Solvency II: Towards implementation
Since late 2005, the Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS) has been involving the European insurance industry in a consultation process to test the impact of proposed Solvency II regulations, known as Quantitative Impact Studies (QISs).
The QISs are performed by companies on a voluntary basis. These have been run by CEIOPS, at the request of the Commission and are the primary means for testing the design of the future standard formula for required capital, as well as the main route for determining the correct calibration. The QISs are also instrumental in collecting data on the potential impact of the new regulations.
QIS 1 was intended to test an approach for setting the technical provisions, while QIS 2 focused on a methodology for assessing required capital (the Solvency Capital Requirement, or ‘SCR’). QIS 3 focused on improving the calibration of the standard formulae for the minimum capital requirement (MCR) and SCR calculations.
CEIOPS released the draft specifications of QIS 4 on 21 December 2007. QIS 4 was officially launched in April 2008 based on these specifications and will run until 7 July for solo companies and until 31 July for groups. See figure 1.
CEIOPS has also recently released Consultation Paper 23 (CP 23) that gives guidance on how to determine the technical provisions (Best Estimate and Risk Margin) when data and/or expertise are missing in the company. These simplified methods will also be tested in QIS 4.
Overview of QIS 3
Overall, QIS 2 participation was good but QIS 3 was even better — to the point where it almost doubled compared to the previous study. This shows the high interest from companies for Solvency II. See figures 2 and 3.
QIS 3 was the first study to have focused specifically on calibration and it has provided a wealth of information on a wide range of elements of the Solvency II proposal. For some elements, multiple approaches were tested with the aim of being able to choose among the approaches after its completion. The main objectives of QIS 3 were to:
>> Measure the financial impact on balance sheets of the adoption of QIS 3 specifications
>> Collect feedback from the participants on the practicality and suitability of the calculations
>> Test the effects on insurance groups for the first time.
Financial impact
The numerical results of QIS 3 provided some interesting insights into the potential impact of Solvency II on the regulatory balance sheets of insurance companies.
>> Technical provisions tended to decrease, with the release of the prudence margin in Solvency I being greater than the additional risk margin under Solvency II.
>> This tended to be more than matched by an increase in the capital requirements and, on balance, the overall solvency ratio tended to decrease. This was most significant for non-life companies, where the solvency ratio decreased for the majority of participants. Life companies showed a combination of increases and decreases in the solvency ratio.
>> QIS 3 showed 98% of firms meet the MCR and 84% of firms meet the SCR, so a large-scale capital injection is unlikely. However, some reallocation of capital may be needed, reflecting the increased risk sensitivity of the proposals.
In addition to the CEIOPS assessment, the Chief Risk Officer (CRO) Forum, an industry body representing the CROs of the 13 largest European (re)insurers, benchmarked the results of the QISs to the internal models of its members. They found that the QIS 3 specification tended to result in an SCR circa 40% higher than internal models, largely due to differences in the calibration of the insurance risk SCR.
Feedback on practicality and suitability of the calculations
QIS 3 results were largely well received and the participants were generally positive about moving towards market-consistent valuation and risk-sensitive solvency requirements. However, there were a few key areas that participants felt should be addressed:
1. Risk margin.
It was suggested that the risk margin approach should be more clearly defined and that market risks should be excluded to avoid double counting. QIS 4 addresses these issues by providing guidance on simplified approaches for projecting the SCR and changing the specification to exclude market risks. Participants felt that the 6% cost of capital rate was too high, however, CEIOPS has not addressed this in QIS 4, pending more industry research into an appropriate factor.
2. Reduction for profit sharing.
The method proposed for assessing the risk-absorbing ability of future discretionary profit sharing (known as the ‘KC’) was strongly criticised as a key weakness of the QIS 3 specification. This method was based on calculating all SCR components assuming no change to discretionary profit-sharing benefits in stress scenarios, aggregating these, and then reducing the overall SCR by a factor to reflect the risk-absorption capability of future profit sharing. The calculation of this factor proved to be problematic and QIS 4 has moved to a calculation of SCR risk categories ‘net’ of management actions that is more aligned with current market practice.
3. Operational risk.
The QIS 3 feedback was that the formula was insufficiently risk-sensitive, includes an inappropriate charge on unit-linked business and does not diversify with other risks. While the formula and diversification benefit have not been changed in QIS 4 (CEIOPS stated that this was better addressed under Pillar 2), the charge for unit-linked business has now been based on expenses, which aims to address the industry feedback.
4. Life underwriting risk.
The life correlation ratio has been changed between mortality and longevity following industry feedback. From previously assuming no correlation, it has now been set to -0.25. This is likely to have a significant effect on firms exposed to both risks.
5. Non-life underwriting risk.
Calibration of the non-life underwriting risk module of standard formula SCR was regarded as unsatisfactory in QIS 3. Some factors have been recalibrated for QIS 4 and there is now an option to replace these factors with entity-specific values, calculated on historic data using standardised techniques.
6. Life catastrophe risk.
The QIS 3 SCR included an assessment of lapse catastrophe risk, based on a mass lapse of 75% of the company unit-linked portfolio. This was criticised as being too onerous and has been reduced to 30% in QIS 4. The new rate is now extended to apply to all businesses where there is a surrender strain on lapse.
7. Non-life catastrophe risk.
A simple formulaic underpin (Layer 1) will apply in QIS 4 if regional scenarios (now referred to as Layer 2) are not available. QIS4 also introduces Layer 3, where insurers will calculate capital based on the ‘personalised’ catastrophe scenarios that are regarded as appropriate to their business.
8. MCR.
The approach set out for the MCR was believed to be too volatile and impractical and demonstrated an unintuitive relationship with the SCR for many life insurers. QIS 4 is now proposing a ‘linear’ approach, which is less risk-sensitive. However, as a mitigating factor, the final specification includes a link between the SCR and the MCR via a 50% cap and 20% floor. CEIOPS had made it clear that this is an area in which the QIS 4 methodology is by no means a final view on the potential methodology.
The fact that a number of issues were addressed by CEIOPS in QIS 4 highlights how the consultation process allows the industry to contribute to the overall form of future Solvency II requirements.
Issues for insurance groups
QIS 3 was the first consultation to address the issues related to management of insurance groups, and included limited guidance in this regard. QIS 4 has aimed to address this by including more guidance and focusing on important areas such as group diversification benefits.
A criticism from members of the industry was that the QIS 3 treatment of insurance groups did not allow for adequate geographical diversification benefit. The approach to non-EEA subsidiaries was also challenged. To address this, QIS 4 has requested a series of calculations so that different approaches can be investigated for group diversification. One of these is an approach that assumes consolidation of non-EEA entities, which aims to apply more consistent standards and include diversification benefits.
New issues that companies need to address in QIS 4
The main objectives of QIS 4 are to test:
>> Simplifications for smaller insurers (the principle of proportionality)
>> Simplified approaches to estimating technical provisions
>> Further refinements to the calculations and calibrations of the SCR standard formula.
A further use of these techniques is in the projection of the SCR in future time periods as proposed by the risk margin calculation for those who cannot project the SCR. This can be a viable alternative to the formula-based SCR run-off.
Next steps
Solvency II seeks to improve risk management and reward good practice. While updating risk management processes and practices takes time, companies are already working hard to implement the changes before the new regulations come into force — and possibly using QIS 4 as a testing ground to see how well they comply. Further QISs in 2009 and 2010 are also probable, providing further opportunities for actuaries and risk managers to familiarise themselves with the key elements of the new solvency framework.
Phil Vermeulen is a senior manager and Jeev Muthulingam is an associate. They both work for Ernst & Young in Zurich.
