Legal provisions of SEC(2010)840 - Document de travail des services de la Commission - Analyse d'impact accompagnant le livre blanc sur les fonds de garantie d'assurance

Please note

This page contains a limited version of this dossier in the EU Monitor.

dossier SEC(2010)840 - Document de travail des services de la Commission - Analyse d'impact accompagnant le livre blanc sur les fonds de garantie ...
document SEC(2010)840 EN
date July 12, 2010

EN

EUROPEAN COMMISSION

Brussels, 12.7.2010

SEC(2010) 840


COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

PART I

Accompanying document to the

WHITE PAPER

on Insurance Guarantee Schemes

{COM(2010) 370}
{SEC(2010) 841}

 

TABLE OF CONTENTS

COMMISSION STAFF WORKING DOCUMENT IMPACT ASSESSMENT Accompanying document to the White Paper on Insurance Guarantee Schemes 1

INTRODUCTION 4

1. PROCEDURAL ISSUES AND CONSULTATION OF INTERESTED PARTIES 6

1.1. THE COMMISSION WORKING GROUP ON IGS 6

1.2. THE OXERA REPORT ON IGS 6

1.3. INVOLVEMENT OF CEIOPS 7

1.4. OTHER CONSULTATIONS 7

1.5. EUROPEAN PARLIAMENT AND COUNCIL 7

1.6. IMPACT ASSESSMENT BOARD AND INTER-SERVICE STEERING GROUP 8

2. PROBLEM DRIVERS 9

2.1. POLICYHOLDERS LACK AND CANNOT PROCESS IMPORTANT RISK-RELATED INFORMATION (PROBLEM DRIVERS 1 AND 2) 9

2.2. INSURANCE UNDERTAKINGS CAN FAIL AND PRODUCE SUBSTANTIAL LOSSES (PROBLEM DRIVER 3) 10

2.2.1. Reasons for insurance failures 10

2.2.2. The probability of default of insurance undertakings in the EU 11

2.2.3. Failure of an insurance undertaking can produce substantial losses which are passed on to policyholders or taxpayers 12

2.3. THE INSUFFICIENT (PROBLEM DRIVER 4) AND UNEVEN (PROBLEM DRIVER 5) PROTECTION OF POLICYHOLDERS 14

2.3.1. The fragmented landscape of IGS protection in the EU 14

2.3.2. Loopholes in the protection of policyholders as a result of the (non) existence of IGS 15

2.3.3. Loopholes in the protection of policyholders as a result of heterogeneous design features of existing IGS 15

2.4. CROSS-BORDER INSURANCE ACTIVITY IN THE INTERNAL MARKET (PROBLEM DRIVER 6) 16

2.4.1. The non-negligible (and growing) cross-border insurance activity in the EU 16

2.4.2. Insufficient protection of cross-border insurance activity by existing IGS 17

3. PROBLEMS AND CONSEQUENCES 18

3.1. SUBSTANTIAL LOSSES PASSED ON TO LARGE GROUPS OF POLICYHOLDERS OR TAXPAYERS (PROBLEMS I AND III) 18

3.1.1. Negative consequences of losses passed on to policyholders in a domestic context (no ex-post State intervention) 19

3.1.2. Negative consequences of losses passed on to taxpayers in a domestic context (ex-post State intervention) 20

3.1.3. Negative consequences of losses passed on to policyholders in a cross-border context (no ex-post State intervention) 20

3.1.4. Negative consequences of losses passed on to taxpayers in a cross-border context (ex-post State intervention) 21

3.2. MISMATCH BETWEEN THE RISK PREFERENCE OF CONSUMERS AND THE RISK OF DEFAULT OF AN INSURER (PROBLEMS II AND IV) 22

3.2.1. Sub-optimal allocation of losses from insurance failure in a domestic or cross-border context (on policyholders or taxpayers) 22

