Internal Credit Rating according to market segmentation at Agribank

1

2

INTRODUCTORY

3. Objects and scope of research.
3.1. Objects of research.
- The thesis research Internal Credit Rating operation according to
market segment to individual customers having credit relations to Agribank
in Vietnam.
3.2. Scope of research.
3.2.1. Space of research

1. Rationales
According to the International Treaty about the safety standard for
capital, Internal Credit Rating has become an efficient tool for commercial
banks to manage the risk in lending and control credit risk, especially for
Agribank.
According to Decision No. 214/QĐ-NHNN on 30th, January, 2011
of the Governor of The State Bank of Vietnam, Vietnam Bank for
Agriculture and Rural Development (Agribank) was transformed into Stateowned one-member Limited Company. As the main bank operating in the
field of agriculture, rural and farmers, credit is the major operations of the
bank: Credit surplus accounts for approximately 79% over total assets,
Income from credit activities takes up 70% total income. Therefore, the
risks in the business of Agribank tend to focus on credit operations.
According to Decision No 1197/QĐ-NHNo-XLRR, written on
October 18th, 2011, Agribank has been put into use, operate the Internal Credit
Rating consulted by Ernst & Young Company since 2007 and it has worked on
decreasing risks in credit operations. However, the applying and making this
Internal Credit Rating has some defects in content as: Customer classification;

rating criteria,…
From practical operation and research, Internal Credit Rating
system of Agribank, the Author choose the topic of “Internal Credit Rating
according to the market segment at Agribank” to make researches and
decrease shortages on both reality and theory.
2. Research objective
- General objective: Making the Internal Credit Rating Models
according to market segment.
- Specific objective: Analyzing and evaluating elements come from
customers; analyzing discrepancies about the influence level of elements
from come from customers from different regions; evaluating the
practicality of applying Internal Credit Rating at Agribank in order to create
the Internal Credit Rating Model according to market segment; suggesting
solutions to complete the Internal Credit Rating system at Agribank.

Research about market segment theories; practical research about
Internal Credit Rating according to market segments to enterprise
customers, household customers, individual customers, financial institution
customers,… of Agribank and some of the largest commercial banks in
Vietnam. On that basis, this offers the Credit Rating system to individual
customers of Agribank according to market segment.
3.2.2. Time of research
- Secondary data: The statistics system about credit and statistics in
relation to customers of Agribank system and other commercial banks in
Vietnam… through websites, statistics from General Statistics Office in 5
years since 2010 to 30/6/2015.
- Primary data: Using questionnaire which are sent to the individual
customers of Agribank in 48/63 provinces, cities in all regions of Vietnam.
The survey includes approximate 210 forms each market segments,
performed from 11/2014 to 4/2015.

4. Research Methodology
4.1. Researching Method.
4.1.1. Qualitative method.
The thesis uses the aggregation, analysis, statistics, comparison and
experting methods to clarify and to prove the ideas and evaluations. The
thesis also uses the logical methodology to point out opportunities and
challenges in Internal Credit Rating operation according to market segment.
4.1.2. Quantitative method.
Analyzing by running secondary data of Agribank and the
commercial banks system for over 5 years; secondary data are collected


3

4

directly from individual customers of Agribank with 1123 observes
completed over 1470 observes established.
Applying statistical model to test and applying econometric model
to suggest models for each of the market segment.
4.2. Research questions
(1) Which are the criteria for market segmentation? What are the
connections between market segment and creating Internal Credit Rating
criteria system? Influence Level?
(2) How to create the Internal Credit Rating system which adapts to
different market segments? What are the baremes to exactly evaluate
customer’s ability?
4.3. Research model
The author recommends this research model:

implement and complete the Internal Credit Rating operation at
Commercial Banks in Vietnam.
5.2. In terms of reality
The research results can be immediately applied in Internal Credit
Rating operation at Agribank; helping researchers and banking
administrators have a more detail look at a new methods in measuring and
evaluating criteria of Internal Credit Rating system approached. The results
can be used as a useful referenced statistics system for people who care
about Internal Credit Rating of Commercial banks.

