Customer scoring solutions based on SAP Hana – a real-time approach to risk minimisation


When the bell tolled, ringing in the financial crisis, it also ushered in an existential issue for all banks: risk management. Banks now have less and less time to furnish different stakeholders with information, among them supervisory boards, boards of directors and rating agencies. As the rule books have expanded – including SolvV, MaRisk and the newly introduced Basel III guidelines – the regulatory authorities in particular are demanding adequate risk management arrangements.

What does this mean in practice? Before issuing credit, banks must examine the relationships between opportunities and risks, summarise these and then keep them under control. And no dilly-dallying! The more precisely and quickly the process is implemented, the more effectively any of the default risk associated with issuing credit can be minimised. At the same time of course, risk provisions can be kept down.

Customer scoring in real time

The customer scoring solution created by GFT makes it possible to analyse, simulate and validate creditworthiness factors in real time. The system is based on SAP Hana and it marks an important step forward in significantly reducing the time it takes to compute complex simulations. As well as improving the quality of financial consultants’ work, it also makes it easier to conform to legal requirements. Without SAP Hana, it would not be possible to evaluate the huge volumes of data involved, not to mention make precise forecasts regarding credit assessment.

Customer scoring: a continuous process involving five steps
Customer scoring: a continuous process involving five steps

The technical foundation: in-memory computing

In-memory computing makes it possible to process large volumes of data in real time. This is because instead of working on an application level the process works on a data level. One thing is important, however: first you carry out the calculations, then you transfer the data. This makes it possible to map what-if scenarios. No longer do programmers have to fantasise about aggregating millions of individual items of information in seconds. This is the new normal, as is importing the data and analysing it in real time.

The bottom line: benefits for customer scoring solutions on several fronts

The scorecard captures each parameter necessary to carry out the evaluation. Even if a single parameter changes, the financial consultant can immediately revise the result. This makes it easier to maintain an overview of a portfolio and actively control it. Also, because it’s now possible to tweak parameters in real time, the validation process can be accelerated. This is because of the in-memory solution, which ensures that data is processed as quickly as possible. As the analysis takes place in actual time, risk optimisation is improved and forecasting is much more accurate. The result is a professional report based on expertly analysed evidence in real time. This provides key decision-makers with a ‘dashboard’ of information, displaying the important risk KPIs at a glance.