Applying intelligent decision support systems to the reorganization of lending processes

Traditional lending is very much the result of close cooperation between the customer, consultant, and lending department representative, but the number of loans issued without the customer and bank having any personal contact is rising. The transparency of applicable terms and conditions created by online publications and comparison websites encourages potential borrowers to shop “anonymously” and to take out loans from other banks instead of their main bank.

 

Banks are faced with an obvious challenge in these cases: an applicant who is unknown to them and without an internal credit history has to be assessed as quickly and objectively as possible. This sales channel simply leaves no time for individual assessments, which are often long-winded and in some cases end up in decisions that are based on nothing more than a “hunch”.

 

Modern mathematical-stochastic methods, so-called scoring, can objectify and support lending decisions. The loan application is assessed on the basis of existing application data. The results of external credit ratings and information provided by public authorities also flow into the decision. Depending on the score, a recommendation can be made on whether to approve or decline the application.

But if a system were to make these decisions entirely autonomously within a pre-defined framework, the lending processes would be completely streamlined. This would eliminate the routine tasks of the loan officer and considerably reduce the time taken to arrive at a lending decision. The resulting savings would have a positive effect on the margin.

It is advisable to work with a credit bureau with experience in IT and processes to structure the complex lending decision process.

 

CRIF is a leading provider of risk management solutions and has both the required know-how for modeling customer-specific scoring systems and an extensive pool of information on credit ratings and data provided by the authorities. CRIF also provides tried-and-tested software applications for workflow management and the analysis of decision-making processes. Cost-optimized CRIF and public information is accessed at various points during the automated decision-making process. Depending on requirements and the area of application, data validation and the results of fraud pool queries or specific product conditions can be integrated in the decision-making process. The CRIF decision support system feeds back an assessment that takes a configurable acceptance level into account. Potential outcomes of the decision-making process are either “declined”, “approved” or “transfer application to an expert team”.

 

Combining the use of innovative assessment methods (scores, credit rating information) and intelligent workflow management is essential for realizing the potential for increasing the efficiency of the lending process. Inconsistent data from various sources should be replaced by an organized and transparent information infrastructure. CRIF AG has showcased its expertise in numerous customer projects and supported many lenders in realizing efficient lending processes in the retail and corporate segments.