Group of customer’s information with bank elements (independent
variable)
Impact by different targets

Group of Relationship between customer and bank criteria

Group of Relationship between customer and bank criteria (already
influenced on customer information with bank) (Independent variable)

Influence on different targets

Customer’s ability to repay debt (dependent variable)

5. Thesis contribution
5.1. In terms of theories
The thesis systemizes general theoretical about Internal Credit
Rating, criteria, elements which have impacts on the result of Internal
Credit Rating. The thesis also evaluates, measures system of criteria in
evaluating Internal Credit Rating according to market segment; other
shortages and the result of the thesis is the premise for other researchers to

CHAPTER I: THE OVERVIEW OF THE RESEARCH
1.1. Some Internal Credit Rating Models
The first part of the thesis has introduced 6 well-known Credit
Rating models now being used worldwide and in Vietnam. According to
that, the thesis has evaluated the advantages and disadvantages of these
models which include:
- The Grading model: is the model which is used popularly by Credit
Rating agencies worldwide such as: Moody’s; Standard and Poor; Fitch.
- The Scoring model of Altman: is a quantitative model according to
the modeling of relationships between variables thereby reflecting credit
quality and those elements which affects credit quality from the customers.
- The Logistic model: Is a mathematical regressive model in which
the variable (Y) is a dependent variable and all others are independent.
- The Personal Credit Scoring model of FICO: Is a means of credit
control attached to each individuals in some developed countries, assisting
credit organization to estimate credit risk
- The Personal Credit Scoring applied for retail banks in Vietnam
(2006).
- The Individual Credit Rating of Ernst & Young (E&Y).
1.2. Some scientific paper involving in Internal Credit Rating
The thesis has presented 5 other PhD thesis involving in Internal
Credit Rating Operations of Commercial Banks in Vietnam. These contents
are presented from 1.2.1 to 1.2.5. From those, author has brought ideas,


5

6

opinions and evaluation about the advantages and disadvantages of Internal
credit rating .
1.3. Some articles, researches involving in Credit Rating
The author has done some researches and represented 12 scientific
paper and published on world famous magazine involving in Credit Rating.
1.4. The gap between researches
On the base of researching different models, researches and articles
about Credit Rating, the author has discovered some problems to delivery
research on
Firstly, Credit Rating methods, making criteria system, bareme and ratings
in Commercial banks are different from each others.
Secondly, Internal Credit Rating system is not made for different kinds of
group of objects.
Thirdly, the criteria in Credit Rating system are often identified using the
experting method, thus the final results remain some shortages.
Fourthly, the group of individual customer is not completely focused.

That role specifically expresses in 4 contents: identifying customer’s
choice; customer’s credit demand orientation; early identifying the ability
of finishing duties of customer; predict, prevent and cope with credit risk.
2.1.3. Market segment basis
The thesis has presented criteria which are often used in market
segment. One of them is segmentation according to economic region. The
thesis has mentioned Circular No 193/UB-VP on 11/2/1963 which is about
doing research on economic region dividend. In different periods, for the
sake of economic develop, Vietnam economy has divided the economic
into different regions. At present, the General Statistic Office has made
surveys on the statistics of the economic and the population of 63 provinces
which are divided into 7 regions which are: (1) The Northern midlands and
mountainous area (including 14 provinces); (2) The Red River Delta

(including 11 provinces); (3) The North Central area (including 6
provinces); (4) The Central Coastal area (including 8 cities and provinces);
(5) The Central Highlands (including 5 provinces); (6) The South East
(including 6 provinces); (7) The Mekong River Delta (including 13
provinces).
The thesis has summarized some common features of these 7
regions and researches and statistics evaluation have been made to identify
the discrepancies of the impact of the criteria on customer debt repayment
ability using the model in Chapter IV and Chapter V according to market
segment of Agribank
2.1.4. Reality of Vietnamese economic region dividing
At present, in Vietnam, the economic regions are usually divided
by geographical dividend; resident qualification or by income.
2.2. Internal Credit Rating in commercial banks
2.2.1. The Internal Credit Rating definition
By doing researches on theories and reality about Credit Rating, on
the aspect of an researcher on Internal Credit Rating Operation of one
Commercial Bank, the author has brought the definition: Internal Credit
Rating is a risk measurement tool of Commercial Bank on the basis of
analyzing, evaluating, rating customers through the customers’ information.

