Insurers are complying with Solvency II but data is proving problematic, according to Craig Skinner, leader for insurance and investment management at PwC.
A review of insurers in the Internal Model Approval Process (IMAP), published by the Prudential Regulation Authority (PRA), was designed to assess the quality of data being used.
Skinner believes that although insurers have made strides in complying with Solvency II, many are struggling to articulate, design and implement effective controls over material data.
He said: “Organisations are creating, collecting and storing ever increasing amounts of more and more complex data. Coupled with the complexities of the processes and using multiple—and often legacy—IT systems, firms must apply ever more controls and ongoing assurance over material data in a pragmatic but rigorous manner.”
He believes that insurers that treat their data as a strategic asset will find success and lay the foundations for achieving Solvency II compliance as well as address other regulations and legislation.
He explained: “The General Data Privacy Regulation taking effect in 2018 will, on top of all the work completed to date regarding Solvency II, require firms to provide even more robust and transparent assurance over their data.”
Yalini Pathy, UK insurance investment management data governance leader of PwC, added: “The PRA findings are intrinsically linked and highlight the need for an effective data governance framework driven by bringing structure and consistency to an organisation’s approach to data.”
Pathy suggests that firms must be careful to consider design principles to enable a solution is relevant and pragmatic such that it can be successfully embedded.
She claims: “Data governance, when pragmatically applied, does underpin wider commercial benefits. When data is protected and managed appropriately, there are benefits far broader than simple Solvency II compliance.”
“The PRA findings complement our perspectives and experience, from supporting clients in the market, on the governance and controls that are to be applied to data used in the internal model.”
