analyze the effect on NPLs in 30 commercial banks in Vietnam.
Factors were observed and estimated by quantitative method Ordinary Least Square in
order to declare the relationship between them and the rate of changes in NPLs. The
results show that non-performing loans this year will positively affect those in the next

year. In addition, a raise in average return on equity and credit growth also leads to the
reduction in non-performing loans from banks. Regarding macroeconomics factors, the
result implies that GDP is significant and negatively affects NPLs, meanwhile inflation
has positive effects related to NPLs. However, inflation resulted in an insignificant
relationship with NPLs contrary to our expectation. The result of this research is useful
to assist financial institutions and the regulators for policy formulation so as to minimize
the negative effects of NPLs to the Vietnamese banking system.

1


INTRODUCTION
1. Rationale of the study
A sound financial system is crucial for every economy since financial
institutions, especially commercial banks, not only facilitate the credit flow in the
economy but also promote the productivity of business units via funding
investment.Therefore, it is difficult for a country’s economy to develop sustainably if
its financial system is inefficient and unstable
During past decades, studies have shown that most banking failures or crises are
caused by nonperforming loans (NPL) (Brownbridge, 1998) As the main operations of
commercial banks are to accept deposits and provide loans, they are exposed to the credit
risk of having bad loans, which are known as NPL. NPLs is considered as blood clot of
the economy which prevents the development of the economy . Bad debts can affect
liquidity risks, reduce operating profits, and the bank’s reputation with customers. As
the increase in NPL has been found to put banks in danger of bankruptcy and financial
crises in both developing and developed countries explained by Barr and Siems (1944)
and Khemraj and Pasha ( 2009)
NPL are also claimed as one of the main reasons causing a significant decrease
in the Vietnamese banks’ profitability during Vietnam’s economic slowdown in 2012
when the ratio of NPLs in Vietnam sharply increased, appearing in most of the

commercial banks’ announcements. The annual NPLs growth rate had sharply increased
from 2007 to 2012, with an average growth rate of approximately 43.11% per year.
Therefore, this study is conducted to explore the reasons behind these NPL.Currently,
recent research on the problems of NPLs in Vietnam was not easy to find out,
particularly quantitative research.
There was a little quantitative research which used an econometric model to find out the
main factors (particularly the endogenous factors) that influence the rate of changes in
NPLs in Vietnamese commercial banks.
Therefore , it is significantly necessary to make the studies on the factors
affecting the capital structure and propose the recommendations, to solve the nonperforming loan of commercial banks in Vietnam. This is the rationale for the author to
choose the research topic:Factors affecting non-performing loans of commercial
banks in Vietnam in the period 2010-2019. We also expect that the research can
contribute a part into the financial management of commercial banks to help
policymakers and bank administrators to devise policies of minimizing risks, limiting
non-performing debts, as well as provide appropriate solutions for commercial banks,
thereby contributing to raise efficiency of banking operations.

2


2. Research methodology
To conduct the study, we used statistical,comparative and regression methods
from annual financial data of 30 Vietnamese commercial banks in the period of 20102019. Accordingly, the method applied mainly is the regression analysis that runs the
econometric model to examine and investigate the correlation between the independent
variables and dependent variables. In specific, a multivariate regression model using
OLS method (Ordinary Least Squares) was employed to test the factors and its impacts
on the non-performing loans ratio of commercial banks in Vietnam.
3. Goals and Purpose
The main objective of this study is to detect and discuss the factors that determine
the rise or fall of the rate of NPLs in commercial banks; specifically, analyzing the

sensitivity of NPL to the macroeconomic factors and bank variables within 10 years.
Based on the results obtained , it also aims to enable policymakers and bank managers
to develop risk-reduction policies and solutions which will help reduce bad debt and
improve banking efficiency.
4. Research subject and scope
Research subject: Determinants affecting the ratio of NPLs in Vietnamese
commercial banks in the period 2010-2019
In terms of Scope: Employing data of 30 commercial banks in Vietnam. Banks are
selected based on size diverse and having continuous operate from 2010 to 2019
In terms of Time: Collect data in the period of 10 recent years from 2010 to 2019
.
5. The Structure of report







The structure of study is organized into three main section as followings:
Section 1: Overview of the topic
The first section aims to provide literature reviews related to in the research topics
including definition, classifications, economic theories and research hypothesis
used in the study.
Section 2: Model Specification
In this section, we present research methodologies to collect and analyze data.
Section 3: Estimated results and statistical references
This part is to demonstrate the results of estimated model, tests for the model's
possible problems and correct them.
Section 4: Conclusion and Recommendations

3


Based on findings and results in the Section 3, give conclusions and some
recommendations.