CHAPTER II: THEORETICAL BASE AND THE REALITY OF
INTERNAL CREDIT RATING ACCORDING TO MARKET
SEGMENT OF COMMERCIAL BANKS IN VIETNAM
2.1. Market segments of commercial banks
2.1.1. Market of commercial banks
From researches about theory of the market, the author has
generalized the market of commercial banks is a place where customers are
guided and satisfied with their banking and financial services using.
2.1.2. Market segment for Commercial banks

On the researching basis involving in market segment, the author
has brought the definition of market segment for Commercial bank:
“Market segment for Commercial bank is the dividend of customers into
different smaller groups of customers who have the same routines or
culture or hobby; or who have the same potential credit ability, etc.
The thesis has analyzed and pointed out: market dividing plays an
important role in business in general and in risk management in specific.


7

8

2.2.2. Internal Credit Rating Method
At present, Commercial Banks usually use 2 Internal Credit Rating
Methods which are experting and econometric method. Especially, the
econometric method including 2 main models which are Logistic regressive
models and factor analysis model. The thesis also pointed out the
advantages and disadvantages of the 2 mentioned methods.
2.2.3. The role of risk management in credit operation at
Commercial Banks
The role of risk management in credit operation at Commercial
Banks specifically expresses such as Decrease and Prevent credit risk in the
bank; decrease the damage, help customer avoid risk in credit operation and
in finishing debt obligation therefore contribute to the stability and
development of the whole society.
2.2.4. The role of Internal Credit Rating in credit risk
management at Commercial Banks
The thesis has quoted the definition of bad debt from Basel
Committee about bank supervision (2005) which exaggerate the most

important operation in bad debt management which is increase solutions to
prevent bad debt increasing. The author has pointed out 8 solutions to
conduct to prevent credit risk, 4 of which has been applied in Internal
Credit Rating operation. It could be said to be one of the most important
solutions which help Commercial Banks classify customers and make the
efficient credit decision.
2.3. Internal Credit Rating according to market segment of
Commercial Bank
2.3.1. Internal Credit Rating according to market segment of
Commercial Bank Definition
On the definition basis about Internal Credit Rating of one
commercial bank at 2.2.1 and the definition about market segment at
2.1.2.2, the author has brought the definition that Internal Credit Rating
according to market segment is rating and evaluating the ability of repaying
debt of the credit customer of one commercial bank based on the criteria
system which suit for the customer’s features on each segment which are
suitable for the bank’s operation.

2.3.2. Internal Credit Rating according to market segment
Features
There are 3 features: (1) Internal Credit Rating system created
based on internal market segment of commercial bank; (2) Internal Credit
Rating system which has suitable criteria system for groups of customers in
the same segment; (3) different criteria system on different segment
2.3.3. The role of Internal Credit Rating according to credit
customer in limiting credit risk at commercial bank
The role of Internal Credit Rating according to market segment has
been analyzed in detail and presented in 3 main content: The role in credit
risk management; in customers management and in policy planning
2.4. The reality in Internal Credit Rating according to market segment

operation at commercial banks in Vietnam, experiences
2.4.1. The foundation and development of Internal Credit Rating
according to market segment in commercial banks in Vietnam
The foundation and development of Internal Credit Rating is
divided into 2 periods: The period of implementation of the centered and
planned economy (from 1952 to 1988) and the period of gradually open
economy converted to market mechanisms (from 1989 to date)
From 1989 to present, State bank of Vietnam has established many
rules to control and manage the bad debt situation at commercial bank.
Enterprise Credit Rating was first mentioned in Decision No. 57/2002/QDNHNN to solve the bad debt rapidly, make the finance healthier, reduce
risk for credit organization and enterprise and improve credit growth for the
economy.
2.4.2. The reality in Internal Credit Rating according to market
segment operation in some commercial banks in Vietnam.
The Internal Credit Rating system in Commercial Bank in Vietnam
nowadays is usually created for 4 groups of customers: enterprises;
individuals; individual businesses; financial institution. The author has
modeled the grading Internal Credit Rating of commercial bank in Vietnam
using this model:


