How to Make a Credit Scoring Model and its Examples
How do we create an accurate credit scoring model? This question must have crossed your mind when you were about to meet a potential debtor. The reason is, credit scoring must be done accurately and periodically. Credit scoring is intended so that the resulting data can provide an overview of the background of the prospective debtor, such as the ability to make loans and have low credit risk.
The presence of credit scoring will certainly make it easier for lenders to verify so that prospective borrowers can apply for the required loan. So, what should be done when trying to create a good credit scoring model?
What is Credit Scoring?
Broadly speaking, credit scoring is a system implemented in financial institutions to determine the ability of prospective borrowers to make loans. To obtain credit scoring information, financial institutions need to conduct a review of personal data, credit history, age, status, length of service, job position, and other financial data that will be used as an assessment indicator.
This data review presents the condition of the prospective borrower and the risks that will be obtained when making a loan. In conducting credit scoring, systematic stages and careful credit analysis are required with an assessment guide that includes the 5C principles (character, capital, capacity, condition, and collateral). Other assessments are also concluded in the 5P principle (party, purpose, payment, profitability, protection).
The results of the credit scoring review will determine the credit score that determines a financial institution to determine approval or rejection of a prospective borrower’s credit application. A carefully designed credit scoring model can also help reduce credit risk and promote a better lending experience.
How to Check the Credit Scoring?
Before applying for a loan from a financial institution such as a bank, prospective debtors should first check their credit scores. This check aims to check credit history, carry out evaluations, find out credit score levels, and review several factors that affect credit scores.
Knowing this procedure is very useful before making a loan. Information from checking credit scoring is also a consideration for creditors to approve or reject credit applications from prospective debtors.
Prospective debtors can check by accessing the Financial Information Services System (SLIK) and obtaining related information. One of the institutions that provide this service is the OJK (Financial Services Authority) through the debtor information request application (iDebku). This process was previously known as BI Checking which was under the management of Bank Indonesia.
This check is useful as a reference for applying for a loan. In Indonesia, a good credit score ranges from 800 to 900. This score indicates that the debtor can fulfill his obligation to repay the loan properly. Conversely, a score that is below the adequacy value is more difficult to get a loan from a financial institution.
According to BI Checking, credit scoring is divided into 5 score levels that consist of:
- Score 1: The debtor is disciplined in making payments without being in arrears.
- Score 2: The debtor is recorded as having been in arrears for 1-90 days and is under special attention.
- Score 3: The debtor is not paying smoothly and credit installments are in arrears for 91-120 days.
- Score 4: Credit installments are overdue for 121-180 days.
- Score 5: Credit has been bad and credit installments are overdue for more than 180 days.
How to Make the Model
The process of creating a credit scoring model is carried out through variable determination, data collection and analysis, scoring model, model validation, and implementation. All of these stages become indicators of assessment for making credit scoring. The ways to create a credit scoring model are as follows:
1 . Variables
The first indicator is the variables used by financial institutions. These variables are useful for determining whether a prospective borrower has a good credit history and background.
Within this variable, the information used also varies, ranging from credit history, payment history, delays, and arrears owned. The amount of debt is also classified as information that will be considered. In addition, financial parties want to ensure credit risk which can be assessed by age, occupation, income, number of credit cards owned, credit period, and type of credit.
2 . Data Collection and Analysis
Data is an important part of creating a credit scoring model. The finance company will collect and analyze the data carefully. Data can be obtained from credit institutions and contains personal information, credit history, the amount owed, debt-to-income ratio, age, occupation, and income. This data will be used to evaluate and build a suitable credit scoring model.
3 . Model Scoring
A scoring model is a method built from predetermined variables and contains several historical and statistical information or data. This method is useful for estimating the eligibility of prospective borrowers to receive bank loans and the probability that the loan will be repaid. This data also includes possible delays or arrears that could occur.
The scoring model also displays credit scores after an in-depth analysis process. This figure indicates eligibility to receive a loan. This credit score involves processing very large data so that the reference from the scoring model can also minimize credit risk.
4 . Model Validation
Model validation is useful for evaluating how well the credit scoring model is used to estimate a person’s credit risk. This is done so that the scoring model that has been produced from data processing can be ensured to be accurate and credible. Model validation also goes through several stages, starting from testing, evaluating, and improving the model.
This stage serves to strengthen the results of data processing. So if inaccuracies are found in the model, then the process must go through each stage until the model is improved or optimized.
Read Also: How to Solve Bad Credit and Its Examples
Examples of Credit Scoring Simulation
Generally, a credit scoring model simulation requires several variables about pieces of information that have been carefully examined. These variables are then prepared to be processed through data processing. The aim is to estimate credit risk. Through financial marketplace services that provide credit scoring services, you can try to process personal data by matching some of the required information.
The credit scoring model simulation requires types of personal information, including age, occupation, income, marital status, number of dependents, home ownership, and number of credit cards. The data provided will then be processed by the service provider and provide an overview of credit scoring from data processing and your history.
The following is an example of a credit scoring model simulation as an illustration of the information provided. Mr. Frans is a 35-year-old man, married, and has one child. His job status is as a staff member in a private company in Jakarta with a fixed income of around IDR 7,000,000 per month. Mr. Frans has worked for 5 years at a privately owned house and does not have a credit card.
From the example above, the credit scoring results obtained are in the range of 452. This figure is considered sufficient and can be improved. However, this value is below average and there may be obstacles to applying for credit.
This simulation does not involve other complex variables, namely historical data relating to credit, arrears, debt ratios, financial ratios, and other information deemed necessary to be processed in a data processing system.
Creating a credit scoring model is very important to do in determining the creditworthiness of prospective debtors. Judging from how important it is to make this model, only use Credit Scoring AdIns that can optimize every single credit scoring process that you’ll do. Contact us right now to get more information about this service!