4


SECTION 1: OVERVIEW OF THE TOPIC
1.1.

Overview about commercial bank

● The term commercial bank refers to a financial institution that accepts deposits,
offers checking account services, makes various loans, and offers basic financial
products like certificates of deposit (CDs) and savings accounts to individuals
and small businesses. A commercial bank is where most people do their banking.
● How commercial banks work:
○ Commercial banks provide basic banking services and products to the
general public, both individual consumers and small to mid-sized
businesses. These services include checking and savings accounts, loans
and mortgages, basic investment services such as CDs, as well as other
services such as safe deposit boxes.
○ Banks make money from service charges and fees. These fees vary based
on the products, ranging from account fees (monthly maintenance charges,
minimum balance fees, overdraft fees, non-sufficient funds (NSF)
charges), safe deposit box fees, and late fees. Many loan products also
contain fees in addition to interest charges.

○ Banks also earn money from interest they earn by lending out money to
other clients. The funds they lend comes from customer deposits.
However, the interest rate paid by the bank on the money they borrow is
less than the rate charged on the money they lend. For instance, a bank
may offer savings account customers an annual interest rate of 0.25%,
while charging mortgage clients 4.75% in interest annually.
● The primary functions of commercial banks are:
○ Commercial banks accept various types of deposits from the public,
especially from its clients, including saving account deposits, recurring
account deposits, and fixed deposits. These deposits are returned
whenever the customer demands them or after a certain time period.
○ Commercial banks provide loans and advances of various forms,
including an overdraft facility, cash credit, bill discounting, money at call,
etc. They also give demand and term loans to all types of clients against
proper security. They also act as trustees for the wills of their customers
etc.
○ The function of credit creation is generated on the basis of credit and
payment intermediary. Commercial banks use the deposits they absorb to
make loans. On the basis of check circulation and transfer settlement, the
loans are converted into derivative deposits. To a certain extent, the
derivative funds of several times the original deposits are increased, which
5


greatly improves the driving force of commercial banks to serve the
economic development.
1.2.

Overview about non-performing loans

Non-performing loans are, in a way, a natural result of the banking system and
the whole of the financial sector’s operation. Financial institutions assume risks, some
of which materialize, turning loans into non-performing loans.
● Definition
According to Bank for International Settlements (2004), Non-performing loans
(or Bad Debts, Doubtful Debt) are loans that customers do not meet the bank's ability
to repay debts for more than 90 days. According to Circular No. 02/2013/TT-NHNN
dated January 21, 2013, bad debts are debts of group 3 ( non-standard loans ), group 4
(doubtful loans) and group 5 (loss-making loans). Therefore, the bank bad debt data with
groups 3, 4,and 5 were used in this research.
● Reasons for Non-performing loans
a. Reduced attention to borrowers
This is related to the Hawthorne effect. Researchers at Hawthorne Electric
Company in the US in the 1920s wondered what effect changes in lighting, heating and
similar variables would have on factory workers. To the researcher's amazement,
productivity increased throughout the study, during which time lighting varied greatly
from normal to dim to brilliant and back, the heat was turned up and down, etc. The
puzzled researchers eventually concluded that the workers were responding positively
because they were the subjects of interest, not because of changes in their working
conditions. Workers' perceptions that someone is paying attention to them get better
results than perceptions of being ignored. Borrowers may also perform in this manner.
b. Lenders lack plans to deal with risk
Donor-funded credit programs are usually designed without a clear focus on risk.
In microfinance promotion there seems to be no clear vision of risk or no industry-wide
concern about means of addressing it, other than running a tight ship. The literature is
largely concerned with outreach, measured by number of borrowers, and covering
administrative costs. The jury is still out on micro-lender performance, which is
currently supported by a tidal wave of donor funds that lifts all but the most leaky of
ships. This inattention to risk may be called The Pollyanna Effect.
c. Borrowers probe a credit operation's weaknesses

