In detail
A scoring model is a type of algorithm that evaluates credit report data and other information to produce a credit score. These models are developed by independent companies, such as FICO and VantageScore, and are used by lenders, landlords, and other entities to assess the likelihood that a borrower will repay a debt as agreed. The model assigns weights to various factors found in credit reports, including payment history, the amounts owed, the length of credit history, new credit inquiries, and the mix of credit types. Different scoring models may emphasize these factors differently, which is why a person can have multiple credit scores that vary from one another. The exact formulas and weighting schemes used by scoring models are proprietary and not disclosed to the public. Scoring models are typically trained on large sets of historical data to identify patterns that correlate with credit risk. Once developed, a model is applied consistently to credit report data to produce a score, often within a range specific to that model. Lenders may choose which scoring model to use based on their own underwriting criteria, and some may use custom models developed in-house. The Fair Credit Reporting Act governs how consumer reporting companies handle the data that feeds into these models, including requirements for accuracy and dispute resolution. Because scoring models are updated periodically to reflect changes in data and risk patterns, the same credit report data may yield different scores when evaluated by different models or different versions of the same model.