Solvency II in focus – Proving your Model, Now and into the Future
In the third part of the serialisation of a Solvency II study, published recently by Clear Path Analysis, a roundtable discusses how to ‘prove your model now and into the future’, and also how risk management could be affected by Solvency II.
The full report is available at https://www.clearpathanalysis.com/reports/solvency-ii-the-global-dimension-2012/
Noel Hillmann, managing director, Clear Path Analysis: How have your plans for data and risk reporting evolved in the past year and what are the main considerations for changes in the next twelve months?
Dan Wilkinson, head of enterprise risk, Liberty Syndicates: Undoubtedly data reporting has developed more formality. This means that data processes have been mapped out, the controls looked at and the materiality f the data flows assessed. We have also started to incorporate primary validation; where data owners design and follow a process to ensure that the controls are operating and that the data can be relied upon. This leads o the identification and reporting of data deficiencies to a monthly management committee.
We have recently reviewed risk reporting to better align it with the internal model. In part this has been done through the grouping of risks so that relatively immaterial risk (from a capital perspective) are not given undue prominence in the risk register. Risk event reporting is now more closely linked to trigger events and back testing that would provide suggestions for where a key assumption or parameter needs revision. We are also using risk reporting to flag when our expectation is, or could be, wrong with our internal model to create a feedback loop.
The role of ‘expert judgement’ (to augment or in lieu of data) is needed to justify why better data cannot be accessed to remove the reliance on subjectivity. Some expert judgement will always be needed for extreme tail events, dependencies and operational risks.
Others techniques are required for assessing parameters and methods used to approximate the risk profile.
Risk reporting essentially needs to move away from 1 in 200 quantification towards more foreseeable points in the risk distribution.
It is probable that 12 months from now, as data discipline increases, the reliance on end of process sense checks, reconciliation and comparison to benchmarks will evolve to be a more data driven validation. Solvency II directs you towards these data and process evolutions.