HEWLETT-PACKARD WORKSTATIONS AT WORK AT... Credit Lyonnais -- Paris, France Credit Lyonnais has learned the value of using detailed statistical analysis to play a key role in focusing the bank's retail lending and credit strategies. As one of Europe's largest retail banks, Credit Lyonnais is aware that the ability to fine tune its policies in this way will have a significant impact on the profitability of its lending and credit operations in retail branches. Based in Paris, the Groupe de Recherche Operationnelle (GRO) is a small specialist team of mathematicians and statisticians that operates within Credit Lyonnais as an in-house consulting service. GRO's ability to refine a mass of statistical data into a meaningful analysis upon which to base policy decisions has made its work of value in many areas of the banks' activities. In addition to developing the bank's retail credit scoring methodologies, GRO is also involved in the evaluation of new capital markets products and financial instruments, and in the targeting of direct mail strategies to maximize response. Sophisticated statistical analysis All the GRO's research projects begin with large volumes of statistical data which must eventually be refined into a practical strategy that will help to optimize the bank's performance while minimizing its risk exposure. Bruno Moatti, in the GRO, explained a typical project: "In credit scoring we will usually analyze a sample of 4000 customer profiles, split equally between good and bad risk accounts. "The data contains raw information on the customer, such as income, age, family details, employment details, and the history of their account behavior -- such as if they make regular savings, maintain an overdraft, or have ever exceeded an agreed overdraft limit. Certain criteria such as the sex of the account holder are not allowed to be used. "We then undertake a variety of both factor and discriminant analysis. The process takes some time and involves the continuous analysis and re-analysis of the data, until we have arrived at the optimum series of results. This information is then distilled into a simple-to-use method of credit scoring which the bank's personnel in its many retail branches can use for approving credit and loan applications." Assuming a strategic importance With retail banks throughout Europe forced by the difficult economic circumstances to constantly monitor their lending policies, the work of the GRO has recently assumed even greater importance. In many instances the difference between profit and loss in a bank's retailing operations is the ability to control risks and reduce the level of bad debt. The GRO's credit scoring analysis allows Credit Lyonnais to be more sensitive in identifying and measuring potential risks, offering the bank a vital competitive edge in assessing which of its customers are likely to be good or bad risks. To assist them in the manipulation and analysis of data the GRO uses the SAS System, a sophisticated software application from the SAS Institute. This gives users complete control over data access, management, analysis, and presentation. Productivity benefits Bruno Moatti finds one of the key benefits of the SAS System lies in its powerful data management facilities. #Using our previous software, we needed to pre-program new data files in Fortran. This could take several days just to prepare the files for analysis. Using SAS, the same task takes only a few hours. This means that we are able to spend more time refining the raw information and conducting even more thorough analysis, rather than using our time as programmers." Added power of new workstations The GRO system consists of a network of HP Apollo workstations. They are linked together on a local area network with a number of PCs to help provide wider access to the group's information resources. Credit Lyonnais was among the first of the leading French banks to use HP's Series 700 family of high-performance PA-RISC workstations and they are now used for a number of applications within the bank. The added power that is a notable feature of the HP workstations has also proved highly valuable, as Moatti confirms. "With the high performance of the HP workstations, we are now able to handle larger samples more efficiently. The fact that we are able to base our work on a larger cross-section gives it a stronger empirical basis." Because the underlying economic circumstances which influence both the bank and its customers are constantly changing, the GRO's research is also continually monitored and reviewed. A policy that is ideal today, may well become inappropriate in a few short months. The productivity gains offered by the new system allows the GRO to concentrate all its efforts on keeping the analysis up to date. Presentation of data - the SAS System Dealing with such a complex array of statistical information, it is not surprising that the GRO, like many other users, enjoys the SAS System's wide range of data presentation facilities. A variety of presentation features help to communicate the significance of complex and sophisticated research through simple graphic displays. An integrated allows users to create their own reporting models, templates, and graphs. These features have established the SAS System as a market leader in data analysis and presentation. #####