Credit Scoring

Credit scoring is used to evaluate the potential risks involved in lending money, and in particular credit. The Adastra solution is ideal for companies which have a large number of reltively anonymous customers such as credit institutions, banks, insurance companies and telecommunications operators, making it virtually impossible for the corporations to know a great deal about these customers. A credit score, computed by data mining models using a customer’s credit file, estimates the risk of an applicant failing to repay a loan or other debt.

There are two categories of credit scoring:

• Application scoring is used to approve a new client’s application for a product or service. The score is based on the application data (socio-demographic profile and characteristics of the required product) and information from public records.
• Behavioural scoring (or performance scoring) assesses current customers by analyzing a customer’s financial history in terms of transactions such as paying a loan back on time. The score is primarily calculated from a customer’s record with additional information from public records.

The output of the credit scoring produces a set of rules known as a scorecard.

Adastra’s added value comes from a model that improves considerably on how to distinguish a good applicant from a bad one leading to a decrease in the number of bad debts (loans, frauds) while having the same percentage of refused debts (thus increasing profits and reducing prices), or, leading to the opportunity to give more credit while maintaining the same degree of risk. These variants are based either on cutting costs or a growth in earnings, which, in both cases, helps to generate greater profits. As a result, return on your investment comes rapidly.

A scorecard can be used in a variety of ways such as approving a loan, deciding whether or not a subsidized flat-rate telephone should be approved, or reducing the likelihood of a loss.