3.2.2. Distorted competition in the Internal Market for insurance services 22

3.3. SUB-OPTIMAL DEVELOPMENT OF THE INTERNAL MARKET FOR INSURANCE SERVICES 23

4. BASELINE SCENARIO, POSSIBLE ALTERNATIVES, SUBSIDIARITY AND LEGAL BASIS FOR ACTION, INTERNATIONAL COMPARISONS 24

4.1. HOW WOULD THE SITUATION EVOLVE WITHOUT ACTION AT EU LEVEL? 24

4.2. ARE THERE VIABLE ALTERNATIVES TO SPECIFIC EU ACTION ON IGS? 24

4.3. SUBSIDIARITY ANALYSIS AND LEGAL BASIS 26

4.4. IGS OUTSIDE EUROPE 27

5. OBJECTIVES OF AN EU ACTION 27

5.1. OBJECTIVE 1: ENSURE AN EVEN AND COMPREHENSIVE PROTECTION OF POLICYHOLDERS 27

5.2. OBJECTIVE 2: AVOID COMPETIVE DISTORTIONS 28

5.3. OBJECTIVE 3: REDUCE ADVERSE INCENTIVES 28

5.4. OBJECTIVE 4: ENSURE COST EFFICIENCY 28

Minimise welfare costs of protection 28

Minimise set-up and operational costs of protection 29

5.5. OBJECTIVE 5: ENSURE MARKET CONFIDENCE AND STABILITY 29

6. ANALYSIS OF AVAILABLE POLICY OPTIONS 29

6.1. THE NATURE OF A POSSIBLE EU ACTION 30

6.2. THE GUARANTEE SIZE (IGS FUNDING NEEDS) 35

6.3. TOOLS FOR AN EU ACTION ON IGS 39

6.4. MINIMUM VS MAXIMUM HARMONISATION 41

7. EXPECTED ECONOMIC AND SOCIAL IMPACT OF RETAINED POLICY OPTIONS 44

7.1. IMPACT ON POLICYHOLDERS 46

7.2. IMPACT ON INSURANCE UNDERTAKINGS 47

7.3. IMPACT ON TAXPAYERS 47

7.4. IMPACT ON EXISTING IGS SCHEMES 48

7.5. IMPACT ON SMALL OR MEDIUM ENTERPRISES (SME'S) 49

7.6. IMPACT ON SUPERVISORY AUTHORITIES 49

7.7. INTERNATIONAL IMPACT 49

7.8. ENVIRONMENTAL IMPACT 50

7.9. IMPACT ON FINANCIAL STABILITY 50

7.10. IMPACT ON THE ECONOMY 50

7.11. IMPACT ON SOCIAL WELFARE 50

7.12. ADMINISTRATIVE BURDEN 51

8. FOLLOW-UP MEASURES – MONITORING AND EVALUATION 51

ANNEXES 52

INTRODUCTION


The recent financial turmoil has made people far more conscious of the existence and limits of consumer protection/guarantee schemes in all financial sectors. In the insurance sector many EU Member Statesi have no consumerii protection arrangements in place, or have implemented guarantee schemes that only cover specific types of insurance. In order to remedy the existing regulatory loopholes and inconsistencies, the report of the de Larosière Group has recommended the setting-up of harmonised Insurance Guarantee Schemes (IGS) throughout the EU.iii


In light of this, the Commission announced in its Communication of 4 March 2009 'Driving European recovery' that it would review the adequacy of existing guarantee schemes in the insurance sector and make appropriate legislative proposals. To this end the Commission will adopt in 2010 a White Paper setting out a European approach to IGS including indications on appropriate follow‑up measures.


IGS provide protection to consumers when insurers are unable to fulfil their contractual commitments. They thus protect people from the risk that their claims will not be met if their insurance undertaking becomes insolvent. IGS provide protection either by paying compensation to policyholders for their claims, or by securing the continuation of their insurance contract. This can be done either by facilitating the transfer of the policies to a solvent insurer or by directly taking charge of the policies.


The main objectives linked to the establishment of IGS in a national/domestic context are to avoid significant reductions in the wealth of large groups of policyholders, to protect consumers' confidence in the insurance sector and financial markets, to prevent possible slowdowns of the real economy, to avoid a suboptimal allocation of insurance failure losses and to preserve the stability of financial markets. In the broader EU context, IGS also serve the purpose of protecting consumer confidence in the Internal Market, of avoiding potential disputes between Member States on the allocation of the losses stemming from defaulted insurers and of avoiding competitive distortions between EU insurance undertakings.


Guarantee schemes have been set up in other sectors of the financial services industry. All EU Member States have deposit guarantee and investor compensation arrangements and minimum protection standards were harmonised at European level by the 1994 Deposit Guarantee Scheme (DGS) Directive and the 1997 Investor Compensation Scheme (ICS) Directive.iv However, there is no such common European framework in the insurance sector.v


This Impact Assessment (IA) does not deal with the issue of consumer guarantees related to the activity of occupational pension funds, because relevant EU legislation on occupational pension funds is currently under revision in a parallel workstream eventually leading to an amended draft proposal in the mid-term. The scope of this IA does not extend to reinsurance undertakings either because consumers are, in general, not directly affected by the failure of a reinsurance undertaking.vi


This IA is structured as follows: Section 1 presents the main procedural issues, including the consultation of interested parties. Section 2 focuses on the six main problem drivers: (i) the fact that policyholders lack important risk-related information; (ii) the fact that policyholders cannot process important risk-related information; (iii) the fact that insurers can fail and produce substantial losses; (iv) the fact that protection of consumers in some Member States is low or insufficient; (v) the fact that protection of consumers in several Member States is uneven; (vi) the fact that cross-border activity in the EU is growing.


Section 3 presents the two main problems generated by these six drivers: the fact that substantial losses can be passed on from insurance undertakings to large groups of consumers or to taxpayers, and the fact that there is the possibility of a mismatch between consumer risk preferences and the risk of default of insurance undertakings. These two problems are analysed both in a domestic and in a cross-border context. Section 3 also explains the main consequences of each of these problems.


Section 4 highlights what would happen if the EU took no action and examines the case for EU action in the light of the subsidiarity principle and the existence of a legal basis for such action in the EU Treaty. Section 5 introduces the objectives of EU action on IGS. Section 6 analyses the main options available in terms of the nature, the tool and the content of possible EU action. Section 7 analyses the expected economic and social impacts of the retained set of policy options.

1.PROCEDURAL ISSUES AND CONSULTATION OF INTERESTED PARTIES


The Commission's attention was drawn to IGS as long ago as 2001 by Ireland after the collapse of a major UK insurance undertaking - Independent Insurance - which also operated cross-border. To date, the failure of Independent Insurance, which initially affected 190,000 policyholders, has generated some 738 million EUR losses.


1.1.THE COMMISSION WORKING GROUP ON IGS


In 2001, the Insurance Committee, the predecessor of the European Insurance and Occupational Pensions Committee, set up a working group which was mandated to examine IGS related issues. The working group quickly recognised that the subject was probably even more complex in the insurance field than in the banking and securities markets areas where EU Directives already require all Member States to have a national guarantee scheme in place. At the final meeting of the working group at the end of 2005 most Member States indicated that they were in favour of some European coordination in this area, although there was no consensus on the extent and content of such coordination.vii


1.2.THE OXERA REPORT ON IGS


In order to have a comprehensive picture of the situation in EU Member States and a better insight into the functioning of existing schemes, the Commission contracted Oxera Consulting Ltd to prepare a report - Oxera (2007) - on IGS in the EU both for life and non-life insurance (excluding motor insurance). The report was finalised at the end of November 2007 and was published on the Commission's website in January 2008viii.


1.3.INVOLVEMENT OF CEIOPS


In its letter of 5 May 2009, the Commission asked CEIOPSix to give its view on the feasibility of the various design features of a possible European approach to IGS and to update the Oxera report's description of existing IGS. Moreover, CEIOPS was asked to give its view on whether, if the EU were to introduce a European regime for IGS, this regime should be extended to the pensions sector. On 30 June 2009, CEIOPS submitted its reportx to the Commission.


