Behavioural Scoring

Behavioural scoring is applied to data collected at the point of sale for a consumer credit providers “through the door” population.

Proaxiom applies multivariate statistical techniques that have become the industry standard among scorecard developers for more than three decades. Proaxiom also have a participative approach to scorecard building and welcome client participation throughout all project phases. This ensures that Proaxiom scorecards are developed to the highest analytical standards and provide the optimum business outcomes for our customers.

Behavioural scorecards are designed and built using multi-variate statistics to predict the future behaviour of a customer based on past actions. This is based on data a customer provides from their completed application forms in combination with their credit bureau file and applicable data from a range of other sources.

Behavioural scorecards are highly accurate and have been proven to reduce the occurence of bad debt by between 15 to 25% over the traditional ‘one shoe fits all’ approach.

A scorecard is no more than a statistically sound set of business rules. It may be as simple as “if a customer does not meet the minimum repayment amount then refer file” or it may be as complex as hundreds of rules. It isn’t a standalone process. If a scorecard is to work for the business it needs to be properly developed and integrated into that business.

There may be better solutions than building a scorecard

Most business advantage is gained not from having a scorecard per se but from the rules and processes that the business implements around that scorecard.

Proaxiom has over 22 years’ experience specializing in understanding both business and credit processes. Proaxiom is not just a “scorecard building company” and because of this it is free to consider and recommend the solutions that best suit the individual business. This may or may not involve the implementation of a scorecard.

A typical application scorecard project occurs in a number phases:

  • Sample Design
  • Preliminary and Sub-population Analysis
  • Finalisation
  • Scorecard Delivery
  • Documentation Delivery

As a specialist in the credit industry Proaxiom has developed, in conjunction with expert statisticians, a number of customized scorecards in the past. In each case, these scorecards have been based upon each individual company’s data, developed solely for each individual company, and implemented successfully into each individual company. This approach avoids a “one size fits all” attitude.

Proaxiom supply Generic Scorecards to organisations who are not yet able to adopt fully customised solutions. This occurs when either the target portfolio is too small for a customised model, or when an organisation wishes to start lending in an area where they have no previous experience.

The duration of projects depends on the quality of data and the number of scorecards identified during sub-population analysis. The average project duration is 6 weeks from sample design to scorecard delivery.