9

10
Grading objectives segment

individual loan for
consumption customer

individual business

customer

collecting information

- Family/household
- The possibility of debt repayment
- Collateral

enterprise
customer

others

- BY INDUSTRY SEGMENT
- BY SCALE SEGMENT
- BY POSSESSION SEGMENT

Grading according to
criteria system

- Financial Criteria
- Non-financial Criteria

Summarize customer rating
Model 1.2: Internal Credit Rating according to segment model in commercial banks in Vietnam

On basis of researching reality of segmenting Internal Credit Rating of
some big commercial bank in Vietnam, the author has come to a
conclusion: Internal Credit Rating of commercial banks in Vietnam had
market segment. However, the segmenting is still simply and only focus on

enterprise customer.
2.4.3. Assessment of reality of Credit Rating according to
market segment operation at some commercial banks in Vietnam
By assessing the reality of Credit rating operation at commercial
banks in Vietnam, the author has achieved some results such as: Creating
index system, bareme which are in detail; dividing customers into the
ability of using and repaying loan; assessing customers with varied
perspectives
Besides, the thesis has confirmed that the Internal Credit Rating
operation in commercial banks in Vietnam has 4 basic limits and pointed
out the origin of each limit such as: (1) lack of unification in creating
Internal Credit Rating of commercial banks; (2) criteria system has not yet
created based on the influence level on loan use and repay ability of the
customer; (3) lack of local influence element; (4) the limit of applying IT in
credit risk prevention of commercial banks.
2.4.4. Experience for Agribank
Agribank need to complete the Internal Credit Rating system in 3
ways:
Firstly, it is necessary to create the market segment based on seriously
assessing different elements of local, environment, etc. which is the
foundation to create the Internal Credit Rating system.
Secondly, the criteria system created must be suitable for each segment
based on quantificating the criteria through applying economic
mathematics model.
Thirdly, it is necessary to synchronize the technology system to decrease
the sensation elements influencing credit rating markers.


11

12

CHAPTER III: THE REALITY OF INTERNAL CREDIT RATING
ACCORDING TO MARKET SEGMENT AT AGRIBANK
3.1. Agribank overview
3.1.1. The foundation of Agribank
From the foundation on March 26th , 1988 to October 10th, 1996,
Agribank had its name changed twice: According to Decree No. 53/HDBT
on March 26th, 1988, it was named Vietnam Bank for Agriculture, after
that, the name was Vietnam Bank for Agriculture and Rural Development
and Agribank in English. On January 31st, 2011, according to Decision No.
214/QD-NHNN, Agribank was transformed into one-member Ltd State
own 100% capital Company.
Model organizational structure of executive management of
Agribank is modeled as below:

3.1.2. Overview of Agribank operation
Until July 31st, 2015, the total assets of Agribank was over 797.959
billion VND, the total liabilities was over 742.473 billion VND; the total
authorized capital was 29.605 billion VND; the total loans was 607.242
billion VND; there was approximately 40.000 personnel. There was 2.300
branches and transaction rooms and 4 million customers including about
3.4 million individual customers which took up 85% and 0.6 million
individual business and enterprise customers which took up to 15%. The IT
system, scale, processing speed and total average transaction processing
speed is the greatest in commercial bank system of Vietnam. Agribank
business was flexible by distributing a variety range of products from
traditional to modern ones and there was 253 of those.
3.1.3. The business performance of Agribank
Until April 30th, 2014, Agribank was the leader of the commercial

bank system on the market share in terms of: capital raising, credit, card
issued which respectively took up to 16.6%, 17% and 20%.

MEMBER COUNCIL

CONTROL
BOARD

HUMAN
RESOURCES
COMMITEE

ASSETS
MANAGEM
ENT
COMMITTE

GENERAL DIRECTOR

MEMBER COUNCIL
ASSISTING
DEPARTMENT

RISK
MANAGE
MENT
COMMITT
EE

16.6%

Other state-owned
commercial banks

43.8
CHIEF ACCOUNTANT

CENTERS SYSTEM

JOINT-VENTURE BANKS,
COMPANIES

DEPUTY GENERAL MANAGER

SPECIFIC BOARDS SYSTEM

HEAD OF
INTERNAL
CONTROL BOARD

BUSINESS
UNITS

EXCHANGES, BRANCHES SYSTEM TYPE I,II

BRANCHES SYSTEM TYPE III, TRANSACTION OFFICES

Agribank

39.6%

Joint-stock
commercial banks

Chart 3.2: the capital raising market share of some commercial bank in
Vietnam until December 31st, 2014


