By Barney Wanstall, Director, Insurance, PwC
It is rumoured that no firms actually fell at the final hurdle and that the panels approved (or more accurately, gave conditional approval) to all of the models that came before them. It is difficult to tell definitively because many models are still subject to Group approval from another regulator and because so many firms have dropped out along the journey (circa 86) - whether they left in the final or penultimate stage is not of great relevance.
Model approval bestows more freedom on the firm however, quite rightly, supervisors wish to reserve the powers they do have over approved firms / models. Some approvals were also “downgraded” from full to partial and the power over model change is well known - all major changes have to be approved by the supervisor. In addition, supervisors will want to keep on top of all changes in the model (major, minor or data driven), especially those which have any impact on the SCR. Most major firms expect to have some major changes every year, the supervisors’ ability to approve or reject these therefore means that the formal IMAP approval is the start of a process of continued close scrutiny (perhaps annual re-approval in some circumstances) rather than a license to drive unsupervised. In addition, many of the submissions would have been made on 2014 year-end data and models will have been 'frozen' for the last six to nine months while the application was in progress, so they now need to be updated.
A number of approvals have also come attached with conditions as part of their internal model transition plan. This feedback must be considered and addressed as part of the ongoing regulatory relationship. The approval is also generally subject to the firm continuing to meet all of the requirements, failure to do so will obviously lead to regulatory action.
So, reflecting back on the IMAP process, what are some of the lessons we can learn from those firms who have seen this process through to the end?
1. It will take longer than you think: Don't be too ambitious or underestimate the resource and time commitment the process requires – many did in the first phase and often spent a multiple of their original estimate in terms of both time and expense.
2. Documentation: A boring fact, but documentation is key, the model is only as good as how it is documented and recorded. Once the model is built and fully validated, bringing together all this documentation and getting it through the necessary quality review and approval is challenging and time consuming – but also essential. This is especially true when it comes to being transparent on model limitations, and how you deal with them.
3. Board approval: The Board must approve the Internal Model before the application is submitted. Board approval is not a rubber stamping exercise – to do this properly requires months of effort to train and guide the Board through the model and the approval being sought. Allow at least 6 months.
4. Models appreciate - a used model is worth more than a new one: Elements of model use have a long lead time, the most valuable uses connected with strategy and business planning require comprehensive training on the model too. You can’t fake use so start early (you’ll need at least a year) so you have time to build credible model use evidence.
5. Document as you go: Developing IMAP documentation as part of model development helps capture much of the underlying thinking and justification needed for the IMAP process – if you don’t document at the time when decisions are made it can be almost impossible to go back.
6. Show how it all fits together: Such is the volume of documentation connected with the IMAP process that many firms created a documentation hierarchy that demonstrated how all the materials fitted together. This was very helpful in mapping and navigating the evidence presented.
7. Develop an intimate familiarity with the SAT early: Understanding of the SAT in detail needs to be spread far and wide – if the whole picture is confined to a small number of people pulling the application together there will much more risk that the underlying evidence contains anomalies.
8. Get it reviewed: Validation of the model is well talked about, but assurance on the IMAP submission is not. The firms who had an independent QA process found it invaluable. It is easy to get lost in the detail of a SAT - having an independent quality review to ensure the programme is on track and documented output meets the relevant requirements has been vital for ensuring applications proceeded on schedule and meets the standards required.
Wave 2 firms will doubtless find the process much easier – many of these lessons were hard learned by firms who fell out of the process or experienced significant setbacks but the recent success shows it is achievable. The PRA also now has far more experience which will make the approval process run more smoothly - now might just be the time to follow in their footsteps.
|
