Data Process Challenges

What are the most important aspects of managing risk to consider for managing data operations?

One of the main challenges is understanding where the problems are in getting data right to input into risk management calculations. Many risk managers cannot see where the data is coming from because the fields they pull information from may be filled with dummy codes.

When trying to run a new risk analytics report, the set-up is not always from the same sources. So teams sometimes have to put together workarounds to patch that data together, but inaccuracies may then go into those data fields. That's where we see a lot of process changes and challenges. The bigger challenge is understanding the data you have, where it comes from and where the problems are, because it might not be immediately obvious.

What is the best way to handle data to support risk management?

There tends to be a disparity between how you approach data and the way that data is managed. Risk calculations vary as well, so there are different focuses. It's very hard to develop best practices when there is such a lack of standardization.

In an operational and cultural sense, end-users giving input on what they want data to look like at the end of the process means you need consistency in ownership of that data.

How are stress-testing regimens and requirements affecting risk data management?

People pay lip service to BCBS 239. They're probably struggling because it's quite a big project to take on. Risk is pulled from all types of different data sets, so it's not an easy project by any stretch of the imagination. Anyone trying to do a six-month project to get information right ... is never going to get it completely right the first time. It would have to be an ongoing program to try and proof risk data.

A lot of firms are trying to address risk management issues post-crisis. They don't have massive budgets and most are trying to make do with what they have.