Risk-Based Pricing is the practice of setting the interest rate, fees or premium for each customer according to the credit or insurance risk that customer presents, instead of charging everyone the same price. A loan's rate is built from cost of funds, expected credit loss, operating cost, capital cost and margin, with expected loss taken from PD multiplied by LGD and borrowers grouped into grades with rate bands. A flat price overcharges low-risk borrowers, who leave, and tilts the book towards riskier grades. In the US, the Fair Credit Reporting Act requires notices for less favourable terms and the Equal Credit Opportunity Act bars pricing on protected characteristics; the EU AI Act treats credit-scoring systems as high-risk.
How risk-based pricing works in lending
A loan's price is built from components: the lender's cost of funds, expected credit loss, operating cost, the cost of capital held against the loan, and a profit margin. Expected loss comes from probability of default (PD) multiplied by loss given default (LGD), both produced by credit scoring and underwriting models. Borrowers are grouped into risk grades, and each grade gets a rate band.
Risk-based pricing example
Illustrative: a lender's cost of funds is 4%, its operating cost 2% and its target margin 2%. A grade-A borrower with a 1% annual PD and 50% LGD has an expected loss of 0.5%, so the rate is 8.5%. A grade-D borrower with a 6% PD and 60% LGD has an expected loss of 3.6%, so the rate is 11.6%.
Charging both a flat 10% would lose money on grade D and overcharge grade A. Grade-A borrowers would then move to a lender that prices them correctly, leaving a book weighted towards the riskier grade: adverse selection.
Pricing also changes who accepts the offer. Higher rates can attract borrowers who are more likely to default, so lenders validate PD estimates on loans actually booked at each price, not only on applicants.
Rules on risk-based pricing
In the US, the Fair Credit Reporting Act's risk-based pricing rule requires a notice when a consumer receives credit on materially less favourable terms than other consumers because of their credit report; lenders can instead give every applicant a credit score disclosure. The Equal Credit Opportunity Act prohibits pricing differences based on protected characteristics such as race or sex.
In the EU, an AI system used to evaluate the creditworthiness or establish the credit score of individuals is high-risk under the AI Act. Its obligations apply from 2 December 2027, after the Digital Omnibus on AI moved the date from 2 August 2026. Under the Court of Justice's SCHUFA ruling (C-634/21, December 2023), a score that a lender relies on heavily is itself an automated decision under Article 22 GDPR. The applicant then has the right to human intervention and to contest the outcome.
Risk-based pricing in insurance
The same logic sets premiums by claim likelihood. Telematics motor policies, for example, price on observed driving data such as speed, braking and time of day rather than on age and postcode alone, and can reprice at renewal as that data accumulates.
For a buyer of a lending platform, the pricing model is the revenue engine. Diligence compares the PD each grade was priced at with realised defaults by vintage, and checks whether model documentation will meet AI Act requirements by December 2027.