Credit programs have no special claim to infallibility. A borrower may be
determined to repay on time but is unable to do so due to unforeseen circumstances. If
the lender does not follow up promptly with a query, the borrower will take note. She
6


may simply be grateful not to have been embarrassed. A second way in which borrowers
are tempted to probe a credit program's weaknesses is when some borrowers blatantly
refuse to pay on time or skillfully avoid payment. Borrower's probing of a lender's
weaknesses may be called the Jurassic Park Effect. The dinosaurs in this popular film
tested the structures and devices used to contain them within certain areas of Jurassic
Park and eventually gained control over the entire park to the dismay, discomfort and
eventual departure or demise of their human captors. In addition, the park's dinosaurs
became more aggressive after the developers lost control of dinosaur breeding as a result
of unexpected risks. This was captured by the remark of one actor that, “life finds a way.
d. Lack of good models
Another possibility is that lenders are simply not familiar with successful
examples of dealing with bad and doubtful debts. This is likely in transaction economics
in North and Central Asia where commercial banking is still something of a novelty
compared to banking in service to economic planning. It also occurs, as in Bangladesh
and Nepal, where state domination of the banking system has been accompanied by a
high tolerance of non-repayment associated with politicization of financial markets.
Legal recourse in these situations is remote, costly, and uncertain. This lack of credible
models can be called High Default Culture Effect.
● Non-performing loans issues in Vietnam
In Vietnam, bad debts also hit the economy and result in many consequences.
Under the impacts of the global crisis in 2007, the Vietnam economy has experienced a
gloomy period with the buildup of bad debts from 2010 up to now.
A backlog of old debts has not been handled efficiently, and now new difficulties
have also arisen. New difficulties relate to the NPL ratio in 2020 being at a low level,

but debt Group 4 and Group 5 were on strong increase. At Techcombank, the bank with
the lowest NPL ratio in the system in 2020, the total bad debt ratio decreased by 58%,
to VND 1,295 billion, and the NPL ratio was only 0.47%. However, in terms of absolute
value, Group 4 debt increased by 75%, reaching nearly VND 534 billion. NamABank is
also in the group of banks with a bad debt ratio below 1% after total bad debts decreased
by 44% to VND 744 billion, but Group 5 debt increased by 77% compared to the
beginning of the year, at nearly VND 468 billion. The bank's accrued interest will also
increase by 100% in 2020, at VND 2,632 billion.
Group 5 debt at BIDV increased by more than VND 5,000 billion, to VND 16,525
billion, equivalent to an increase of 46% compared to the beginning of the year. MB's
total bad debt was nearly VND 3,248 billion, of which Group 5 debt increased by 124%,
accounting for VND 1,384 billion. Another phenomenon is the remarkable debt at
Group 2 at the end of 2020 at some banks which increased dramatically. Specifically,
OCB increased by 118%, VIB by 76%, and Vietcombank by 70%. This is also a matter
7


of concern because if customers continue to be unable to repay their debts, these debts
will jump from Group 2 to bad debt groups.
1.3.

The determinants affecting non-performing loans

The literature identifies two sets of factors to explain the evolution of NPLs over
time. One group focuses on external events such as the overall macroeconomic
conditions, which are likely to affect the borrowers’ capacity to repay their loans, while
the second group, which looks more at the variability of NPLs across banks, attributes
the level of non-performing loans to bank-level factors. Empirical evidence, however,
finds support for both sets of factors.

Exhibit 1.1: Research model

1.3.1. Bank-specific determinants
● Previous NPL:
According to Circulars 02/2013/TT-NHNN:
NPL=(Debt group 3,4,5 /Total loans to customers)*100%
Financial analysts frequently use the NPL ratio to compare the quality of loan
portfolios among banks. They may view lenders with high NPL ratios as engaging in
higher-risk lending, which can lead to bank failures. Economists examine NPL ratios to
predict potential instability in financial markets. Investors can view NPL ratios to choose
where to invest their money; they can view banks with low NPL ratios as being lowerrisk investments than those with high ratios.
8


Bad debts arise significantly from the weakness in the debt collection process,
the disproportionate provision for unresolved bad debts, creating a burden to deal with
future debts. Besides, the bad debt backlog will lead to the situation that the balance
sheets of banks still account for a high percentage of bad debts, banks cannot lend, and
the tangible and intangible costs for bad debts are larger.
In terms of outstanding loans provided by banks, bad debt ratio is generally
considered to be the main factor affecting bad debt (Sinkey and Greenwalt, 1991;
Keeton, 1999; Salas and Saurina, 2002; Jimenez). and Saurina, 2006) In the study, the
author uses the bad debt ratio in the previous period as bad debt in year t-1.
Hypothesis H1: The bad debt ratio in the previous period has a positive effect on the
current bad debt ratio (Louzis et al, 2010; Salas and Saurina, 2002).
● Equity-to-asset ratio
The asset to equity ratio reveals the proportion of an entity’s assets that has been
funded by shareholders, answering one question: What percentage of a company's assets
do investors own?
Formula to calculate EAR:

