Credit scoring assigns points to a consumer’s credit information and totals the points to create a three-digit credit score. When a consumer applies for credit with a bank, the lender uses a credit scoring system to then see the applicant’s credit score, and thus make lending and pricing decisions. An auto lender, for example, may use the FICO Score 9 to gauge a consumer’s likelihood of defaulting on a new auto loan, approve or deny the application, and assign an interest rate.
How Credit Scoring Works
Credit scoring uses information from consumer reporting agencies to produce a three-digit credit score that allows businesses to quickly, sometimes automatically, gauge a consumer’s likelihood of defaulting on a debt obligation. In addition to analyzing credit information, credit scoring models can factor in changes in consumer behavior and broader economic patterns like a recession, too. Businesses in various industries use credit scoring, including lenders, insurance providers, electric and gas companies, and internet service providers. To create a credit scoring model, developers look for patterns and trends within a sample of consumer data to determine risk characteristics. Those characteristics are broken down into specific attributes and assigned points to indicate whether the attribute has a good or bad outcome. Attributes include things like:
Total number of accounts in good standing Total number of mortgage accounts, or Average balance on the consumer’s credit cards.
The characteristics, attributes, and point values of the consumer segment create a table called a scorecard. A specific credit scoring model may be generated using several scorecards to account for risk with a specific group of consumers. For example, there may be scorecards for consumers with a high amount of credit card debt or a thin credit file. Let’s look at an example of how scoring might work in regard to payment history. Attributes for the number of months since the consumer’s last delinquency could appear on a scorecard. In this case, the more recent a delinquency, the fewer points they will receive. In this example, if the consumer has no serious delinquencies, they would receive 75 points. On the other hand, a consumer with recent delinquencies (between zero and five months) would receive only 10 points. A person’s credit score is essentially the total of points received on a scorecard. Therefore, having several risky or negative attributes on a scorecard means fewer points and ultimately leads to a lower credit score.
Credit Scoring Breakdown
While there are hundreds of possible attributes, credit scoring models like the FICO and VantageScore generally group them into broad categories. Different types of information receive different weights in the credit score. For example, the FICO Score considers:
Payment history: 35%Amounts you owe, or credit utilization: 30%Length of your credit history: 15%Mix of your credit accounts: 10%New credit accounts: 10%
By comparison, the VantageScore considers:
Payment history: 41%Length and mix of credit history: 20%Credit utilization: 20%New credit: 11%Balances owed: 6%Available credit: 2%
Business Use of Credit Scoring
Many businesses use a credit scoring model that has been tailored for their industry. The formula breakdown for industry-specific credit scoring models isn’t available for the general public. This means we don’t know which specific credit information helps lenders predict certain information. For example, we don’t know how lenders predict that a borrower will default on an existing auto loan within the next 12 months.
Updates to Credit Scoring Models
Credit scoring companies periodically introduce newer versions of their credit score models, improving the analysis of consumer payment habits. For example, later versions of the FICO Score exclude paid third-party collections and include alternate data like rental tradelines, telecom, and utilities when they’re on the consumer’s credit report.
Laws Affecting Credit Scoring
The Fair Credit Reporting Act largely ensures consumers have an accurate credit report, and also requires credit score providers to allow consumers to access their credit score. Additionally, when a consumer’s credit score is used to deny credit or charge a higher interest rate, the consumer must receive a copy of their credit score and the risk factors that contributed to their score.By law, credit scoring cannot use certain types of information that may discriminate when landing on a score. This includes race, color, religion, national origin, sex, marital status, age, and whether you’re receiving money from public assistance.
Types of Credit Scoring
There are several types of credit scoring, mainly designed to help businesses make accurate decisions based on credit information specific to their industry.
General Lending
Businesses across industries may use base credit scoring models rather than an industry-specific version of the credit score. Credit scores produced by a general lending model are referred to as “generic” scores.
Auto Lending
Auto dealers and lenders can assess the risk level of consumers seeking auto financing and determine the likelihood of default within a 12-month period.
Credit Card Decisions
There are credit scoring models created for credit card issuers to determine the default risk of credit card applicants.
Mortgage Lending
Mortgage lenders use a specific credit scoring model to approve applications and to set an annual percentage rate (APR). FICO, for example, provides a mortgage credit scoring model for each credit bureau: FICO Score for Experian, FICO Score 5 for Equifax, and FICO Score 4 for TransUnion. Some mortgage lenders use a general FICO score version created before FICO 8.
Energy
Electric and gas companies can assess the amount of risk in providing services to customers by identifying people who may become delinquent within a 12-month period.
Broadband
Cable and broadband service providers can determine which prospective customers may become delinquent within a 12-month period.
Proprietary
Some businesses or contract credit reporting agencies create their own credit scoring models. These proprietary or custom versions are generally not publicly available and may include data that are not available on a consumer’s credit report. Credit-based scoring models may be able to predict your income and debt-to-income ratio, as well as the risk of creating losses for an insurance company. Models also help banks determine the number of losses or revenue they can generate on an account.