1.4.Other consultations


On the basis of the Oxera report, the Commission carried out a public consultation exercise in 2008xi. It received 30 contributions, from European and national associations, insurers, supervisors, Ministries, a consumer panel, an IGS and CEIOPS. A public hearing was also held on 2 June 2008. The results of the consultation and hearing were put together in a summary feedback statement which was published on the Commission's website.xii


In May and June 2009, the Commission met with representatives of CEA, FINUSE, AMICE, CEIOPS, EFRP and EFDI to discuss the content of the forthcoming White Paper. The minutes of the meetings are published on the Commission's websitexiii.


1.5.European Parliament and Council


In one of its recommendations arising out of the Equitable Life Committee of Inquiry (No 25)xiv, the European Parliament called on the Commission to go ahead swiftly with preparing legislation on IGS.


In addition, Article 242 of the Solvency II Directivexv entering into force in 2012 requires the Commission to take into account developments and progress on a harmonised and adequately funded EU-wide solution for IGS and to report on this to the European Parliament and to the Council by 2014.


1.6.Impact assessment board and inter-service steering group


An inter-services steering group was set up to monitor progress and to feed in views. The group comprised representatives from SG, the LS, the JRC as well as JLS, COMP, SANCO, ECFIN, EMPL, ENTR and TAXUD. The minutes of the last steering group meeting have been sent to the IA Board.


The IA Board held its meeting on 10 March 2010 and issued its opinion on 12 March 2010 asking for some modifications to the IA. The main recommendations included:


- a more precise indication of the extent of the problem, explaining why EU intervention is needed and how this initiative relates to other policies in the field;
- a clearer presentation of the objectives and of all relevant policy options, with an analysis of subsidiarity and proportionality aspects;
- a more comprehensive overview of the expected impacts of the options, including alternatives that do not include an IGS;
- a more explicit indication of the planned next steps in the development of policy on IGS.


These proposed amendments were taken on board and a revised draft IA was resubmitted to the IA Board on 12 May 2010. The Board issued its opinion after a written procedure on 28 May 2010. It recognised that the report had been improved on a number of issues mentioned in the Board's first opinion and welcomed the fact that stakeholders will be able to provide feedback on the White Paper and that any follow-up measures will be accompanied by a further IA. In addition, the Board requested:


- to explain more clearly the likelihood of a default of insurance undertakings, the need for enhanced consumer protection and the need for relevant action at EU level;
- to simplify the presentation of the objectives and the most relevant policy options;
- to improve the understanding of the report by removing repetitions and unnecessary details.


These recommendations led to a revision and redrafting of the text in order to make the text easier to read. The section on the likelihood of default including the analysis of potential losses to be incurred by policyholders or taxpayers has been further developed (see in particular 2.2). The report explains more clearly the importance of enhanced consumer protection in a domestic and in a cross-border setting (see in particular 3.1 and 3.2). It shows that although alternative measures to EU action exist, they do not sufficiently address the shortfalls identified (see particular 4.2). The sections on objectives and relevant policy options have been streamlined and the relevant analysis has been more focuses more clearly on the relevant key elements and questions at stake (see in particular section 5 and 6).

2.PROBLEM DRIVERS


The integrated, competitive and stable functioning of the Internal Market for insurance services is affected by four problems created by six problem drivers. Figure 1 shows the problem drivers (1 to 6) and the problems (I to IV) they lead to, as well as their consequences. In this section, the problem drivers will be discussed one by one. The resulting problems and consequences will then be presented in Section 3.


Figure 1 - Problem tree


2.1.Policyholders lack and cannot process important risk-related information (Problem drivers 1 and 2)


For a number of reasons, it is almost impossible for consumers to assess the quality/security of insurance services:


First, there is a significant information gap on the side of policyholders, which prevents them from choosing between insurance services on the basis of their level of security. In fact, while policyholders can compare insurance undertakings' products on the basis of the premiums they would pay for any specific product, they hardly have any reliable information on the risk of failure of individual insurance undertakings. Moreover, policyholders are usually unaware of the existence (or not) of an IGS protecting them (and up to what coverage level) when they take out an insurance policy.


Second, even if policyholders know about the risk of failure of individual insurance undertakings operating in the market, they can hardly put a price on such a risk.xvi


Due to the policyholders' lack and – in general – inability to correctly process important risk-related informationxvii, they are more exposed to the risk of choosing insurers which are not financially sound. This may lead to a systematic mismatch between policyholders' risk aversionxviii (supposedly high, as they are looking for insurance) and the risk of an insurer's default which they continue to run.


2.2. INSURANCE UNDERTAKINGS CAN FAIL AND PRODUCE SUBSTANTIAL LOSSES (PROBLEM DRIVER 3)


2.2.1.Reasons for insurance failures


Failure of an insurance undertaking may have different origins. These may or may not be linked to financial markets.xix


Non-life insurance undertakings are less concerned by financial market developments. Their losses tend to arise from non-financial liabilities. In fact, losses by non-life insurers are typically caused by higher than expected claims (due, for example, to natural catastrophes, etc.) rather than by investment losses (see Error: Reference source not found for the high variability over time of losses from natural catastrophes and man-made disasters).


Life insurers are much more exposed to financial market developments. Their losses are mainly generated by financial liabilities. Life insurers are certainly exposed to insurance losses from non-financial events as well, such as unexpected rates of mortality due - for example – to pandemics or increased longevity. But market/investment risk is typically the main source of risk for life insurers: indeed, on most policies, life insurers offer an investment performance guarantee to policyholders. When financial markets fall, life insurers are normally hit by losses arising from their financial liabilities.xx


In general terms, life insurance undertakings are more exposed to losses:


- When interest rates fall (thus reducing returns on assets or the discount rate applied to liabilities);
- During periods of high market volatility (as an increased volatility increases the value of guarantees to policyholders)
- When there are falls in equity or bond markets, driven by increased spreads.