13

17%

14

Agribank
NHTM quốc doanh

51%
32%

NHTM cổ phần

Chart 3.3: The credit market share of commercial bank system in
Vietnam until December 31st, 2014
By credit investment, Agribank has contributed to poverty
elimination, economic development and policy implication.
3.2. The reality Internal Credit Rating according to market segment
operation at Agribank
Creating and completing Internal Credit Rating system is one of the
biggest solution which will help Agribank exactly assess the ability of

capital using and debt repayment of the customers in order to effectively
prevent credit risk in the system.
3.2.1. The reality of credit market segment at Agribank
Currently, there are not any documents about credit market
segment. However, documents system involving to credit operation and
credit customers market segment analysis was clearly divided in each
guidance document involving to credit operation. Beside credit segment
customers below, Agribank also manages the credit operation on 10 regions
at Annex 23 attachment.
With the credit management dividend method of Agribank, its
market segment is quite similar with the way The General Statistics Office
divide economic regions which was presented at Chapter II
3.2.2. The reality of Internal Credit Rating according to market
segment operation of Agribank
Implementing regulations at Article 7 Decision No. 493/2005/QDNHNN on April 22nd, 2005 issued by State Bank, Agribank promulgated
Decision No. 1197/QD-NHNNo-XLRR which guided how to use and

operate the Internal Credit Rating system. The Internal Credit Rating
system of Agribank was created based on the segment of 5 groups of
customers: (1) Enterprises; (2) Individual; (3) Farm Household; (4)
Household Business; (5) Financial Institution.
With each of the groups above, Agribank has a distinct procedure
and criteria system.
The author has aggregated the Internal Credit Rating according to
customers segment procedure of Agribank which will be presented below:
Customer information
Register

Customer segment

Economic organization/
institution customer
Enter financial, non-financial
information

Individual customer

Enter personal data;
debt repayment ability;
business plan

Rating, classify
debts
Report
customer rating

Collateral
rating
Summarize and
make decision

The thesis has described the scoring stages, criteria set, scoring methods,
bareme of groups of customers in detail. Such as:
- The enterprise customers is scored by using the method of
assessing financial criteria and non-financial criteria, which were created


15

16

based on 34 branches divided into 10 groups. Each criteria set has 60
criteria including 14 financial criteria and 46 non-financial criteria with 7
stages.
- With individual customers, Agribank created a unique criteria set
to rate used for all the individual customers. The criteria sets include 20
criteria divided into 3 parts: personal data (12 criteria); the ability of debt
repayment (4 criteria); collateral information (4 criteria).
- With farmer customers, Agribank only created 01 criteria system
for scoring including 22 criteria
- With business units customers, Agribank created enormous criteria
sets and scoring methods including 55 criteria divided into 4 big parts
- With financial institution, there are 2 criteria sets: financial
criteria set and non-financial criteria set.
3.2.3. Some circumstances of scoring borrowers at Agribank
In the research range presented at Chapter I, the subject the author
chose to research was the individual consumer at Agribank. The author did
research some circumstances of individual consumer in detail at Agribank
from the beginning of 2012 to 2014. The result of research on the reality of
Internal Credit Rating at Agribank on individual customers show that there
are some cases: the credit rating was very good but there were still bad
debts and vice versa

3.3. Assessing the reality of Internal Credit Rating according to market
segment operation at Agribank
3.3.1. Achievements
The Internal Credit Rating according to market segment operation
at Agribank has achieved some results which are presented in the 6
following contents: Scoring model for Internal Credit Rating at Agribank
was created according to the instruction of The State Bank; Agribank has

created criteria to identify, measure, control risk and step by step increase
risk management according to the international standard; having
established policy documents about debt classifying and risk provisioning;
having segmented customers into 4 groups; credit rating system for
enterprise customers has created based on customer segments such as
branch, scale and form of business ownership.
3.3.2. Shortcoming & Causes
- Because of the large quantity of customers, there are quite many
difficulties in implement Internal Credit Rating documents.
- When scoring the Internal Credit Rating, some criteria are not
suitable and influenced by sensation
- The scoring system and the bareme are not clearly calculated, it is
because the criteria has not been assessed by experting method.
- Market segment in Internal Credit Rating is only clearly
implemented for enterprise customers.
From the below analysis shown, Agribank need to overcome
difficulties in order to improve the Internal Credit Rating system.

