Carding Forum
Professional
- Messages
- 2,788
- Reaction score
- 1,244
- Points
- 113
Along with traditional, financial institutions also use alternative data sources that help determine the reliability of the borrower
By issuing a loan, banks or financial companies risk not getting their funds back. To make the right decision - to lend or refuse a loan - credit scoring is used in the world. It is a mathematical or statistical model by which the lender determines the likelihood that the borrower will repay the loan on time. Scoring provides for the assignment of points - the higher it is, the lower the risk. Banks and MFIs use both their own scoring systems and data calculated by credit bureaus. Credit scoring models may differ slightly in the way a loan is scored. The Fair Isaac Corporation system, known as the FICO score, is the most used system in the financial industry. Another popular credit scoring model, VantageScore, is created by three leading credit reporting agencies: TransUnion, Experian and Equifax. 2. Most scoring models take into account the following data:
Some experts criticize credit scoring for failing to clearly account for current economic conditions. If, for example, borrower A has a credit rating of 800 and the economy goes into recession, then borrower A's credit rating will remain static until its behavior or financial situation changes. FICO tried to address this gap by introducing the FICO Sustainability Index in April 2020. Advances in areas such as machine learning and other analytic-friendly computer languages are helping to improve the accuracy of credit risk modeling.
Alternative scoring sources
With the development of fintech, many alternative scoring sources have emerged. Alternative sources may include transaction data, from mobile and other devices, social media, behavioral factors, and the like. For example, transactional details can include account settlement behavior (such as a credit card) as well as e-commerce data. Social media can also help to understand the reliability of the borrower. Several studies have shown that the number of publications and their frequency provide an indication of lifestyle, spending and willingness to repay. Another useful source of credit scoring is analyzing the history of utility bills. Behavioral factors analyze how a person fills out the questionnaire, how the slider moves, the duration of actions. The logic is
For example, the German lending service Bintbond requires access to several profiles - PayPal, Amazon, an online bank account. The scoring model of the American company Branch is based on data received from mobile phones - call history, contact list, SMS-logins.
FICO offers two alternative data scoring products - the FICO Score XD, which uses payroll data for telephones and television bills, and UltraFICO Score, which uses escrow account information. David Schellenberger, vice president of valuation and predictive analytics at FICO, says data such as utility bills, telephone and television bills, or escrow account information can "reach millions of consumers and help them get their first loan."
By issuing a loan, banks or financial companies risk not getting their funds back. To make the right decision - to lend or refuse a loan - credit scoring is used in the world. It is a mathematical or statistical model by which the lender determines the likelihood that the borrower will repay the loan on time. Scoring provides for the assignment of points - the higher it is, the lower the risk. Banks and MFIs use both their own scoring systems and data calculated by credit bureaus. Credit scoring models may differ slightly in the way a loan is scored. The Fair Isaac Corporation system, known as the FICO score, is the most used system in the financial industry. Another popular credit scoring model, VantageScore, is created by three leading credit reporting agencies: TransUnion, Experian and Equifax. 2. Most scoring models take into account the following data:
- Does the potential borrower repay the loan on time? If the bills are paid late, the applicant has filed for bankruptcy at some time, this is likely to negatively affect the prospects for obtaining a loan;
- has the maximum been reached? Many systems compare the amount of outstanding debt with credit limits. If the amount of debt is close to the credit limit, this increases the chances of hearing a refusal from the lender;
- How long has a potential borrower had a loan? Typically, rating systems take into account the credit experience of the borrower. A short credit history can be harmful, but paying bills on time can compensate for this;
- Have you recently applied for a new loan? If a bank client has recently applied for many new loans, this reduces his prospects for getting a new one;
- How many credit accounts are open and what are they? While it is generally considered a plus to have open credit accounts, too many credit cards can reduce your chances of getting a loan approved.
Some experts criticize credit scoring for failing to clearly account for current economic conditions. If, for example, borrower A has a credit rating of 800 and the economy goes into recession, then borrower A's credit rating will remain static until its behavior or financial situation changes. FICO tried to address this gap by introducing the FICO Sustainability Index in April 2020. Advances in areas such as machine learning and other analytic-friendly computer languages are helping to improve the accuracy of credit risk modeling.
Alternative scoring sources
With the development of fintech, many alternative scoring sources have emerged. Alternative sources may include transaction data, from mobile and other devices, social media, behavioral factors, and the like. For example, transactional details can include account settlement behavior (such as a credit card) as well as e-commerce data. Social media can also help to understand the reliability of the borrower. Several studies have shown that the number of publications and their frequency provide an indication of lifestyle, spending and willingness to repay. Another useful source of credit scoring is analyzing the history of utility bills. Behavioral factors analyze how a person fills out the questionnaire, how the slider moves, the duration of actions. The logic is
For example, the German lending service Bintbond requires access to several profiles - PayPal, Amazon, an online bank account. The scoring model of the American company Branch is based on data received from mobile phones - call history, contact list, SMS-logins.
FICO offers two alternative data scoring products - the FICO Score XD, which uses payroll data for telephones and television bills, and UltraFICO Score, which uses escrow account information. David Schellenberger, vice president of valuation and predictive analytics at FICO, says data such as utility bills, telephone and television bills, or escrow account information can "reach millions of consumers and help them get their first loan."