In the recent financial crisis, for instance, losses to life insurers have mainly been caused by a fall in equity values (see Error: Reference source not found) and by the widening of spreads on corporate bonds (see Error: Reference source not found).xxi Some important European insurers have reported particularly severe losses and have been forced to inject large amounts of new capital (for example, Allianz, losses of 7.3 billion USD and a capital injection of 2.0 billion USD; Aegon NV, losses of 7.9 billion USD and a capital injection of 4.1 billion USD; AXA, losses of 1.8 billion USD and a capital injection of 2.0 billion USD - see Error: Reference source not found).


Apart from operational causes, losses for insurance undertakings might also be generated by fraud and, more generally, by the severe agency problems that insurance undertakings are potentially subject to. These agency problems are mainly caused by the length and the 'inversion' feature of the insurance cycle, i.e. the fact that premiums are cashed in at an early stage and that claims are paid off only at a much later stage.


2.2.2. The probability of default of insurance undertakings in the EU


Default occurs when an insurance undertaking is unable to meet its financial obligations. Because of prudential requirements established by EU law, failures of insurance undertakings have not been very frequent in the past. Over the period 1996 to 2001, around 85 insurers have failed. And between 2001 and 2004, at least another 48 insurers (31 non life, 14 life and 3 composites) have defaulted in the EUxxii.


The Oxera report (see in particular sub-section 4.1.3) provides a calculation of the Probability of Default (PD) by a major European insurer rated by Standard & Poor's and arrives at the average value of 0.065%. This value corresponds with that arrived at by the Commission during the preparation of this Impact Assessment using data from Standard & Poor's on European insurers, updated to the year 2008 (see Error: Reference source not found).


Using Moody's Kealhofer, McQuown and Vasicek (KMV) model, the ECB monitors equity values and their volatilities of insurers and thus calculates the expected probability of default for the Euro area insurance sector. The median expected probability of default calculated by the ECB has increased significantly - to some 0.5% - during the financial crisis (see Figure 1). This figure is coherent with the default rate in 2008 of major European insurers rated by Standard & Poor's, which rose to 0.404% (see Error: Reference source not found).


Neither the current (Solvency I) nor the future (Solvency II) EU solvency regimes create or can create a zero-failure environment for insurance undertakings.xxiii On the basis of both historical data and model estimations, and for the purposes of this IA, one may assume that the PD of insurance undertakings ranges, according to economic conditions, by and large between 0.1% in normal conditions and 0.5% in exceptional conditions such as a financial crisis or the existence of particular conditions of weakness for insurers in a specific EU country.


On the basis of this range of probability of default of insurers, and taking into consideration the number of insurance undertakings present in each Member State, one can estimate the expected number of years between defaults of insurance undertakings in each Member State under normal market conditions to be: in DE 2 years; in UK 2.3 years; in FR 2.6 years; in LU 2.8 years; in IE 2.9 years; in NL 3.3 years; in ES 3.4 years; in IT 4 years (see Table 22). Against this background, it can be concluded that failure of one or more insurance undertakings may be expected to happen in the EU on average once every year.


2.2.3.Failure of an insurance undertaking can produce substantial losses which are passed on to policyholders or taxpayers


Error: Reference source not found presents the Exposure at Default (EAD) of the insurance sector in each Member State and in the EU. The EAD is an estimation, based on technical provisions, of the maximum losses for society that would occur in each Member State and in the EU in case of failure of the entire insurance sector.xxiv These hypothetical maximum losses would either hit policyholders or taxpayers, depending on the existence of IGS or on the possible intervention of public authorities. Error: Reference source not found presents EAD/GDP ratios: in the EU EAD of the insurance sector represents 52.69% of GDP.


It is important to bear in mind that losses incurred by policyholders might be different in nature depending on the contract and on how the failure is resolved. Failure of a life insurer may cause the loss of expected policy benefits, which can be significant particularly if the policy was purchased to provide for retirement income. Losses on savings and investment products may equally result in important wealth losses, when guarantees given cannot be honoured. With regard to non-life insurance failures, losses to policyholders may result from the loss of the policy benefit (e.g. protection) as well as from the loss of premiums already paid in advance.


With a probability of default ranging between 0.1% and 0.5%, it is clear follows that not all insurers will default, and not all at the same time. What lessons can we learn from this with regard to future developments?


One recent indicator is the 2009 failure of five insurance undertakings of the Greek Aspis Pronia insurance group, which held about 16% of the Greek life market. This failure has affected an estimated number of 200,000 life insurance and 600,000 non-life insurance policyholders. The estimated loss for consumers and taxpayers is estimated to be higher than 200 million EUR.xxv Apart from this recent default, Error: Reference source not found provides other examples of losses generated by selected defaults of European insurers.xxvi It clearly shows that losses derived from past failures provide only a very general and rough indication of losses that might hit consumers in other Member States in the future.


Another indicator is the identification (see Error: Reference source not found) of the average loss produced by a failed insurer in each country or the loss happening in each country when its largest insurer defaults.xxvii However, losses in each Member State can easily be higher or much higher than the average loss.xxviii Likewise, it is also a matter of fact that losses in each Member State will, in general, be lower than those produced by the failure of the largest insurer. It follows, therefore, that both the average loss and the loss produced by the default of the largest insurer represent only a very rough indication of future losses possibly hitting policyholders and beneficiaries in Member States.


Another way chosen by the Commission to estimate the losses that might hit policyholders in the future is to use a reasoned theoretical model. The model in question allows to estimate policyholders' losses combining the effect of various elements, such as: the EAD, the PD, the correlation of defaults between insurers (how probable is it that defaults happen at the same time), the concentration of the insurance market (how many insurers dominate the market), and the severity (Loss Given Default) of the losses in the case of default.


