What do credit card type A and type B mean?

By implementing a scoring system for customers, we can judge whether the customers are of high quality or not. Card A (ApplicationScoreCard), application scorecard

Card B (BehaviorScoreCard), behavior scorecard

Since retail credit business has the characteristics of large number of transactions, small single amount, and rich data , determines the need for an intelligent and probabilistic management model. The credit scoring model uses modern mathematical statistical model technology, and through in-depth data mining, analysis and refinement of borrowers' credit history and business activity records, it discovers knowledge contained in complex data that reflects consumer risk characteristics and expected credit performance. and rules, and summarize them through scoring as a scientific basis for management decision-making.

Differences:

1. The time of use is different, focusing on before the loan, during the loan, and after the loan.

2. The data requirements are different, A card generally does The data of 0 to 1 year before the loan, B card is carried out after the customer has certain behavior and has larger data, usually 3 to 5 years,

3. All models of each score card Differently, A card mostly uses logistic regression, while the latter two commonly use multi-element logistic regression, which has better accuracy.

B Card

1. Definition: Predict the future overdue probability based on the lender’s behavior after lending (observation behavior)

2. Usage scenario: Loan issuance The time period before maturity

3. Pay attention to the observation period, performance period, and time slicing issues

Division

1. Based on repayment willingness and repayment Depending on the ability, different risk levels are divided

Mild: Good repayment willingness and repayment ability, overdue for special reasons

Medium-light: Good repayment willingness and repayment ability appear Problem

Moderate: Willingness to repay has deteriorated, but ability to repay is available

Severe: No willingness to repay, ability to repay has deteriorated or been lost

2. Collection Process

SMS collection, phone collection, on-site collection, legal proceedings, third-party collection (overdue assets are packaged and sold)

3. Model composition

Repayment Rate model: predicts the rate of debt collection after collection

Aging rolling model: predicts the probability of the number of overdue people converting from mildly overdue to severely overdue

Loss of contact model: In the overdue stage, predict the probability of loss of contact for those who can still be contacted

4. Common indicators

Number of overdue days

Historical repayment rate information

Proportion of overdue amount

Proportion of debt burden

Personal information (gender, age, income, job, education, etc.)

Bank’s Credit rating standards for credit card holders

For reasons of confidentiality, the name of the bank cannot be published, but there is still a lot of sensitive information in it. An experienced person should be able to guess it at a glance.

The scoring standard is just a scoring standard. It is a rough assessment of the customer. It does not have a decisive impact on the final review. It is not the only basis for decision-making. It is subjective and excludes green channels, situation verification and other promotion policies. outside of decision-making. But indeed, we can find many problems from it.

Bank A adopts a hundred-point system, which consists of four parts: guarantee support, economic support, stability and personal background:

1. The maximum score for guarantee support is 15 points

(1) The maximum score for housing rights is 8 points

No house

0 points

Renting

2 points

Unit welfare room allocation

4 points

Own or purchase

8 points

(2) Maximum score with or without mortgage 7 points

Secured

7 points

Unsecured

0 points

2. Financial support The maximum score is 34 points

(1) The maximum score for personal income is 26 points

Monthly income is more than 6,000 yuan

26 points

Monthly income 3,000~6,000 yuan

22 points

Monthly income 2,000~3,000 yuan

18 points

Monthly income 1,000~2,000 yuan

13 points

Monthly income 300~1,000 yuan

7 points

(2) The maximum score for monthly debt repayment is 8 points

No debt repayment

8 points

10~100 yuan

6 points

100~500 yuan

4 points

More than 500 yuan

2 points

3. The maximum score for personal stability is 27 points

< p>(1) The maximum score for employment status is 16 points

Civil servants

16 points

Public institutions

14 points

State-owned enterprises

13 points

Joint-stock enterprises

10 points

Others

4 points

Retirement

16 points

Unemployment with social assistance

10 points

Unemployment without social assistance

p>

8 points

(2) The maximum score at the current address is 7 points

More than 6 years

7 points

< p>2~6 years

5 points

Less than 2 years

2 points

(3) The maximum score for marital status is 4 Points

Single

2 points

Married with no children

3 points

Married with children< /p>

4 points

4. The maximum score for personal background is 24 points

(1) The maximum score for household registration is 5 points

Local< /p>

5 points

Overseas

2 points

(2) The maximum score for education level is 5 points

Junior high school and below

1 point

High school

2 points

Technical secondary school

4 points

College and above

5 points

(3) The maximum score for age is 5 points

Female over 30 years old

5 Points

Male over 30 years old

4.5 points

Female under 30 years old

3 points

Male 30 points Under the age of

2.5 points

(4) The maximum score for breach of trust is 9 points

Not investigated

0 points

No record

0 points

Breaking trust once

0 points

Breaking trust more than twice

-9 points

No breach of trust

9 points

Bank B also adopts a hundred-point system, which is divided into four categories: natural situation, professional situation, family situation and relationship with the bank. Part of the content consists of:

Project

Scoring criteria

Score

Automatically

ran

< p>Circumstances

Situation

Age

Under 25 years old

26~35 years old

36~50 years old Years old

Over 50 years old

2

4

6

4

Gender

Male

Female

1

2

p>

Marital status

Married with children

Married without children

Single

Others

< p>5

4

3

2

Health status

Good

General

Poor

5

3

—1

Educational level

Graduate degree or above

Undergraduate degree

College degree

8

6

4

Technical secondary school, high school

Others

2

1

Nature of household registration

Permanent household registration

Temporary household registration

2

2

Occupation

Occupation

Love

< p>Categories

Types of units

Governments and institutions

State-owned enterprises

Collective enterprises

Military

6

4

3

5

Sole proprietorship

Sole proprietorship< /p>

Foreign-funded enterprises

Others

2

2

5

1< /p>

Unit

Economic status

Good

General

Poor

4

2

—1

Engaged in the industry

Development prospects

Good

General

Poor

4

2

—1

Nature of the position

Unit supervisor

Department supervisor

General staff

6

4

2

Position length

More than 2 years

1~2 years

Within 1 year

3

2

1

Professional title

Senior

Intermediate

Junior level

No professional title< /p>

4

2

1

Monthly income

More than 10,000 yuan

8,000 ~10,000 yuan

5,000~8,000 yuan

4,000~5,000 yuan

12

10

9< /p>

8

3000~4000 yuan

2000~3000 yuan

1000~2000 yuan

Below 1000 yuan

6

4

2

1

Home

Family

p>

Situation

Situation

Family monthly

Average income

More than 5,000 yuan

4,000 ~5000 yuan

3000~4000 yuan

2000~3000 yuan

9

6

5< /p>

4

1000~2000 yuan

Below 1000 yuan

2

1

Related to

this

line

whether

this Bank employee

Yes

No

2

Bank account

Have a credit card

< p>Have a debit card

No

6

4

Deposit balance

Higher

Lower

None

6

4

Business

Frequent visits

Frequent

Normal

Rarely

4

2

Others

Borrowing status

Never borrowed money

Has borrowed money but has paid it off

Has default record

4

5

—5

Total score

This scoring standard is also used by the bank as a reference standard for loan limit review. The basic corresponding segment quotas are: above 90 points, the quota is 600,000 yuan; 80 to 89, 100,000 yuan; 70 to 79, 50,000 yuan; 60 to 69, 10,000 yuan; 50 to 59, 5,000 yuan; 40 to 49, 3,000 yuan. 40 or less, 0 quota. Generally speaking, according to the credit limit, 50,000 basically corresponds to the starting point for credit card platinum card customers.

I still have a complete phone number and rating form from a bank, but they are not of much help, so I won’t send them out again. The following content is more critical:

However, in the credit scoring standards of various commercial banks in China, we do not seem to have seen some relevant indicators that we think may be included in the scoring standards. Now let’s take a look. Some excerpts from the two most widely used scoring models in the world:

The David Durand credit score derivative model has the following provisions:

Credit surveys in the past 6 months

p>

1~3

More than 3 times

8 points

2 points

—5 points

The regulations in the most widely recognized FICO credit score model in the world are more clear:

Within one year

Number of inquiries

1

2

3

4

5~9

No record

3 points

p>

11 points

3 points

—7 points

—7 points

—20 points

0 points

Credit file

Case years

0.5

1~2

3~ 5

5~7

7

0 points

5 points

15 points

30 points

40 points

Number of revolving credit overdraft accounts

1~2

3~5

5

5 points

12 points

8 points

—4 points

Credit Quota

Utilization rate

0~15%

16%~30%

31%~40%

< p>41%~50%

50%

15

5

—3

—10

—18

Maybe in our impression, the more fully the quota is used, the better, but this may not be the case. Of course, the scoring standards for requesting a quota increase may be stipulated in this way. .