Equity-to-Asset ratio= Net Worth / Total Assets
Therein lies the key to the equity-to-asset ratio, which is to determine what
percentage of a company's assets are owned by investors and not leveraged and therefore
could come under the control of debtholders (such as banks) in the event of bankruptcy.
A low ratio indicates that a business has been financed in a conservative manner,
with a large proportion of investor funding and a small amount of debt. A low ratio
should be the goal when cash flows are highly variable, since it is quite difficult to pay
off debt in this situation. A higher ratio is tolerable when a business has a long history
of consistent cash flows, and those cash flows are expected to continue into the future.
A high asset to equity ratio can indicate that a business can no longer access additional
debt financing, since lenders are unlikely to extend additional credit to an organization
in this position. Also, if a business has a high ratio, it is more susceptible to pricing
attacks by competitors, since it must maintain high prices in order to generate the cash
flow to pay for its debt.
Starting with the bank-level indicators, the estimations show that higher equityto-assets ratio leads to lower NPLs.
Hypothesis H2: There is a negative relationship between EAR and NPL ( Kerlin 2013)
● Return on Equity (ROE)
Return on Equity (ROE) is the measure of a company's annual return (divided by
the value of its total shareholders’ equity.
9


ROE= Net income (annual) / Shareholders Equity
Return on Equity is a profitability metric that is used to compare the profits
earned by a business to the value of its shareholders’ equity.
A higher ROE is always preferred as it indicates efficiency from the side of the
management in generating higher profits from the given amount of capital.
Most empirical studies have shown that bad debt and bank profitability have a
negative relationship such as research by Klein (2013), Ghosh (2015), Le and Mai
(2015), KT Nguyen and Dinh Dinh (2015). Indeed, a bank with high profitability will

have little incentive to engage in high-risk lending activities. In contrast, inefficient
banks will try to make a profit by providing substandard credit, so it is easier for these
banks to generate bad loans. This problem is also reasonable when the profits of
Vietnamese banks are mainly obtained from credit activities, so when profits are high,
the quality of loans of banks is good, capital and interest are fully recovered enough,
leading to low bad debt (KT Nguyen & Dinh, 2016).
Furthermore, for the period from 2003 to 2012, Abid, Ouertani, and ZouariGhorbel (2014) used a pair of panel data from sixteen Tunisian banking institutions to
assess non-performing loans, and concentrated on both bank-specific and
macroeconomic features. Experimental findings showed that ROE was found to be
negatively associated with NPLs.. Moreover Makri, Tsakanos and Bellas (2014) also
show that there is a negative relationship between ROE and NPL ratio in European
countries. This indicates that poor management leads to riskier behavior and poorer
business results.
Hypothesis H3: there is a negative relationship between ROE and NPL ratio (Abid,
Ouertani, and Zouari-Ghorbel (2014)
● Credit Growth
Credit growth is the rate of increase of this year's credit balance compared to the
previous year
Credit Growth(t)= Loan balance(t)- Loan balance (t-1)/ Loan balance (t1)
*t:the year studied
Credit growth represents the scale of capital supplied to the economy . This
indicator is used to compare the growth rate of credit balance over the years to assess
the bank's lending situation. Competition in lending market share causes banks to race
for credit growth, high credit growth often the consequences that follow when the
economy is in crisis, banks are at risk of bad loans leading to bad debt in the future.
Salas and Saurina (2002) studied Spanish banks and found that loan balance
growth is related to loan default. Weinberg (1995) suggests that bad debts increase with
an increase in credit.
10