The Methodological report (MR) explains in detail how the Commission, by means of a Vasicek model, has estimated the losses that might hit consumers in each Member State in a one year time horizon.xxix The order of magnitude of the estimated loss distributions has been tested on the basis of selected past failures in the EU. Past failures tend to fall in a range between the 75% and the 99% percentile of the estimated loss distributions.xxx


This means in conditions of serious stress, and in the total absence of IGS in Member States, that losses resulting from failures of insurance undertakings happening in a one year time horizon, that might (with a 99th confidence level) be passed on to policyholders or taxpayers, may amount to:


- 51.5 billion EUR for total (life and non-life) insurance in the whole EU, which is some 4.9% of total EU annual gross written premiums;
- 45.8 billion EUR for life insurance only, which is some 6% of annual gross written life premiums;
- 6.6 billion EUR for non-life insurance only, which is some 2.3% of annual gross written non-life premiums.xxxi


In conclusion, when EU insurance undertakings fail, EU policyholders or taxpayers can incur very significant losses.xxxii


2.3.The insufficient (problem driver 4) and uneven (problem driver 5) protection of policyholders


2.3.1.The fragmented landscape of IGS protection in the EU


The question whether an IGS needs to be introduced depends on the risk of failure of insurance undertakings and the potential impact that such failures could have on consumers. Given that clear evidence suggests that the latter can be considerable, the question arises as to the ability of the current (fragmented) framework of IGS to mitigate the risk or insurance failure or to reduce the losses for policyholders and beneficiaries if the risk materialises.


Unlike the banking and securities sectors, the insurance sector is not covered by any European legislation on guarantee schemes. Of the 30 EEA countries, 12 operate one (or in some cases more than one) general IGS as shown in Error: Reference source not found. In particular, six countries cover both life and non-life (excluding motor) insurance (ES, FR, LV, MT, RO and UK); three countries cover life insurance only (BG, DE and PL); and another three countries cover non-life insurance only (DK, IE and NO).xxxiii


History, including the recent financial crisis, has shown that public authorities generally tend to be more reactive than proactive towards handling risks of negative shocks hitting the financial sector. This is illustrated by the fact that many IGS were introduced following a major default of one or more insurance undertakings or have been triggered by insurers experiencing serious financial difficulties in a given Member State. Where no IGS exists, this is normally due to the absence to date of major defaults.


2.3.2.Loopholes in the protection of policyholders as a result of the (non) existence of IGS


Given the limited number of existing IGS, a large number of policyholders in the EEA have no IGS protection whatsoever against the risk of failure of an insurance undertaking (both life and non-life (excluding motor insurance)).xxxiv According to the Commission's estimate (see Error: Reference source not found) the share of the EEA market - in terms of gross written premiums - which is not covered by any IGS is 35% for the whole insurance sector, 26% for life and 56% for non-life.


Error: Reference source not found shows the estimated funds available in existing national IGS. This table read together with Table 31 shows that there might be situations whereby even already existing IGS are not able to fully absorb total losses. For example, notwithstanding the existence of an IGS, major defaults in DE, FR and UK may amount to: 1.3, 1.9 and 2.0 billion EUR respectively in normal times and 6.9, 10.1, 14.1 billion EUR respectively in situations of crisis.


This means that, in a situation of market stress and taking into account existing IGS, losses that might (with a 99th confidence level) be passed on to policyholders or taxpayers for failure of an insurance undertaking happening in a one year time horizon may amount to:


- 46.5 billion EUR for life and non-life (total) insurance together in the whole EU;
- 41.3 billion EUR for life insurance only;
- 5.9 billion EUR for non-life insurance only.xxxv


In conclusion and taking into consideration funds available in existing IGS, significant losses stemming from the failure of insurance undertakings can be passed onto EU policyholders or taxpayers.xxxvi


2.3.3.Loopholes in the protection of policyholders as a result of heterogeneous design features of existing IGS


Loopholes in the protection of policyholders can also stem from differences with regard to the design features of existing IGS. There are significant differences between national IGS not only in terms of whether a scheme exists at all and whether it has a general or a specific coverage, but also in relation to other aspects such as: geographical scope (home country principle, host country principlexxxvii, etc), eligibility restrictions, protection limits, nature of intervention, funding arrangements, financial capacity, etc.


Error: Reference source not found provides a detailed analysis of the design features of existing IGS.xxxviii It is very difficult to analyse in detail the consequences (in terms of loopholes in the protection of policyholders in Member States) of all these differences because of the complexity of the elements involved.xxxix It has therefore not been possible to carry this out in this IA. Attention has instead been focused on two of the main design features for IGS: policies covered and geographical scope.xl


2.4.Cross-border insurance activity in the internal market (problem driver 6)


Sub-section 2.4.1 describes the size and the features of cross-border insurance activity in the EU. Sub-section 2.4.2 quantifies the losses that could hit policyholders from cross-border insurance activity in the EU.


2.4.1.The non-negligible (and growing) cross-border insurance activity in the EU


Although cross-border activityxli is still relatively limited in the major EU insurance markets, it has increased over time and it is likely to increase further in the future. Some major European insurers (for example AVIVA) have, for instance, recently announced their intention to turn their EU subsidiaries into branches.xlii This should help them to make better use of their capital particularly in light of the fact that Solvency II will not introduce the group support regime as initially proposed by the Commission.


In 2007, the volume of exported insurance services in the EU - in terms of gross written premiums – amounted to 42.8 billion EUR, of which 11.8 billion EUR have been sold via branches and the rest via Free Provision of Services (see Error: Reference source not found). Cross-border insurance activity is mostly related to life insurance, which in 2007 amounted to 33.2 billion EUR (see Error: Reference source not found), while non-life insurance activity covered in total 9.6 billion EUR (see Error: Reference source not found).


