Dexia Bank Uses SAS^® for Fast and Accurate Credit Scoring

Dexia Bank receives up to 700 requests for consumer loans a day, which is why the bank relies on an automated risk assessment tool. Thanks to the end-to-end use of SAS technology – from model building and data extraction to analytics for decision support – 85 percent of loan requests are processed automatically. Scoring parameters can be easily adjusted based on new legislation, market evolutions or litigation rates. As a result, the productivity of the Risk Management Department is on the rise, while the improved risk position benefits the bank's overall financial health.

With a broad offering of banking and insurance products and services, Dexia is Belgium's second largest financial group. Like many other financial institutions, Dexia Bank uses a credit scoring procedure to assess the risk of granting a loan to a customer. The sheer number of parameters involved makes this a time-consuming exercise.

Assessing Expected Default
Dexia Bank uses two different methods to assess solvency. The choice of method depends upon whether the applicant is already a customer or not. For new applicants, the bank uses a product scorecard based on static variables such as age, marital status, profession, type of product, loan duration and income. For existing customers, it relies on a behavioral scoring method that is used to estimate the expected default as required by the Basel II legislation. This gives the probability that a customer will be unable to pay back their debt within the next 12 months. It takes into account a greater number of variables including excess overdrafts, credit card usage, timeliness of reimbursements, and movements on the bank account.

"Our credit scoring method based on expected default analyzes all the information we have on existing customers," explains Sophie Labenne, risk manager at Dexia Bank. "The more we know about the applicant, the better we can assess their ability to pay back the loan. In addition, we must take into account the requirements set out by the Basel II legislation regarding the amount of capital and reserves a bank must keep to cover loan risks."

SAS Scores High
Given the vast number of parameters involved and the quantity of applications Dexia receives, they needed powerful software to help them make the right decisions faster. The application had to lighten the load for analysts by enabling the highest possible degree of automation in deciding whether to grant a loan or not.

Dexia Bank chose to build their scorecard with SAS. "We had been using SAS technology since the late 1980s for data-gathering purposes," notes Labenne. "Over the years, we added complementary SAS tools to exploit data and produce reports. With the recent addition of modules to develop, model, and monitor scoring systems, we now rely on an end-to-end SAS credit scoring solution that meets all of our requirements perfectly."

From Raw Data to Decision Support
The SAS solution implemented in the Risk Management Department of Dexia Bank first extracts raw data from all customer, product and financial databases, and then all this information is merged into a central server running SAS Mainframe. All programs and applications also run on this server, which increases processing speed and automation. In addition, SAS Enterprise Miner is used to develop and fine-tune the scoring systems.

To assess the risk related to each customer, Dexia Bank has developed the so-called ERIC (Evaluation Risk Client) application. The ERIC reports are generated using SAS. Depending on the scoring model, ERIC generates an automatic response: yes, no, or undetermined. In the latter case, the request is forwarded to a classical system based on a product scorecard or to an expert analyst. The target is to limit the number of undetermined cases in order to increase automation. At the same time, Dexia needs to maximize the acceptance rate while minimizing recovery litigation.

Close Monitoring
Monthly reports are generated in SAS to monitor product profitability and acceptance rates, and to follow up on the number of recovery actions. This rate is analyzed in detail for each scoring category. Whenever it becomes too high for a given score, the relevant variables are adjusted in the customer scoring application. These parameters are also adapted following evolutions in payment habits. For instance, the growing use of credit cards for online transactions and the gradual decline in the use of checks are both factors that influence the risk assessment of an applicant.

Rapid Adjustment to Changing Demands
"An important benefit of SAS is that it reduces the need for training to a minimum," adds Labenne. "The SAS solution, for instance, is a point-and-click application that is very easy to use. It allows us to create powerful scoring models without the need for advanced IT knowledge. We make the scoring models ourselves, choose the variables, and then ask the IT department to put the model into production. This is an extremely important feature since our scoring models need to be fine-tuned continuously in the face of growing competition and major changes in legislation such as IFRS and Basel II. In this respect, SAS provides us all the flexibility we need. The query features are equally simple to use. Graphic visualizations enable us to instantly spot which ratios and trends we need to focus on."

Preparing the Ground for Marketing Campaigns
The behavioral scoring system is also used by the marketing department for customer segmentation. It is extremely useful, for instance, in helping identify the most suitable customer profiles for commercial mailings. The system allows detailed profile analyses based on a customer's total assets and also enables correlation analyses between risk and return.

"With over 2.2 million Dexia customers, many of whom have several cards and accounts, you can imagine the quantity of information our marketing department has to deal with," stresses Labenne. "SAS predictive analytics helps us identify which of these customers are likely to be interested in a specific product. This significantly increases the success rate of our marketing campaigns, and prevents us from alienating customers that have no interest in certain products or services."

Optimal Balance Between Risks and Available Funds
The risk assessment application also benefits the entire institution by helping the bank stand on more solid ground. "All banks must comply with the Basel II requirements, but we try to go a step further by continuously improving both our asset and risk positions," Labenne concludes. "After all, an improved risk management strategy enables us to block fewer reserves. And that freed-up capital can be invested in turn to generate more revenue. It's a winning loop."

The results illustrated in this article are specific to the particular situations, business models, data input, and computing environments described herein. Each SAS customer’s experience is unique based on business and technical variables and all statements must be considered non-typical. Actual savings, results, and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software. Brand and product names are trademarks of their respective companies.

Copyright © SAS Institute Inc. All Rights Reserved.