Customer rating

Total
customers
at Bank A

Đã từng có
NQH trong 12
tháng

%overdue debt
customer/total

customers

AAA
AA
A
BBB
BB
B
CCC
CC
C
D
Total

324
2.134
11.000
2.215
439
165
322
174
453
132
17.358

4
40
1.240
1.269

272
62
21
0
33
69
3010

1.23
1.87
11.27
57.29
61.96
37.58
6.52
0.00
7.28
52.27
17.34

CHAPTER IV: RESEARCHING, MAKING SURVEY AND DATA
PROCESSING PROCEDURE
4.1. Researching procedure
Chapter IV presented researching procedure including:
- Describe how to create scale;
- Set up questionnaire;
- Preliminary survey, primary survey
- Data analysis.

17

4.1.1. The Qualitative analysis result
Qualitative analysis includes
- Making the preliminary scale: preliminary survey scale was made based
on aggregating information from bank; general researches in Chapter I;
criteria system presented in individual credit rating system of commercial
banks in Chapter II. Researching questions used questionnaire from
customers.
- Expert interview: to reassess the suitability of surveyed criteria. The
experts suggest the necessity of the criteria as well as adding or removing
others in the official research.
- Preliminary interview: The preliminary interview is implemented at
Agribank branches in the country, each branch interview 2 leaders about
the criteria used in researching.
4.1.2. Quantitative research
Sample survey: Number of established: 7 (branches) * 30
(customers) * 7 (regions) = 1470 samples
Individual customer list at each region was surveyed based on
random selection from customer list of branches in 7 regions. The survey
result was 1400 collected including 1123 valid. Data collected from valid
samples were encrypted and analyzed.
The suggested model is logistic regressive with 2 signals. This
model is convenient for calculating the customer possibility of customer in
the characteristic of target variable and measuring the impact of factors (in
the role of independent variable)
Including: each target variable Y is correspond with 1 model that
independent variables are selected if the correspond factor can change the
possibility of status. The model formula is:
With factors which are presented using qualitative variables, we

can point out the discrepancies of those qualitative variables by using the
following formula:
In this research, the author selected coefficient P value (sig) < 0.1
which ensures the reliability for statistical tests.

18

Gender
Age
V01 ....

V24
V25

DEBT
REPAYMENT
ABILITY

.........
V30
….
Model 4.1: The impact of factors on customer debt repayment ability
4.2. Data analysis results
The thesis has implemented and presented the analysis results
through 4 parts: the statistics of the features of survey subjects at 4.2.1; the
statistics of the description of the bank relation factors at 4.2.2; the analysis
results of how customers’ features influence on their relations with the bank
at 4.2.3; the analysis results of how bank relations influence on the
customer debt repayment ability at 4.2.4.
The aggregated result show that the models assessing the customer

debt repayment ability using the variables of the relation between customer
and bank has high suitablity and trustworthy and it is also suitable for
analyzing the content above.
CHAPTER V: THE MODEL OF INTERNAL CREDIT RATING
ACCORDING TO MARKET SEGMENT AT AGRIBANK,
RESULTS AND SOME RECOMMENDATIONS
5.1. Suggestions for the Internal Credit Rating according to market
segment at Agribank
5.1.1. The basis theory about customers segment who have credit
relation at Agribank
Based on the theory of the research overview at Chapter I and the
market segment at Chapter II, the author clearly pointed out the research


19

20

orientation on Internal Credit Rating according to market segment of the
thesis: The customer objectives segmented are those who have consumer
loan in Vietnam and 7 regions:
(1) The Northern midlands and the mountainous area including: Son
La, Dien Bien, Lai Chau, Lao Cai, Ha Giang, Cao Bang, Lang Son,
Hoa Binh, Yen Bai, Phu Tho, Tuyen Quang, Thai Nguyen, Bac
Giang, Bac Kan.
(2) The Red River Delta including: Vinh Phuc, Bac Ninh, Ha Nam,
Hung Yen, Hai Duong, Hai Phong, Thai Binh, Nam Dinh, Ninh
Binh, Ha Noi, Quang Ninh.
(3) The North Central area including: Thanh Hoa, Nghe An, Ha Tinh,
Quang Binh, Quang Tri, Thua Thien Hue.