Cross-border activity represents 4.10% of total gross premiums written in the EU. The share of EU-wide exported activity varies however quite significantly between Member States. LU (98.89%), IE (57.24%), MT (43.32%), and EE (32.62%) are the Member States where exported activity is the most developed as a share of total activity. The share of EU-wide imported activity is instead relatively homogeneous between Member States. LT (13.45%), LV (12.60%) and CZ (11.97%), are the Member States where imported activity is the most important as a share of total activity (see Error: Reference source not found, Error: Reference source not found, Error: Reference source not found and Error: Reference source not found)


The highest volumes of exported insurance activity are to be found in IE and LU (see Error: Reference source not found and Error: Reference source not found). Total insurance gross written premiums exported by IE and LU amount in fact to 23.7 and 11.0 billion EUR respectively (81% of the total EU). If one focuses specifically on exported life insurance activity, it can be seen that this is very concentrated in the same two countries (see Error: Reference source not found and Error: Reference source not found): 20.5 and 10.4 billion EUR are the volumes of life insurance premiums exported from IE and LU respectively (93% of the total EU). Exported non-life insurance activity (see Error: Reference source not found and Error: Reference source not found) amounts instead to 9.6 billion EUR (22% of total EU). It is concentrated in a few countries: IE, FR, DE, BE, and DK.


The highest volumes of imported insurance activity are in UK, DE, IT and FR (see Error: Reference source not found and Error: Reference source not found). Total insurance gross written premiums imported amount to 15.5, 6.0, 6.4 and 6.1 billion EUR respectively (77% of the total EU). In terms of imported life insurance activity, the highest concentration is in the same four countries (see Error: Reference source not found and Error: Reference source not found). Imported non-life insurance activity (see Error: Reference source not found and Error: Reference source not found) is instead mainly concentrated in the UK and DE.


2.4.2.Insufficient protection of cross-border insurance activity by existing IGS


Error: Reference source not found shows in detail whether existing IGS cover domestic and cross-border life insurance activity. As set out in Error: Reference source not found, IE and LU do not have a home principle based IGS (see Endnote Error: Reference source not found) in place for life insurance and only four Member States (LV, MT, PL, UK) have a host principle based IGS. It follows that 62% (see Error: Reference source not found) of cross-border life-insurance activity is not covered by any IGS today.


Error: Reference source not found shows in detail to what extent existing IGS cover domestic and cross-border non-life insurance activity. While insurance sold out of IE and FR is covered by a home principle based IGS, insurance sold out of DE, BE, DK, LU and IT is not protected by a similar scheme (Error: Reference source not found). Overall, 23% (see Error: Reference source not found) of non-life cross-border activity in the EU is not covered by any IGS.


Error: Reference source not found and 42 show the losses that can be 'exported' to other Member States when providing cross-border insurance services. In a situation of market stress, losses that might (with a 99th confidence level) result from exported business and hit non-domestic policyholders or non-domestic taxpayers in a one year time horizon, may amount to:xliii


- 1.80 billion EUR for total insurance, which is around 3.5% of total (life and non-life) annual gross written premiums paid in the EU (or 1.77 billion taking into account existing IGS);
- 1.40 billion EUR for life insurance, which is around 3.1% of life annual gross written premiums paid in the EU (or 1.37 billion EUR taking into account existing IGS);
- 0.25 billion EUR for non-life insurance, which is around 3.8% of non-life annual gross written premiums paid in the EU (or 0.24 billion EUR taking into account existing IGSxliv).


Error: Reference source not found demonstrates for each Member State the losses that might hit domestic policyholders or taxpayers from (imported) cross-border insurance which is not covered by existing (home and host) IGS available funds. In a situation of market stress and taking into account the coverage of existing home and host state principle based IGS in the EU, losses that might (with a 99th confidence level) result from imported business and hit domestic policyholders or domestic taxpayers in a cross-border context in a one year time horizon, may amount to:


- 1.05 billion EUR for total insurance;
- 0.82 billion EUR for life insurance;
- 0.14 billion EUR for non-life insurance.xlv


It follows from the above that significant losses stemming from defaults of insurance undertakings operating in a cross-border setting might be exported to non-domestic policyholders. Similarly domestic policyholders might suffer important losses if they have purchased policies from a defaulting insurance undertaking in another Member State, when these losses are not covered by IGS in the home and/or the host Member State.

3.PROBLEMS AND CONSEQUENCES


3.1.Substantial losses passed on to large groups of policyholders or taxpayers (problems I and III)


As shown in Error: Reference source not found, any insurance failure may affect up to several hundreds of thousands of policyholders. Quantifying and estimating the exact number has not been possible in this IA, as statistics on the number of the policyholders of individual insurance undertakings in Member States are currently not available to the Commission.


Alternatively, when an insurer fails, the State may intervene ex-post, and absorb the losses caused by a failing insurance undertaking. In this case, the totality of taxpayers is hit by the losses produced by the failure. Even if the effect on individual taxpayers might be limited, overall, the effect on public finances could be significant.


3.1.1.Negative consequences of losses passed on to policyholders in a domestic context (no ex-post State intervention)


Losses passed on to policyholders can substantially reduce their wealth and income, particularly for households.


Life insurance policies are generally important components of households’ savings. Protecting life insurance policies, therefore, means securing people's life savings and thus protection them and their families from financial hardship. It can be estimated that policyholders' equity in the life EU insurance system amounts to some 5,696 billion EUR (see Error: Reference source not found). By contrast, losses that policyholders may incur under present conditions may amount to 41.3 billion EUR.


In the non-life insurance sector, the reduction in policyholders’ wealth may also be important but it presents different features. These losses generally affect only those policyholders with outstanding (or already incurred but as yet unreported) claims against the failed insurer, i.e. only a percentage of all policyholders.xlvi In concrete figures, the aggregate value for non-life claims can be estimated at around 821 billion EUR (see Error: Reference source not found) while the estimated losses to be incurred by policyholders in the non-life sector are around 5.9 billion EUR.