In addition, the research results of Klein (2013), Do Quynh Anh and Nguyen Duc Hung
(2013) and Nguyen Thi Hong Vinh (2015) also share the same opinion.
Keeton (1999) uses data from 1982 to 1996 and a vector autoregression model to
analyse the impact of credit growth and loan delinquencies in the US. It reports evidence
of a strong relationship between credit growth and impaired assets. Specifically, Keeton
(1999) shows that rapid credit growth, which was associated with lower credit standards,
contributed to higher loan losses in certain states in the US. In this study loan
delinquency was defined as loans which are overdue for more than 90 days or does not
accrue interest.
Hypothesis H4: There is a positive relationship between the non-performing loan ratio
and the credit growth of PCFs.
1.3.2. Macroeconomics factors
● Gross Domestic Product (GDP):
Gross domestic product (GDP) is one of the most common indicators used to
track the health of a nation's economy. The calculation of a country's GDP takes into
consideration a number of different factors about that country's economy, including its
consumption and investment.
GDP= C+ I + G + (X-M)
C = Private Consumption ;G = Government Investment; X = Export s;
M = Imports; I = Gross Investment

The GDP of a country is calculated by adding the following figures together: personal
consumption; private investment; government spending; and exports (minus imports).
GDP can be expressed in two different ways—nominal GDP and real GDP.
● Nominal GDP takes current market prices into account without factoring in
inflation or deflation. Nominal GDP looks at the natural movement of prices and
tracks the gradual increase of an economy's value over time.
● On the contrary, real GDP factors in inflation. meaning it accounts for the overall
rise in price levels. Economists generally prefer using real GDP as a way to

compare a country's economic growth rate. Real GDP is how economists can tell
whether there is any real growth between one year and the next. It is calculated
using goods and services prices from a base year, rather than current prices, in
order to adjust for price changes.
GDP growth is perhaps the most closely watched and important economic
indicator for both economists and investors alike because it is a representation of the
total dollar value of all goods and services produced by an economy over a specific time
period. As a measurement, it is often described as being a calculation of the total size of
an economy. GDP growth is also a key factor in using the Taylor rule, which is a primary
method used by central bankers to evaluate economic health and set the target interest
rates in an economy
11


The literature reviews anonymously found that there is a negative relationship
between actual GDP and NPLs (Fofack, 2005; Jakubík & Reininger, 2014; Jiménez &
Saurina, 2006; Khemraj, Tarron & Pasha, 2009; Louzis, Vouldis, & Metaxas, 2012).
The relationship can be interpreted by an increase in the level of actual GDP growth to
positively influence the level of profits in line with the decrease in NPLs. In this
situation, the repayment capacity of the borrower is increased thus directly solving the
bad debts issues. Conversely, collapse in the economy; decrease or adverse growth of
GDP leads to unsettled debts.
Quite consistent with the theory, the results that we found show a significant and
negative relationship between the growth rate of GDP, and NPL . The improvement in
the real economy is generating a reduction in non-performing loan portfolios of
commercial banks.
Tanasković and Jandrić (2015) investigated NPL determinants for some selected
CEEC and SEE countries for the period of 2006-2013. They found a negative
relationship between increases in GDP and rise of the NPL ratio.
Hypothesis H5: There is a negative relationship between the growth rate of GDP, and

NPL
● Unemployment rate :
The unemployment rate is defined as the percentage of unemployed workers in
the total labor force.
The unemployment rate provides insights into the economy’s spare capacity and
unused resources. Unemployment tends to be cyclical and decreases when the economy
expands as companies contract more workers to meet growing demand. Unemployment
usually increases as economic activity slows.
Concerning the unemployment rate, we found a positive and significant
relationship with the ratio of non-performing loans at a level of 1%. In fact, unemployed
customers cannot meet their commitments and repay the loans which can increase the
level of non-performing loans
Vatansever and Hepşen (2015) investigated NPL and determined
macroeconomic and bank-level indicators from January 2007 to March 2013 for Turkey.
They found the unemployment rate has a positive effect on NPL
Messai and Jouini (2013) investigated three countries, namely Italy, Greece and
Spain for the period of 2004-2008 for a sample of 85 banks. By employing panel data

12


regression they found that NPL varies negatively with the growth rate of GDP and
positively with the unemployment rate and the real interest rate.
Babouček and Jančar (2005) analyzed the Czech banking between 1993-2006
and found a positive correlation of NPLs with the unemployment rate and consumer
price inflation.
Hypothesis H6: There is a positive relationship between unemployment rate and NPL.
● Inflation:
Inflation is the rate at which the value of a currency is falling and, consequently,
the general level of prices for goods and services is rising.The most commonly used

inflation indexes are the Consumer Price Index (CPI) and the Wholesale Price Index
(WPI).
Inflation=