(4) The Central Coastal area including: Da Nang, Quang Nam, Quang
Ngai, Binh Dinh, Phu Yen, Khanh Hoa, Ninh Thuan, Binh Thuan.
(5) The Central Highlands including: Gia Lai, Dak Lak, Dak Nong,
Lam Dong, Kon Tum.
(6) The South East including: Binh Phuoc, Tay Ninh, Binh Duong, Ho
Chi Minh City, Ba Ria Vung Tau, Dong Nai
(7) The Mekong River Delta including: Long An, Tien Giang, Ben
Tre, Tra Vinh, Vinh Long, Dong Thap, An Giang, Kien Giang, Can
Tho, Hau Giang, Bac Lieu, Soc Trang, Ca Mau.

5.2. The logistic model of Internal Credit Rating following market
segment processing result
5.2.1. Criteria system of each segment analysis
The thesis has presented the result of the influence level and the
comparison of the influence of 30 criteria in each region. The result show
that the criteria of the influence on the customer debt repayment ability in
each region are different. The inspection results are suitable with the
qualitative analysis in each region at Chapter II.
5.2.2. The criteria segmenting result
The thesis has presented the result of implementing Binary
regression logistic analysis method with the dependent variable is the
customer debt repayment ability and the independent variables are those
which describe the relations with the customers. The result is the influence
of the factors in group relation with the bank in 7 regions (including the
influence of 23 personal data criteria in each region).
5.2.3. The customer scoring logistic model combination
The method which determine this possibility is implemented in 5
steps:
Step 1: Collecting customer information
Step 2: Estimating the possibility of variables featured by the

relations with the bank using the results of 5 regressive models
Step 3: Calculating the values of variables showing the relations
between the bank and the customers
Step 4: Calculating the possibility of overdue principals
Step 5: Rating each customer.
All of the above steps are implemented on one file which calculates
each customer using their information automatically. The file consists of 2
sheets: Customer information sheet includes the encrypted information;
The other is used to determine values: The possibility of overdue debt,
rating on the calculation basis. With customers in each region, the index as
well as the coefficient is determined based on each different region.
The thesis has presented the method of scoring Internal Credit
Rating of the customer by performing 1 specific example.

5.1.2. Model suggestions
By researching and implementing model at Chapter IV, models are
used as followed:
• Customer’s relations status assessing model according each target
variable – logistic regression model using variables from V24 to
V30 (correspond with questions 24 to 30). The analyzed results can
set up the customer segment frame in 2 ways:
• The first is dividing possibility level of the status from bad to good
(or vice versa) according to each region
• The second is scoring each status of explanation variable based on
customer rate at different status of the target variable according to
each region


21

22

Following the current rating regulation at commercial bank, the
thesis has recommended the system of 10 levels in Internal Credit Rating
from AAA, AA, A, etc, CCC, C, D according to the following table:
AAA
AA
A
BBB
BB
B
CCC
CC
C
D
3525<
6555908545-54
75-84
95+
44
34
25
74
64
94
89

a. Collecting data: because of time limit as well as scale and range of
research, the survey results of the thesis still remains incomplete.
Therefore, Agribank should do survey in the wider range to ensure the

accuracy and trust level of the criteria in the testing and statistics procedure.
b. About the information system: To guarantee the valued data source of
the customers for the system to process in the Internal Credit Rating
operation, Agribank need to collect information from variety of sources:
from the customers, banks, credit information system of the commercial
banks system; suggesting solutions in order to better off the customers and
markets information source.
c. Continue to make better researches, tests on variance between regions in
order to increase the accuracy level of the credit rating according to market
segment results.

