Data Goverance at the Data for ERM & Solvency II event

Finsbury hosted the session on Data Quality at the recent Insurance ERM Data for ERM & Solvency II event.    There were quite a few important discussions at the event,  which we are going to cover in a couple of blog posts.  There was an interesting session on what it took to achieve good data quality.    The discussion was dominated by governance.   Without effective governance and  management support it would not happen. 

 

Data quality matters for Solvency II because the directive explicitly requires data to be complete accurate and appropriate.   The insurance regulators have learnt from the Basel II experience and the banking crisis, where it became clear that many banks risk data was inaccurate. An informal poll of the audience concluded that less than 15% of the attendees felt that their organisation had effective data governance in place.   

The panel and audience discussed whether having a Chief Data Officer (CDO) was a viable option.   The argument for this is that it means that there is clear responsibility for data quality.  The rationale against is that it absolves other managers from having responsibility.   It may also be hard to find someone with the right skills to do the job or broad enough shoulders to take it on.

 

Our view is that there is a lot that can be learnt from other industries.   For example most trading operations have a product control or middle office function.  Their job is to make sure the data is accurate and that all the trades have been booked correctly.    If a product controller detects an error then this triggers a chain of activity to not only correct the immediate error,  but also investigate what went wrong and put in place any process or IT changes required to stop it happening again.   This type of approach has been shown to work.

So it is not just about having the right governance.  In most insurance companies good data quality requires changes to the organisation structure and some of their core processes.    To be effective the data governance team has to have the authority to get other departments to improve their own processes and to complete the right checks and balances at the right time.   This can be quite hard to get right in an organisation because it impacts, people, processes and of course the data itself.   

 

If you are working on data governance at an insurance company, please contact us if you would like a chat about data quality.  We will be happy to share our experiences with you.   You can contact us here.