𝑪𝑷𝑰 𝒕 − 𝑪𝑷𝑰(𝒕−𝟏)
𝑪𝑷𝑰(𝒕−𝟏)

Inflation aims to measure the overall impact of price changes for a diversified set
of products and services, and allows for a single value representation of the increase in
the price level of goods and services in an economy over a period of time.
To combat this, a country's appropriate monetary authority, like the central
bank, then takes the necessary measures to manage the supply of money and credit to
keep inflation within permissible limits and keep the economy running smoothly
The mechanisms of how this drives inflation can be classified into three types:
demand-pull inflation, cost-push inflation, and built-in inflation
Basically, inflation reduces the value of loans, so customers' ability to repay loans
will be more favorable,however inflation affects the real income of customers, causing
the real income of customers to decrease, wages have a slower growth rate, inflation
will increase the bad debt ratio.
A number of studies have discovered the positive relationship between inflation
and NPLs (Badar & Yasmin, 2013; Fofack, 2005; Nkusu, 2011; Skarica, 2014).
Badar & Yasmin (2013) have conducted a study of banking institutions in
Pakistan from the period of January 2002 until December 2011 measuring the long and
short run between non-performing loans and variables such as inflation, exchange rate,
interest rate, GDP and money supply. The ability of the borrowers to repay their debts
can be influenced by level of inflation and percentage of interest rate. For example, if
the country is facing high levels of inflation, the borrowers have difficulty making
repayment for their loans due to increasing cost of capital.
Fofack (2005) also cites that higher levels of inflation would result in costly
borrowing and finally affects the quality of loan diversification. This is supported by

Skarica (2014) who analyzed banking institutions in Europe and found that higher
13


inflation will reduce the income of households and incidentally influence the capability
of debtors.
Hypothesis H7: There is a positive relationship between inflation and NPL

14


SECTION 2: MODEL SPECIFICATION
2.1 Methodology in the study
The process using the research is called Multiple Linear Regression.This is a
linear approach to modeling the statistical relationship of a dependent variable on one
or more independent variables. Specifically, in our research, it is the statistically
dependent relationship of non-performing loans on the previous year’s non-performing
loans, GDP growth, Inflation rate, Unemployment rate, Equity- to-asset ratio, credit
growth and return on equity. We use OLS method (Ordinary Least Squares) to obtain
the necessary estimated results.
2.2 Methodology to collect and analyze the data
2.2.1 Collect the data
To investigate the effect of bank-specific factors and macroeconomic factors on
Non-performing loans, the research uses panel data of individual bank’s financial
statements from website Vietdata and Vietstock as well as macroeconomic indicators
from World Bank datasets. Data is based on annual frequency for 2010-2019, and
collected from 30 commercial banks in Vietnam. Thus, the study has 300 observations
(30 banks * 10 years = 300).
2.2.2 Analyze the data
Our group has used Stata and Excel to analyze the dataset and interpret the

correlation matrix between variables. After conducting some analysis by using
histogram and scatter command, we see that non-performing loans(dependent
variables), the previous year’s NPL, equity to asset ratio, credit growth and return to
asset are long right-tailed distributed, therefore the log-transformation will be used.
Moreover, the relationship between independent variables and dependent variables also
became clearer after log-transformation. Below are some samples:

15


Exhibit 2.1: The distribution of some variables

1, Before log-transformation

After log-transformation

non-performing loans NPL

ln( non-performing loans) (lnNPL)

After log-transformation

2, Before log-transformation

The previous year’s non-performing loans ln(The previous year’s non-performing loans)
pNPL
lnpNPL

3, Before log-transformation

After log-transformation

16


ln(Credit growth) lnCreG

Credit growth-Credit

Exhibit 2.2: The relationship between and the dependent variable(NPL) and some
independent variables
After log-transformation

1,Before log-transformation

non-performing loans vs the previous year’s ln(non-performing loans) vs ln(the previous
non performing loans
year’s non performing loans)
NPL vs pNPL
lnNPL vs lnpNPL
After log-transformation

