Ask the expert
Risk specialist Bart Baesens on credit risk, credit scoring and Basel II
Bart Baesens, PhD, started his academic career as a researcher in data mining for credit scoring and customer relationship management applications. As a former SAS Student Ambassador, he presented his research findings at the 2002 SAS Forum International conference, which strengthened his partnership with SAS. Currently, Baesens is a Lecturer at Katholieke Universiteit Leuven in Belgium and the School of Management at the University of Southampton in the UK.
Baesens also is a guest lecturer for SAS, teaching Credit Scoring for Basel II and Advanced Customer Analytics. These courses are offered as part of the SAS Business Knowledge Series around the world.
Students and industry experts alike praise Baesens for his knowledge in the areas of risk compliance and analytics. “His inspiring and profound industry knowledge, as well as his enthusiasm for SAS products and solutions, have contributed significantly to SAS’ credibility in the Basel II space,” says Lieve Goedhuys, Marketing Manager in SAS Belgium.
We caught up with Baesens while he was in Hungary teaching a SAS course and asked about his thoughts on credit risk technologies.
What is the role of IT in credit risk, credit scoring and Basel II?
One of the major problems that many banks are facing nowadays is that data is geographically distributed throughout the enterprise. The first role that IT can play is to set up a data infrastructure, for which SAS has provided some solutions, such as SAS Banking Intelligence Solutions. So you set up a data infrastructure in order to gather all this data into one coherent and consistent data format. And that data architecture can then serve as the input for developing all the Basel II rating systems. The role of IT does not stop there, though, because IT needs to provide facilities for implementing the rating systems and for presenting the results of these systems. IT also needs to provide support for monitoring these rating systems and keeping track of how these systems perform. In that area, the role of business intelligence is becoming very important.
What are the current issues in Basel II, credit risk and credit scoring?
Many issues center on data. Banks need to develop rating systems, but they do not have access to the data. There are a couple of very specific modeling issues, and the first step of modeling is providing data. Next, banks need to think about the number of defaults because Basel II is about modeling the default behavior of customers. Some banks don’t have observations on their default behavior, which makes it very hard to develop these models. Another issue in Basel II is how to interpret the guidelines because many of these guidelines have been formulated rather vaguely. That is why it is very hard for banks to put the guidelines into operation and to implement compliant credit risk models that conform to those guidelines.
How do you see banks managing their risks because of the Basel II regulation, or is it more a business issue already?
Banks are becoming a lot smarter. Basel II has put forward a couple of guidelines that not all banks agree with. Basel II basically tells banks how to calculate the minimum amount of capital they need in order to protect their depositors. This is what we call regulatory capital. But banks are also calculating economic capital, or so-called business capital. This is a second amount of capital, which according to them most reflects the risk they undertake. One of the key ideas of Basel II is that you should always model credit risk from a conservative perspective. However, the danger exists that the scenarios assumed are so conservative that they no longer reflect reality.
What are the different trends and practices in measuring risk in Eastern and Western Europe, Asia and the US?
Basel II has been implemented in slightly different ways in different countries because there are different regulations. I’ve taught courses in Asia, Europe and the United States. I think the Western European countries were among the most advanced. Many of them have Probability of Default (PD) models, Loss Given Default (LGD) models and Exposure at Default models in place. I think the Eastern European countries will be coming behind. Most of these countries now have PD models and are starting to think about LGD models as well. The major US regulatory authorities only recently adopted Basel II. Now all big American banks need to adopt Basel II as well, and that is why I am heading back there soon to teach four courses.
Are there specific programs regarding Basel II at universities?
Basel II is an area that is not taught at universities a lot. I have a course at the University of Southampton School of Management, where I teach credit scoring, but credit scoring is only one part of Basel II. At the same time, I do think there is demand from the industry for graduates with this kind of knowledge. I get phone calls and e-mails every week asking, “Do you know good students who could help us do LGD modeling and PD modeling?”
When you develop your curriculum, do you take business expectations into account?
My courses are very much business oriented. The first time I taught this course, I did not have that much business experience because I came from the university. Furthermore, I collaborate with a lot of financial institutions, doing projects or validating and implementing Basel II models, which also has given me a lot of business knowledge in the meantime.
What makes your course so popular?
First of all, the current topic: the Basel II Capital Accord, and because in the course I use SAS as a tool. I teach theory in the morning, and then in the afternoon I teach how to illustrate and implement all these concepts in SAS.
Why do you prefer SAS® for risk analysis?
SAS has the nicest environment to work in. It is user-friendly, powerful and an end-to-end solution. I can do data preparing, model building, model evaluation and more. I can do everything with one application
Bart Baesens, PhD, is a Lecturer at K.U. Leuven and the Southhampton School of Management and a guest lecturer for SAS.