Bank of China Xiamen Branch Great Wall Credit Card admission criteria:

1. Age score

35-55 years old 10

25 -35 years old 8

Over 55 years old 6

18-25 years old 2

2. Marital status

Married 10

p>

Single 3

3. Educational background

Master's degree and above 10

Bachelor's degree 8

College degree 5

Technical secondary school, technical school, high school 3

Junior high school 1

Elementary school 0

4. Unit

Party and government agencies , public institutions 10

State-owned enterprises, listed companies 8

joint-stock or collective enterprises 6

foreign-funded enterprises 4

individuals and others 2

5. Housing

Purchased 10 commercial houses

Purchased 8 low-profit houses

Purchased 7 welfare houses

Renting 2

6. Working experience

More than 8 years 10

5-8 years 8

3-5 years 5

Below 3 years 0

7. Position score

Above bureau level or senior management of large companies 20

Above division level Or mid-level managers in large companies 15

Above department level or 10 general managers in large companies

5 below department level

8. Professional titles

< p>High Normal School 10

Middle School Normal School 8

Assistant teacher 5

Assistant teacher below 0

9. Monthly income

More than 10,000 yuan 20

8000-9999 yuan 15

5000-7999 yuan 13

3000-4999 yuan 10

2000-2999 yuan 8

1000-1999 yuan 6

600-999 yuan 3

Below 600 yuan 0

10. Security deposit

Below 10,000 yuan 20

8,000 yuan 18

5,000 yuan 15

3,000 yuan 10

11. 50 major private business customers

12. 50 major card business customers of our bank

13. 10 major consumer credit customers of our bank

14. Major corporate business customers of our bank Customer 50

Applicants who meet the above individual conditions and have a comprehensive score of 80 points can apply for a personal gold card; those who have a comprehensive score of 65 points can apply for an individual ordinary card. ! ! !

If the overall score is insufficient, can I apply for a credit card?

If the comprehensive score is not enough, you can go to the bank to apply for a loan, or you can apply for several online loans, but the pass rate is very low, so it is not recommended for everyone to try. Borrowing users must know the reasons for insufficient comprehensive scores and solve the problems before they can apply for a loan. Generally, the reasons for insufficient comprehensive score are too many personal credit inquiries, too much debt beyond the borrower's repayment ability, bad credit record and untrue personal information. It is recommended that you find your own reasons and improve them before applying. loan, otherwise the overall score will only get worse. After improvement, you can learn about the following formal platforms with lower requirements.

1. Personal comprehensive credit score refers to the use of scientific and rigorous analytical methods to comprehensively examine the internal and external subjective and objective environments that affect individuals and their families, and to fulfill various economic commitments. Ability to make comprehensive judgments and assessments. For different applications, credit scores are divided into risk scores, income scores, responsiveness scores, customer churn (loyalty) scores, collection scores, credit card issuance review scores, mortgage loan issuance review scores, credit limit review scores, etc.

2. Comprehensive credit risk score - Pengyuan 800. At the end of April 2005, Pengyuan Credit Information Co., Ltd. independently developed the personal comprehensive credit risk score - "Pengyuan 800", which was officially approved by credit institutions. Provide credit score inquiry services to individuals. "Pengyuan 800" conducts statistical analysis on personal credit information by establishing a mathematical model to predict the possibility of default risk in the future, and uses a score to comprehensively reflect the personal credit status.

This credit scoring system has six levels, ranging from 320 to 800, with each 80-point level quantifying the individual's credit status in detail. Each score corresponds to a probability of default. The higher the score, the lower the risk of default. Illustration of the corresponding grades and default probability of the score: the lowest score is 320400480560640720800 and the highest score is FEDCBA.

3. The scoring model uses more than 40 variables related to personal credit, which can be summarized into four categories of variables: personal basic information, bank credit information, personal payment information, and personal capital status. Among them, Bank credit information accounts for the largest weight, close to 50%, and the remaining three types of variables account for roughly the same weight. Currently, customers with bank credit records in the credit information system database only account for 25% of the total population. Since bank credit information is the most important variable that affects personal credit status, for customers without bank credit records, the model selects other factors related to bank credit. variables are replaced. In the future, as the data gradually improves, we will add more variables to the model to continuously improve the accuracy, precision and universality of the model.

What is a credit scoring model?

The credit scoring model is a classification of personal financial authority that has emerged in recent years to ensure the financial security of banks and other financial sectors. Define the model. This model refers to using a certain credit scoring model based on the customer's credit history information to obtain different levels of credit scores. Based on the customer's credit score, the amount authority that the customer can hold is determined to ensure the security of repayment and other services. . As loans and credit cards play an increasingly prominent role in modern society and companies, the development prospects of credit scoring models are immeasurable.

Risk control scorecard model

The scorecard is a tool for quantifying customer risks based on comprehensive customer data and using statistical learning methods. It can be applied to a variety of financial risk controls. Scenario-based risk management.

The general presentation method of traditional scorecards is (example):

Through scorecards, customers can have a score that represents the customer’s credit risk level. Generally, the higher the score, the better the credit qualification. The better. Sesame Credit score and WeChat Pay score are both score cards.

