QIS3: CEIOPS Report and CRO Forum Study

Objectives, key findings and conclusions

The Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS) has scheduled a series of Quantitative Impact Studies (QIS) as part of the consultation process on behalf of the European Commission with a view to establishing a risk-oriented solvency system (Solvency II). QIS3, which took place from April to July 2007, was the third of these studies.

The next impact study, QIS4, is due to commence in April 2008 and is expected to be of significant relevance when Solvency II is implemented. At this stage in the process, it is interesting to see what insights QIS3 has already provided, as they will have a significant influence on QIS4.

next page »

CEIOPS Report

The CEIOPS report sets out the results of QIS3, conducted by CEIOPS to support the European Commission in the development of an insurance solvency regime. This impact study was mainly designed to test the calibration of risk factors and to provide information from a group-supervision perspective. The practicability of calculations was also a key issue. The participants were asked for the first time to report their internal model results (groups on a consolidated basis).

The report incorporates the feedback from over 1,000 insurance companies from 28 countries in the European Economic Area (cf. Figure 1). This is an increase of 100% on QIS2.

The market share of participating companies was almost identical for life/health, non-life and composite companies, and for most countries it was more than 60%.

	Life	Non-life	Reinsurers	Composite	Total
Large	79	63	6	39	187
Medium	135	194	10	79	418
Small	116	254	12	40	422
Total	330	511	28	158	1,027

Table 1: Number of respondents

Objectives

CEIOPS’ report analyses the results of QIS3 with reference to its following four objectives:

• —
  To obtain further information on the practicability and suitability of the calculations involved and the alternatives tested
• —
  To enquire quantitative information relating to the impact on balance sheets, and the amount of capital that might be needed if the QIS3 approach were to be adopted as the Solvency II standard
• —
  To collect information on the suitability of the suggested calibrations for calculating the "Solvency Capital Requirement" SCR and the "Minimum Capital Requirement" MCR
• —
  To test the effect on insurance groups of applying the QIS3 specification

« previous page  |  next page »

Interpretation of selected key findings

Financial impact

Compared to Solvency I, in 30% of companies eligible own funds would increase by more than 50%, whereas in 34% eligible own funds would decrease by more than 50%.¹ Furthermore, 15.7% of companies would need additional capital to meet their SCR (cf. Table 2).

	Life	Non-life	Composite	Total
Large	18.3	23.7	7.3	17.5
Medium	12.4	20.0	7.1	15.3
Small	10.9	18.0	13.2	15.4
Total	13.1	19.5	8.7	15^.7

Table 2: Percentage of firms with additional capital needs to meet SCR

• —
  Interestingly, large life companies appear to have more difficulty meeting the SCR than smaller ones (some supervisors particularly expect smaller monoline non-life undertakings to be more likely to raise additional capital).
• —
  Compared to Solvency I, both life and non-life companies show a decrease in their solvency ratios based on the QIS3 calculations.
• —
  MCR was met by most participants under the approaches tested. However, especially for life companies, the MCR approaches tested provided outcomes that seemed to be inconsistent with the SCR calibration.

Composition of the SCR

One of the most important insights provided by QIS3 is the composition of the SCR. Based on this analysis, companies are able to investigate the risk drivers in their portfolios. Though calibration is not yet finalised, interesting aspects can already be observed by analysing the composition of the SCR.

Figure 2: Composition of basic SCR (top graph: life; bottom graph: non-life); each bar displays the average outcomes of participants in all countries, totaling 100%. Diversification effects are explicitly displayed.

Figure 2 shows that market risk is generally the most significant risk for life insurers, whereas underwriting risk dominates the overall risk exposure for non-life undertakings. There can also be significant differences from one jurisdiction to another. The CEIOPS study also provides a further drill-down into sub-risk categories.

Comparison with internal models

For the first time in the QIS exercises, internal model results were requested in QIS3. 13% of all participants provided figures for this analysis.

For the life insurance companies, the internal models produced on average a total SCR about 15% lower than with the standard approach, but the ratio of internal model SCR to standard formula SCR varies considerably between companies. The internal models of non-life insurers produce SCRs on average 25% lower than with the standard formula, largely due to differences relating to non-life underwriting risk.

For life and non-life insurers the internal models produce on average somewhat lower capital requirements for the market risk component.
The capital requirement for the credit risk component in internal models is far higher (although it has only a subordinate effect on the SCR).
The internal models especially of non-life insurers also produce much higher capital requirements for the operational risk component. Company feedback indicates that it is difficult to depict the effect of reinsurance appropriately in standard factor-based approaches.

---

¹ It should be noted that Solvency I and QIS3 use different definitions of eligible own funds. Internal models may use a third definition of eligible capital.

« previous page  |  next page »

CRO Forum QIS3 Benchmarking Study

Objectives & key findings

The CRO Forum comprises major European insurance groups, the member companies of which have both developed internal capital models and participated in QIS3. The CRO Forum QIS3 Benchmarking Study was undertaken by Tillinghast in association with Prof. Dr. Damir Filipovic and Dr. Daniel Rost in order to obtain a QIS3 calibration benchmark through a thorough comparison of the results of internal models, both for solo entities and for groups.

