Credit scoring

Credit scoring is the algorithm that turns a borrower's history into a single number predicting repayment probability. In the US that number is usually a FICO score (300–850), built from payment history, amounts owed, length of history, new credit, and credit mix — the weights haven't changed much in decades because lenders trust stability. The interesting fintech angle is alternative data: for people who are "thin-file" or "no-file" (young, immigrants, cash economy), traditional scores are useless, so platforms use bank transaction data, rent/utility payments, phone bills, or even psychometric surveys to underwrite. This is what lets BNPL, neobanks, and lenders in emerging markets serve customers a traditional bank would reject — they're not smarter, they're just looking at different data.