The most commonly used scenario for scorecards is credit business, such as credit cards, consumer installments, small cash loans, etc. At different stages of the credit business process, scorecards play different roles and therefore correspond to different types of scorecards. Usually, there are three types of scorecards: Application Scorecard, Behavior Scorecard and Collection Scorecard, namely A card, B card and C card, which have different uses at different stages of credit.

1. Apply for a scorecard

The customer group for applying for a scorecard is all customers who apply for a loan. It evaluates the customer's credit risk at the time of application and assists financial institutions in making decisions. , pass, reject or further review. For example, a customer with a score of 650 or above indicates excellent credit and can directly apply and be given a higher credit limit and a lower preferential interest rate; a customer with a score of below 400 indicates poor credit and can be directly rejected; a customer with a score of 400-650 can directly reject the application. The middle customer group is the most critical. How to manage this customer group can best reflect the risk control capabilities of financial institutions, because the top customer group is the target of all financial institutions. Even if credit is granted to them, they may not be able to do so due to quota and The interest rates did not meet their expectations and they defected to other institutions.

Although the intermediate customer group carries certain risks, good management will also bring greater profits. From the perspective of risk prevention and control, we can further manually intervene in the review, ask customers to provide more data, or provide certain collateral for risk mitigation. At the same time, it is necessary to quantify the score according to the score card, appropriately reduce the amount and increase the interest rate.

2. Behavior scorecard

The customer group of the behavioral scorecard is the customers who have applied for credit or have drawn but have not yet paid off the loan. The purpose is to evaluate the customer's future period. Due to potential overdue risks, financial institutions can take certain measures in advance to prevent the occurrence of risks.

Compared with applying for a score card, in addition to different customer groups, the behavioral score card is supported by data such as customer historical expenditure and repayment and quota usage, so the risk assessment of customers is more accurate.

Behavior scorecards can be scored every time a customer withdraws money, or they can be scored in batches on a regular monthly (quarterly) basis, depending on the type of business and the financial institution's strategy. If the behavior score is low, the institution can reduce or freeze the limit, settle the payment in advance, etc.

3. Collection Score Card

The Collection Score Card is aimed at overdue customers and evaluates the probability that customers can collect their money within a certain period of time in the future. The main purpose of the collection scorecard is to reduce collection costs by optimizing the collection process while improving the collection rate of overdue customers. For example, for customers with higher collection scores and a higher probability of repayment, text message reminders or robot voice reminders can be used. On the contrary, for customers with low collection scores, manual collection can be directly used.

Collection scorecards can be divided into different categories according to different collection stages, such as pre-collection, early collection, and late collection. Separate collection scorecards can be developed at different stages based on different customer groups. The pre-collection score card assesses overdue risks for customer groups that have not yet been overdue but are about to expire. For customer groups with higher overdue risks, early warnings or early collections can be issued in advance; early collections are mainly reminder collections, and those with lower scores can be Customers can intervene in manual collection; the overall recovery probability of late collection is not high. For customer groups with low scores, they can be outsourced for collection or go through legal procedures in advance. Customers with high scores can organize their own collection to achieve the highest possible improvement overall. Improve collection rates and reduce collection costs.

Different credit products have different business risk points. The scorecard is only a quantitative model learned based on data, which requires high data quality. For data with relatively low quality, it is not suitable to be used as a scorecard variable, but it does not mean that this part of the data is unimportant. Therefore, it is far from enough to rely solely on scorecards for credit risk control. Comprehensive risk management must be done in conjunction with risk strategies. On the contrary, scorecards are only a small part of the strategy. For credit strategies, they are comprehensive risk management rules at all levels, including policy and institutional levels, expert experience levels, data levels, and scorecard rule sets, and require rapid iteration, while the frequency of scorecard iterations does not require Not high. I will explain the strategy later.

As the beginning of the scorecard series, this article can give us a macro understanding of the scorecard. In subsequent articles, I will explain the development and construction process of the scorecard from a technical level, so that we can understand the scorecard and the style of the scorecard. Control has a further and deeper understanding.

Model of manual credit card review

The so-called online application is an online application. Customers need to log in to the official website, or scan the official QR code to enter the card-issuing bank’s online application page, page by page. Fill in the application information and complete the application process. The offline channel is to fill in the application form in paper version or electronic device through the account manager of each card-issuing bank. After completing the information, you will enter the credit card application review process.

The credit card review process requires two procedures: electronic review and manual review.

1. Electronic review

Each bank will have an electronic scoring model. After different customer application materials enter the system, the information will be matched according to the system module, big data will be verified, and the electronic system will evaluate Based on the completeness of the information on the application form, we will then use the personal information on the application form to determine whether age, income, etc. meet the requirements of the approval policy, and automatically query the applicant's credit status in the personal credit system as a reference.

That’s it for the introduction of the credit card scoring model.