Furthermore, the failure of insurance undertakings together with the absence of policyholder protection mechanisms, are likely to decrease consumer confidence in the insurance industry. This may result in a weaker insurance sector, and may eventually generate financial contagion between insurers. An OECD study (Yasui T. (2001)) reports for example that "the insurance industry is built on public's confidence in the business, which is in fact vulnerable", so that 'without the ability to appropriately assess the risks of individual companies, the general public may lose their confidence in the soundness of other insurers', and 'the bankruptcy case of a given insurer may cast doubts as to the soundness of other insurers and induce a run on them. Such a run was actually observed in some countries, particularly on companies of poor reputation', similarly to the banking sector, as 'the line of reasoning is in fact analogous to the argument of the banking sector'.xlvii


In addition, unprotected insurance failures may lead to a slowdown of the real economy for two reasons. First, the reduction in policyholders' wealth can severely affect their consumption behaviour. Second, when insurance companies fail, the economy's overall ability to manage risk is reduced.xlviii There is evidence that the collapse of insurance undertakings can significantly harm the development of the economy over the following months or years.xlix The likelihood of such disruptions is clearly greater where the insurance market is concentrated or the collapse affects many undertakings at the same time.


Finally, losses of insurance undertakings passed on to policyholders can also cause or deepen financial market turbulence and instability as policyholders can react to losses with sudden mistrust for the whole insurance sector leading them to surrender their policies en masse. To put it differently, when a large number of policyholders decide to surrender their policies at the same time, this may lead to an exacerbated downward spiral in stock market prices as insurers may have to sell large quantities of assets in order to obtain the necessary liquidity.l


To summarize, when substantial losses are passed onto large groups of domestic policyholders, and if there are no consumer protection mechanisms in place (and the State does not intervene ex-post) to absorb these losses, this may trigger a series of important negative consequences.


3.1.2.Negative consequences of losses passed on to taxpayers in a domestic context (ex-post State intervention)


Losses passed on to domestic taxpayers may lead to a deterioration of domestic public finances which may obviously be more important if the losses are substantial loss and the state of public finances is weak. Error: Reference source not found shows the estimated losses of policyholders as a percentage of GDP: these may amount up to 0.42% of EU GDP. During the recent financial crisis several Member States have, for example, intervened after approval by the Commission to support insurance undertakings through state aid: absorbing impaired assets (e.g. Dexia: 3,1 billion EUR and ING: 0.75 billion EUR) or recapitalising them (e.g. Aegon: 3.0 billion EUR; Ethias: 1.5 billion EUR; ING: 4.75 billion EUR; KBC: 1.5 billion EUR).


Moreover, losses from insurance failures that are absorbed by public finances also eventually lead, due to public budget constraints, to a reduction in public spending (for example on public services offered to citizens) for an amount equal to the loss. This may, depending on the amounts involved, have significant long-run effects on the real economy.


3.1.3.Negative consequences of losses passed on to policyholders in a cross-border context (no ex-post State intervention)


Losses passed onto non-domestic policyholders can cause substantial reductions in their wealth. Error: Reference source not found shows that in 2007 the embedded EU value for non-domestic policyholders in the insurance sector as a share of GDP was 1.70%. This includes an estimated value of 178.7 billion EUR covered by life-insurance policies (Error: Reference source not found) and 31.1 billion EUR attributed to the non-life sector. With regard to losses that non-domestic consumers might incur, taking into account existing IGS, these may amount up to 0.82 billion EUR (life insurance) and 0.14 billion EUR (non-life insurance) respectively.


Furthermore, the failure of insurance undertakings without any policyholder protection mechanisms in place in the home and/or host Member State may decrease consumer confidence in the Internal Market.


Finally, disputes at political level may arise between Member States regarding the allocation and, where appropriate, compensation across countries of the losses generated by the failure of insurance undertakings operating cross border.


3.1.4.Negative consequences of losses passed on to taxpayers in a cross-border context (ex-post State intervention)


Losses passed on to non-domestic taxpayers may have a negative impact on the public finances of another Member-State which may obviously be more important when the loss is substantial and the state of public finances is weak.


Furthermore, the losses from insurance failures that are absorbed by the public finances of another Member State also eventually lead, due to public budget constraints, to a reduction in public spending (for example on public services offered to citizens) for an amount equal to the loss. This may, depending on the amounts involved, have significant long term effects on the real economy.


In addition, when non-domestic taxpayers are asked to absorb the losses of a defaulted insurance undertaking operating cross-border, this may adversely affect public opinion in the Member State(s) concerned as well as upset consumer confidence in the Internal Market. Finally, disputes at political level may arise between Member States regarding the allocation, and where appropriate, compensation across countries of the losses generated by the failure of insurance undertakings operating cross-border.


3.2.mismatch between the risk preference of consumers and the risk of default of an insurer (problems ii and iv)


3.2.1.Sub-optimal allocation of losses from insurance failure in a domestic or cross-border context (on policyholders or taxpayers)


As set out above, individual policyholders are hardly able to process important risk-related information which leads him/her to underestimate the risk of his/her insurer going bankrupt. This creates a mismatch with the risk preference of policyholders which may cause a suboptimal allocation of the losses caused by the failure, because consumers are convinced that they are insured and that they will consequently not be affected by the loss resulting from a default.


The resulting welfare problem stems from the absence of an efficient insurance market. An inefficient insurance market does not allow society to maximise social welfare as some parts of society remain exposed to an excessive amount of risk compared with their individual risk preference as well as with their efforts to be insured/protected against negative shocks. In other words, a welfare loss occurs because of an insufficient redistribution of negative shocks in society.li This argument holds true unless the public authority – a Member State – intervenes ex-post to absorb losses. This, in turn, creates another problem as the cost from absorbing the losses will in that case be incurred by the totality of taxpayers which may produce further efficiency losses.lii


3.2.2.Distorted competition in the Internal Market for insurance services


The coexistence of different systems of IGS (including their total absence in some Member States) may create uneven levels of protection for policyholders purchasing insurance services in a Member State. This argument is particularly important in the context of an internal market that enables and encourages consumers to buy insurance cross-border.


