“Beware the Dog that Didn’t Bark,”Report on QIS5 Results

Location: London
Author: Tricia Holly-Davis
Date: Monday, April 18, 2011

“It is disappointing that EIOPA does not seem to accept that the introduction of a robust USP approach for catastrophe risk could significantly improve the appropriateness of the standard formula whilst giving regulators the comfort that insurers are not able to cherry-pick the lowest catastrophe model results.”

The Willis Re report, “Beware the Dog that Didn’t Bark – Is the European Insurance Industry Really Ready for Solvency II?”, found that the QIS5 study showed a smaller than expected increase in the number of companies failing to meet their Solvency Capital Requirement (SCR), the threshold below which regulators will be required to intervene.

While this figure rose from 11% for 2008’s QIS4 to 15% for QIS5, Willis Re noted that the increase was not as dramatic as expected considering the impact of the financial crisis on insurers’ balance sheets, the sharp increase in standard formula risk factors used in QIS5, and the doubling of the number of small insurers taking part in the exercise. (Small insurers were expected to fare comparatively badly due to their limited diversification).

On a whole, Willis Re said that the QIS5 study, conducted by the European Insurance and Occupational Pensions Authority (EIOPA) between August and November 2010, points to the (re)insurance industry being comfortably well-capitalised and in relatively good shape for the implementation of Solvency II in 2013.

However, a closer analysis by Willis Re of these unexpectedly benign results, finds that:

·         Many participants felt that a number of SCR sub-modules were disproportionately complex which led to widespread use of simplifications. A notable example of this was the counterparty default risk sub-module.

·         The catastrophe risk sub-module was the most criticised. Respondents’ feedback was, to some extent, contradictory as they complained both about the complexity of data requirements and the limited sensitivity to actual risk exposure. However, the latter issue cannot be addressed without further increasing the former. Willis Re believes that a fundamental review of CAT risk treatment is needed – extending the concept of Undertaking Specific Parameters (USPs) to the catastrophe calculation.

·         The number of insurers submitting internal model results was too small to draw general conclusions. However, the results indicate that internal models did not provide materially different outcomes from the standard formula. Quite surprisingly, a significant number of insurers that have already applied for regulatory approval did not use their models in QIS5, for example in the UK.

Commenting on the report, David Simmons, Managing Director, Analytics and Head of International Risk Management for Willis Re, said: “Many companies, large as well as small, struggled with the complexity of QIS5. Some parts of the standard formula, in particular the catastrophe risk module, do work for many companies’ risk profiles. The answer may be internal capital models, but QIS5 shows that companies are making slow progress in this area. It is also debatable whether national regulators have the resources to deal with the increasing number of internal model approval requests.”

Ian Cook, Managing Director, Analytics and Chief Actuary for Willis Re International & Specialty added, "It is disappointing that EIOPA does not seem to accept that the introduction of a robust USP approach for catastrophe risk could significantly improve the appropriateness of the standard formula whilst giving regulators the comfort that insurers are not able to cherry-pick the lowest catastrophe model results."

Whilst the European Commission has raised the possibility of a gradual implementation of Solvency II requirements over a period of up to 10 years, giving companies and regulators some breathing space, Willis Re believes that the relatively benign QIS5 results may remove the pressure for a long transition period.

To subscribe or visit go to:  http://www.riskcenter.com