Exploring the impact of advanced analytics on op risk
- 2 April 2019
The application of advanced analytics – artificial intelligence and machine learning – is increasingly part of everyday life. It's used everywhere, from our smartphones and social media, to autopilot on aeroplanes and virtual assistants such as Alexa or Siri. Financial companies are already using this technology to monitor transactions and spot suspicious transactions. But what impact will this have on operational risk? And how can we use it to enhance risk management?
What is certain is that AI and machine learning will have an impact on operational risk. At ORX, we believe that advanced analytics will become a core part of any strategy for the future management of operational and non-financial risk. This view was echoed by Chief Risk Officers (CROs) in our recent study on The future of operational risk. The industry needs to be ready for this.
"CROs identified advanced analytics (beyond capital modelling and stress testing) as a potential game changer for operational risk management and around half earmarked it as a priority for innovation and investment over the next five years." The future of operational risk, a study of 15 CROs by ORX and McKinsey
Sharing expertise in a new working group
To help operational risk managers be prepared for this, ORX has recently launched a working group to specifically review machine learning and its application to operational risk. Our Machine Learning Working Group is made up of experts from ORX member firms who have a background in advanced analytics. They'll share common challenges, identify potential solutions, and investigate the potential for collaboration. They'll also help to guide any future research we carry out on this topic. We will create clusters of terms with similar meanings, and then analyse these to see whether they have any relationship to losses.
New research on using machine learning to get more from op risk loss descriptions
Alongside the work that our machine learning group will do, we're also conducting a project on text mining later this year. We will look to apply machine learning techniques to operational risk loss descriptions. It's possible that historical event descriptions could contain a wealth of useful information – for example, material risk drivers or event causes. While it would be very resource intensive to manually review these historical descriptions, this could be an opportunity to use advanced analytics.
Working groups and research from ORX
This machine learning focus is just one of many detailed research projects and peer working groups offered by ORX. To find out more about how you could benefit from any of our offerings, contact us today: [email protected]