ADVANCE SCORE performs credit score modelling, drawing data from a combination of the client, ADVANCE’s databases and third party databases. SCORE’s engine uses statistical analysis to create a portfolio for each applicant which entails the applicant’s credit score, default risk and recommended action.
SCORE is unique from other parties’ credit scoring engines in that it uses Machine Learning techniques to constantly improve the model’s algorithms. Through a positive feedback loop that monitors and evaluates application results according to repayment speeds and loan security, SCORE continuously refines its model, maximizing accuracy and minimizing risks in scoring applicants in the long run.
To ensure there aren’t underlying significant risks when approving a loan, an in-depth profiling of an applicant’s details is necessary. A twofold approach is taken to combat fraudulent applications.
Credit History Check
A fastidious check for any issues in the applicant’s credit history is conducted. Historical defaults and historical credit checks are taken into account prior to Score recommending a decision for the applicant. These factors are vital to calculate the probability of default.
The applicant’s provided details are cross-checked with ADVANCE’s database. including their personal particulars and their employment records. Personal particulars including the applicant’s name, gender, place of birth, and date of birth are inspected by Guardian, while their employment details are processed to develop a more accurate score, appending their employment type factor to the applicant’s default risk.
CREDIT SCORE TAKES INTO ACCOUNT
The work environment relative to the company and office location. The work location reflects the profiles of the company and employees.
The residential locations. Attributes such as the average renting price reflects income levels and expenditures.
Details such as the applicant’s age, marital status, number of dependents are used to judge their financial stability.
The job sector the applicant is currently in, their education and their profession contributes to job security analysis.
The applicant’s job income relative to their job position.
TRENDS IN HISTORICAL APPLICATIONS
A ‘bad loan’ is a loan where the repayment time exceeds the allotted time period. Several trends regarding an applicant’s associated risk can be extracted from analysing a broad collection of 10 million historic loan records.
As shown in the diagram below, the correlation between level of education and frequency of bad loans is evident – the higher the applicant’s educational level, the less likely it is for that applicant to make a bad loan.
Another correlation is seen between an applicant’s historical income level and their repayment results. SCORE uses this correlation to predict an average rate for future applicants, partitioning them into separate categories depending on their income.
LET’S GET STARTED
Refine your business reliability while enhancing your customer's experience