2, Before log-transformation
17


non-performing loans vs equity-to-asset

ln(non-performing loans) vs ln(equity-toasset)
lnNPL vs lnEAR

NPL vs EAR

Table 2.1: Variable description
Variab
le

Variable description

U
ni
t

Abbre
viation

Dependent variable

Non perfo
rmin
g
loans

𝐿𝑛

𝑇𝑜𝑡𝑎𝑙 𝑏𝑎𝑑 𝑑𝑒𝑏𝑡𝑠𝑡

%

lnNPL

𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠

Independent variables

18

Expe
cted
sign

Previous studies


Previo
us
year’s
NPL

GDP
growt
h

Inflati
on

Unempl
oyment

𝑇𝑜𝑡𝑎𝑙 𝑏𝑎𝑑 𝑑𝑒𝑏𝑡𝑠(𝑡−1)

%

lnpNPL

+

Sinkey and Greenwalt,
(1991);
Keeton,
(1999); Salas and
Saurina, (2002); . and
Saurina, (2006)

𝐺𝐷𝑃𝑡 − 𝐺𝐷𝑃(𝑡−1)
𝐺𝐷𝑃(𝑡−1)

%

GDP

-

Fofack, 2005; Jakubík
& Reininger, 2014;
Jiménez & Saurina,
2006; Khemraj, Tarron
& Pasha, 2009; Louzis,
Vouldis, & Metaxas,
2012

𝐶𝑃𝐼𝑡 − 𝐶𝑃𝐼(𝑡−1)
𝐶𝑃𝐼(𝑡−1)

%

INF

+

Badar & Yasmin,
2013; Fofack, 2005;
Nkusu, 2011; Skarica,
2014

%

UNE

+

Vatansever
and
Hepşen (2015) Messai
and Jouini (2013)
Babouček and Jančar
(2005)

%

lnEAR

-

%

lnCreG

+

𝐿𝑛

𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠

Annual
unemployment rate

Equity
-toasset
ratio

𝐿𝑛

Credit
growt
h

𝐿𝑛

𝑁𝑒𝑡 𝑤𝑜𝑟𝑡ℎ

𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠

𝐿𝑜𝑎𝑛𝑡 − 𝐿𝑜𝑎𝑛(𝑡−1)
𝐿𝑜𝑎𝑛(𝑡−1)

19

Kerlin (2013), Makri,
Tsagkanos, and Bellas
(2014)

Klein (2013), Do
Quynh
Anh
and
Nguyen Duc Hung
(2013) and Nguyen Thi
Hong
Vinh
(2015);Salas
and
Saurina (2002)


Return
on
Equity

𝐿𝑛

𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒(𝑎𝑛𝑛𝑢𝑎𝑙)
𝑺𝒉𝒂𝒓𝒆𝒉𝒐𝒍𝒅𝒆𝒓𝒔 𝑬𝒒𝒖𝒊𝒕𝒚

%

lnROE

-

Klein (2013), Ghosh
(2015), Le and Mai
(2015), KT Nguyen
and Dinh Dinh (2015)
Makri, Tsakonas and
Bellas (2014)

2.3 Building the research model
Base on related public researches and economic theories, the model used in this
report is constructed to examine the effects of relevant factors on non-performing loans
of commercial banks including its ROE, EAR, Credit, pNPL and macroeconomic factors
like GDP, UNE, INF
NPL=f( pNPL, GDP, UNE, INF, Credit, ROE ,EAR)
To demonstrate the relationship between movie revenue and other factors, the
regression function can be constructed as below:
2.3.1 Population Regression Model
PRF:
𝑙𝑛𝑁𝑃𝐿𝑖𝑡 =𝛽1 + 𝛽2 𝑙𝑛𝑝𝑁𝑃𝐿𝑖𝑡 + 𝛽3 𝐺𝐷𝑃𝑖𝑡 + 𝛽4 𝐼𝑁𝐹𝑖𝑡 + 𝛽5 𝑈𝑁𝐸𝑖𝑡 + 𝛽6 𝑙𝑛𝐸𝐴𝑅𝑖𝑡 +
𝛽7 𝑙𝑛𝐶𝑟𝑒𝐺𝑖𝑡 + 𝛽8 𝑙𝑛𝑅𝑂𝐸𝑖𝑡 + 𝑢𝑖𝑡
Where: 𝛽1 : the intercept term of the model
𝛽2 : the regression coefficient of lnpNPL

𝛽3 : the regression coefficient of GDP
𝛽4 : the regression coefficient of INF
𝛽5 : the regression coefficient of UNE
𝛽6 : the regression coefficient of lnEAR
𝛽7 : the regression coefficient of lnCreG
20