This document discusses past, present, and emerging paradigms in risk modeling and pricing for lending. It covers: 1) Generic credit bureau scores have long been used for lending decisions but custom models using lenders' own data can better rank risk. Custom scores are now widely used, incorporating more contract data. 2) An emerging paradigm fixes a target probability of default and solves for loan terms (amount, APR, term) that meet that risk level rather than basing pricing on credit score alone. This allows matching customers to any vehicle, like a peer-to-peer lending platform for autos. 3) The "Big Data" revolution enables capturing more applicant information to further refine risk assessments and