SPSS - Customers

Cornèrcard

Situation

Established in 1975, Cornèrcard is a key player in the Swiss credit card market. A division of Cornèr Bank, a private and independent banking institution, it was the first company in Switzerland to issue Visa credit cards and introduced MasterCard to its portfolio in 1998.

Thanks to its extensive experience in the credit card market and a reputation for a high-quality service, the company has built a customer base of over 600,000 clients.

Challenge

A successful credit card company must be able to accurately assess its credit risk at any given time. Having an up-to-date picture of which clients may default on payments is critical to this process, not only among new applicants, but existing customers too.

Cornèrcard was using a rules-based system to drive its credit risk management processes. This was created around the company's own experience of delinquency, selecting data elements and creating rules to automate the credit decision process.

However, while the rules-based system helped speed up the credit assessment process, Cornèrcard believed it did not provide a high enough degree of accuracy and effectiveness to the credit evaluation process.

The decision was made to reinforce and improve the performance of the credit risk management system by opting for a behavior scoring solution. Such a system would be able to review thousands of credit risk data elements in order to find the most predictive data elements and develop models with which customers could be assessed. These data elements could then be weighted through scorecarding to further improve the effectiveness of the models used. Not only would this help Cornèrcard improve its credit risk management, but it would also help the company boost its bottom line.

"We felt that changing tack on how we approached credit risk management and looking at a behavior scoring system would help us to better identify accounts that are very likely to experience severe delinquency or possible credit loss in the future," explained Lorenzo Cattaneo, data analyst at Cornèrcard.

"By having better intelligence on which customers may default on payment, we can address potential issues faster, before they become a real problem," he added.

Solution

Recognizing SPSS Inc.'s leadership position in predictive analytics, Cornèrcard opted to work with the company on the initial scorecarding process and then build its entire credit risk management decision support system on SPSS Inc.'s predictive analytics technology.

During the scorecard development process, SPSS Inc.'s leading data mining technology was used to determine the predictive factors upon which the scorecard was built. These included cardholders' demographics, historical performance and ratios based on business knowledge.

Once this was complete, the SPSS Inc. technology was integrated with Cornèrcard's central databases to run the credit risk management decision support system.

Every 24 hours, customer data is downloaded from central databases into the SPSS Inc. software. This includes information from all 600,000 customers, each with several hundred input fields.

The SPSS Inc. solution analyzes the customer information, such as card usage, age, longevity, utilization rates and payment behavior, and converts this into a score, representing the credit risk of that customer.

Next, the score is combined with other account key indicators, such as delinquency level, credit limit utilization rate and type of account, and run through a series of decision trees to determine if the account should enter the debt collection process and the kind of approach to apply.Based on the highly-scalable SPSS Inc. technology, the entire process takes less than 10 minutes, so the information is available the next working day.

The collection department at Cornèrcard can then manage the credit risk of any accounts identified through this process, with particular attention paid to accounts with high balances and, of course, delinquent accounts.

Depending on the risk level identified, the actions of the collection team can take a number of forms. This might include a simple written reminder about the amount owed or more formal warning notices. Where necessary, the team can also block an account holder's card.

"Developing our own solution in-house was the right choice for Cornèrcard. Using PASW Statistics, we've been able to set up a credit risk scoring system that is faster, more cost effective and much more flexible to maintain than anything we could have found pre-packaged," said Cattaneo.

"PASW Statistics is incredibly powerful, enabling us to undertake advanced, intelligent analysis of customer data and use this information to the benefit of our business," he added.

Interested in Cornèrcard? Download the PDF

Results

The main purpose for developing the new credit risk scoring system was to improve the management of credit risk and the overall collection procedure. Early results have shown that Cornèrcard has reduced the unpaid debts by 15 percent or more.

As a result of the implementation, Cornèrcard is now able to better manage the daily collection activities by providing consistent, objective and automatic decisions based on a sound and ongoing risk assessment of its customers.

The company has seen a rise in the early identification of potentially problematic accounts and been able to cut the management cost of the collections process by taking a more efficient approach.

The new credit risk management solution is also providing benefits to other areas of the business and helping drive revenues and maximize profits. It has helped reduce cancellation rates amongst good cardholders who might have previously been incorrectly targeted as potential bad debtors. And, by a more targeted approach to customers, Cornèrcard has been able to increase revenues through up-selling and cross-selling activities by knowing when to offer account holders new products and services, such as gold cards or increased credit limits.