The process of determining the creditworthiness and forecasting future performance of existing or prospective consumer or small business (SME) borrowers or lessees through the use of historical data and statistical techniques is credit scoring, used primarily with consumer and small business (SME) credit applications.  Credit scoring systems (CSS) are largely statistical regression credit models that quantify and show the degree of the probability of the default of a credit applicant on a credit transaction against different risk criteria to determine whether the transaction is likely to be sufficiently profitable to be acceptable.  Credit scoring is also used to confirm internal ratings and to supplement judgmentally assigned ratings.

Credit scoring criteria are individually set by each creditor (lender, lessor) and depend on the size of the transaction, asset type, type of customer, credit history and numerous other factors.  The data are converted into numbers that are added together to arrive at the credit score.  If the total score is above a certain value, credit is granted; if it is below that value, the application is rejected.  Scorecards can show the degree of a transaction’s probability of default, not only the probability of default.

FICO Credit Score Breakdown
Criteria Weight
Types of Credit in Use 10%
New Credit Opened 10%
Length of Credit History 15%
Amounts Owed 30%
Payment History 35%

Credit scoring models can be differentiated from credit-risk models in several ways:

  1. Credit scoring models are largely statistical, regressing instances of default against various risk criteria, whereas credit-risk models directly model the default process calibrated to available data;
  2. Credit scoring underlies app-only programs using only the data submitted on credit applications, while credit-risk modeling relies on the analytical study of financial statements and other microeconomic data;
  3. Credit scoring is usually used for smaller credits of individuals and small businesses (SMEs), whereas credit-risk models are applied more to larger Commercial and sovereign credits.

Related Items