Table 5.32: The individual customer rating table:
All the steps above are implemented similarly with customers from
different regions, there are only differences when using the coefficient
determined in each region mentioned in criteria system and segment in each
region.
5.3. Model result discussion
5.3.1. Analysis result
Firstly, the criteria set did work and influence on the customer debt
repayment ability in 7 regions.
Secondly, the rating model based on each region has calculated the
influence level of criteria on customer debt repayment ability in each region
using the influence on the criteria of relations with the bank
Thirdly, this Internal Credit Rating system can be integrated on the Ipcas
technology background of Agribank based on updating criteria and region
density
Fourthly, the Internal Credit Rating method is quite simple and easy to
apply and save money.
5.3.2. Disadvantages
There are 2 main disadvantages: (1) the information collected

through surveys are the information come from the customers, thus it
cannot be totally legit, (2) research scale of the thesis is still limited in both
space and time therefore influence on the model calculating result.
5.4. Some recommendations
5.4.1. Recommendations on some contents which are necessary to
research on
With the target of improving the Internal Credit Rating system at
Agribank and commercial banks in Vietnam, the thesis recommends some
contents which are necessary to research on including:

5.4.2. Macro solutions recommendation:
The thesis has recommended 2 sets of solutions to complete the
internal credit rating operation at Agribank and the commercial banks
system. Those are:
- Micro solutions for state agencies and the state bank
(1) It is necessary to create strong juridical lobby for the commercial bank
system in the internal credit rating operation. Currently, the document for
guidance and direction for commercial banks to deploy internal credit
rating operation of the State bank is document No. 493/NHNN in 2005,
commercial banks almost separately created the internal credit rating
system for credit customer without synchronization.
(2) Standardizing the Internal Credit Rating method shown in 3 contents:
Firstly, the Internal Credit Rating systems of commercial banks were
approached on the basis of the Basel II standard.
Secondly, the Credit Rating operation must apply the appropriate rate
adjustments in order to decrease the differences in the result of Internal
Credit Rating of the same customer between commercial banks.
Thirdly, the regulations of the necessary and sufficient conditions for
expert to participate in creating the credit rating system is necessary.

23

24

The Internal Credit Rating system must be approached based on the
Basel II standard
(3) Setting up the juridical frame in trading and providing information
about the manufacture and business operation of individuals, enterprises
and credit information between commercial banks in Vietnam
(4) Enhancing the officers quality in Internal Credit Rating operation

for other kinds of customers in the future. After 4 years of research, the
thesis achieved some remarkable results.
Firstly, regional influence on the criteria in the Internal Credit Rating
system are tested and proved by running the model of 23 criteria independent variables and 7 criteria – dependent variables model in a
survey with 1470 observations forms conducted in 7 regions;
Secondly, the influence level of the criteria in each segmented regions has
been calculated while the foundation of the 7 credit rating systems for
individual customer with the criteria set which truly has impacts on the
Agribank credit customer debt repayment ability;
Thirdly, the adjustment coefficient among each region and scoring system
in the board classification ratings is determined
Fourthly, 2 groups of solutions has been recommended such as: micro and
macro solutions to overcome the disadvantages of the Internal Credit
Rating operation in commercial banks system in general and Agribank
especially.
With the mentioned results, the author has completed the target
research plan./.

5.4.3. The micro solutions:
(1) Programming specific software for internal credit rating operation
(2) Applying the segmenting method/model in Internal Credit Rating
operation in Agribank
(3) Setting up the criteria system suitable with each market segment
(4) Enhancing the human resources quality in Agribank system.
GENERAL CONCLUSION
Credit risk management of commercial banks in general and
especially of Agribank is the most important problem for banks to survive
and develop. Credit risk management is the foundation for banks to
increase their ability of competing then develop the national economy.
Internal Credit Rating is one of the most important tools in internal credit
risk management operation in each banks.
As presented, the author has analyzed the generalization of the
credit rating operation in the world and in Vietnam up to the present,
figured out the gap in researches as well as methods, credit rating models.
The author has researched on theoretical and practical matters about the
market and market segment; the market factors which influence on the
internal credit rating operations; the role of market segment on internal
credit rating operations of commercial banks in Vietnam, especially
Agribank. Thereby, the author emphasized the advantages and
disadvantages which is necessary to overcome, solve and improve for the
Internal Credit Rating system of Agribank mentioned in the thesis.
Based on that, the thesis has concentrated on researching and
creating the Internal Credit Rating following market segments system for
Agribank credit individual customers in Vietnam, which will be the pioneer