Objectives defined by the CRO Forum

• —
  To compare the capital requirements produced by CRO Forum members’ internal models with those calculated according to the standard approach defined for QIS3
• —
  To obtain an insight into the reasons for the differences between the standard approach under QIS3 and internal models
• —
  To identify any inherent conservatism/optimism in the QIS3 calibration, or potential problems in methodologies proposed by CEIOPS

Interpretation of selected key findings

The capital requirements for the CRO Forum companies are generally higher on average under QIS3 than under Solvency I. Furthermore, the use of internal models leads to a higher coverage ratio (defined as eligible own funds as a percentage of required capital) than with the QIS3 standard approach.

Figure 2: Coverage ratio – eligible own funds as % of required capital

Non-life solo companies

• —
  The coverage ratio of eligible own funds* to SCR is 175% according to QIS3 and 195% according to internal models (cf. Figure 2).
• —
  The ratio of internal model capital requirement to QIS3 SCR is 80% which approves a more conservative calibration in QIS3 than in internal models.
• —
  Market risk capital requirements are roughly the same in QIS3 and internal models.
• —
  Non-life underwriting risk is the largest risk class and the main driver for differences in capital requirement (capital requirements for non-life underwriting risk are 29% lower using internal models than under QIS3; a 25% lower credit for diversification across risk types under QIS3 than under the internal model also contributes to the difference).

* Internal models and QIS3 use different definitions of eligible own funds → a stand-alone comparison of the capital requirement itself between internal models and QIS3 can be to some extent misleading.

Life solo companies

• —
  The coverage ratio of eligible own funds to SCR is 186% according to QIS3 and 220% according to internal models.
• —
  The ratio of the internal model capital requirement to QIS3 SCR is 87% and approves therewith a more conservative calibration in QIS3 than in internal models. The largest risk class, market risk, produces similar capital requirements under QIS3 and in internal models (although the higher requirements for equities/property in the internal models compared to the QIS3 standard approach offset lower requirements for interest-rate and spread risks).
• —
  The relative composition of life underwriting risk also differs considerably in the QIS3 standard approach from the internal models for some jurisdictions (cf. Figure 3).

The key reasons for the difference in SCR are a 29% higher capital requirement for life underwriting risk and a 10% lower credit for diversification across risk types under QIS3 than under the internal model.

Figure 3: Life u/w risk by risk type as % of total requirements for all risk types before diversification

Groups

• —
  The coverage ratio of eligible own funds to SCR for groups is 142% according to QIS3 and 211% according to the internal models.
• —
  The ratio of the internal model capital requirement to QIS3 SCR is 71% which approves a more conservative calibration in QIS3 than in internal models (at group level this effect is more pronounced than at solo entity level).

A difference lies in the treatment of Non-EEA countries. Whilst internal models assess Non-EEA countries on a full economic basis, the QIS3 standard approach uses the local statutory basis. This can have a considerable effect at group level depending on the significance of the operations in the Non-EEA countries.

« previous page  |  next page »

Summary and outlook

Many studies have been performed on the basis of the QIS3 results, including the CEIOPS report and CRO Forum Benchmarking Study, which provide a deep insight into how Solvency II will impact the insurance industry.

The following are the main findings of these reports:

• —
  Financial impact: There are considerable differences between QIS3 and the existing Solvency I regime. Whilst some companies will have higher coverage ratios, others will have additional capital needs.
• —
  Internal models: A comparison of internal models to QIS3 reveals interesting aspects of the calibration of various factors in the standard QIS3 approach. While market risk is assessed in a similar way, the underwriting risk requirements differ significantly both for non-life and life underwriting risks. In each case, QIS3 results in higher requirements than those produced by internal models.
• —
  Insurance groups: The treatment of Non-EEA countries in QIS3 is very different from that in internal models. A comparison of QIS3 and internal models at a group level is therefore very difficult.
• —
  Risk mitigation: Factor-based approaches do not currently capture risk mitigation, such as reinsurance, appropriately.

Field study – QIS4

The next major step on the way to Solvency II will be the fourth quantitative impact study, QIS4, which will be conducted by the European Commission from April to June 2008.

Preparations are in full swing. Unlike all preceding studies, this one is not being conducted by CEIOPS but by the European Commission itself. The Commission requires QIS4 to test various aspects, for example the extent to which less complex methods of calculating capital requirements exist and can be implemented in practice.

The results of QIS4 at European level are expected to be published in November 2008. The envisaged timetable is as follows:

• —
  20 December 2007: CEIOPS submits the final QIS4 specifications draft to the European Commission.
• —
  End of January–mid-February 2008: Consultation phase regarding the QIS4 specifications draft. The Commission is planning to hold public hearings and meetings with stakeholders and with CEIOPS.
• —
  April 2008–July 2008: QIS4 study to be carried out.

Stakeholders have been invited by the European Commission to comment on the draft QIS4 specifications as a part of the formal consultation process running until 15 February 2008.²

As QIS4 will be of significant relevance for Solvency II, insurance companies can gain an early insight into how they will perform once Solvency II has been introduced in 2012.

---

² http://ec.europa.eu/internal_market/insurance/solvency/index_en.htm#qis4

« previous page