To illustrate this point, a different protection of policyholders takes place in an (importing) Member State in the following situations (see also Error: Reference source not found):


• an IGS is in place in the exporting Member State based on the home state principle, while there is no IGS in the importing Member State (cross-border activity is more protected than domestic activity);


• an IGS is in place in the exporting Member State based on the host state principle, and in the importing Member State based on the home state principle (cross-border activity less protected than the domestic activity);


• an IGS is in place in the exporting Member State based on the home plus host state principle, while there is no IGS in the importing Member State (cross-border activity more protected than the domestic activity);


• no IGS is established in the exporting Member State, while there is an IGS in the importing Member State based on the home state principle (cross-border less protected than domestic activity).


Error: Reference source not found and 48 show the cases of an uneven level of protection in (importing) Member States with regard to cross-border life and non-life insurance activity.


Differences in IGS treatment between domestic and foreign EU insurers may result in an unlevel playing field and may cause distortions in competition between these two groups of insurers. These competitive distortions are closely related to the general inability of consumers to correctly process complex risk-related information. Consumers may, for instance, prefer to buy policies that are covered by an IGS to the detriment of insurers offering policies that are not covered. On the other hand, belonging to an IGS entails additional costs which will ultimately be borne by policyholders. Alternatively, consumers might prefer to buy lower priced insurance services because they are incapable of appreciating correctly the importance of IGS protection. This would again distort competition to the detriment of insurers which offer protected insurance products.liii


3.3.SUB-OPTIMAL DEVELOPMENT OF THE INTERNAL MARKET FOR INSURANCE SERVICES


Ideally, in a perfectly functioning internal market, cross-border and domestic activity should receive the same protection under IGS. Error: Reference source not found clearly shows that such is not the case. A large part of cross-border activity (54%) in the EU remains unprotected as compared with domestic activity (34%). In other words, existing national IGS are designed in such a way that domestic insurance activity is better protected than cross-border activity.


For life insurance, 62% of cross-border activity is not covered (as compared to 25% of domestic life activity), while only 23% of cross-border non-life activity lacks relevant protection (as compared to 57% of domestic non-life activity).


There is, in conclusion, an important discrepancy in the coverage provided by existing IGS with regard to domestic and cross-border insurance activity. This argument holds particularly true for the life insurance sector.

4.BASELINE SCENARIO, POSSIBLE ALTERNATIVES, SUBSIDIARITY AND LEGAL BASIS FOR ACTION, INTERNATIONAL COMPARISONS


It follows from the above that the coexistence of different national approaches to IGS raises concerns about comprehensive and even consumer protection in the EU. It may also lead to competitive distortions and may hinder the development of a single market in insurance. The question now arises whether these problems can be best addressed by Member States andr whether there are adequate alternatives to specific EU action on IGS.


4.1.How would the situation evolve without action at eu level?


Despite the introduction of a more risk-based solvency regime, Solvency II will not create a zero-failure environment. A certain residual default risk will continue to exist. In the case of failure, the loss will be passed on to policyholders.


Although existing IGS regimes lead to an uneven and insufficient protection of policyholders within and across Member States, there are no signs that Member States are taking or planning initiatives in order to remedy the situation.


On the other hand, the scale of cross-border insurance activity in the EU is expected to increase. This is not only due to growing market integration in Europe, but also to the recently introduced Solvency II requirements. The lack of recognition of group support has already prompted some international insurance groups to turn some or all of their EU subsidiaries into branches. An increasing number of branches might alter the existing IGS landscape in the EU, thereby possibly exacerbating the existing shortfalls.


4.2.ARE THERE VIABLE ALTERNATIVES TO SPECIFIC EU ACTION ON IGS?

The importance of introducing an IGS depends on the risk of failure of insurance undertakings and the potential impact that such failures could have on consumers. This raises the question as to what alternative protection mechanisms are available at national or at European level to mitigate the risk of insurance failure or to reduce the losses for policyholders if the risk materialises.

Prudential regulation and risk management: The Solvency II Framework Directive1 which will become applicable by 31 December 2012 provides for a risk-based, economic approach to solvency. It requires insurance and reinsurance undertakings to hold sufficient capital to cover their obligations over a 1-year time horizon subject to a 99.5% VaR confidence level. This should ensure that failure of an insurer occurs no more often than once in every 200 cases. Effective risk management and comprehensive governance structures are cornerstones of the future solvency system, in addition to capital requirements and appropriate supervisory powers of varying degrees of intensity. In spite of the many safeguards contained in the new solvency regime, Solvency II will not amount to a zero-failure regime. It is widely acknowledged that it would be too costly to set solvency requirements at a level that would be sufficient to absorb all unexpected losses.

Preferential treatment of policyholders in winding-up proceedings: in the event of the winding up of an insurance undertaking, current EU winding-up legislation offers Member States a choice between two alternatives in national law for giving priority treatment to policyholders over other creditors of the insurer in liquidation2. However, reliance on winding-up proceedings may not be workable in practice. Firstly, there may not be a sufficient amount of assets for the protection of policyholders giving rise to uncertainty over whether policyholders will be compensated. Secondly, winding-up proceedings of insurance undertakings are not only complex but also expensive and time-consuming. This may create serious liquidity shortages for policyholders with outstanding claims at the time of insolvency, if their claims cannot be satisfied within a reasonable period of time.

Case-by-case government intervention: case-by-case solutions such as ex-post government interventions, while by their nature flexible, also have serious drawbacks. Unequal interventions may raise concerns regarding fairness and transparency, as relevant decisions are made on an ad-hoc basis rather than according to a set of pre-designed rules. In addition, case-by-case intervention may be perceived as privileging larger undertakings thereby incentivising risk and creating moral hazard through the assurance of safety nets for which others have to pay. Ad-hoc interventions may create uncertainty both for policyholders and, depending on their financing, for taxpayers and the industry.

Additional information and enhanced transparency: Approaches which enhance transparency and information requirements seek to strengthen policyholders' capacity to choose the most appropriate insurance product for themselves. These approaches rely on the assumption that relevant information is properly understood and incorporated in the decision-making process of policyholders. Particularly in Member States where the policies of domestic and incoming insurers are subject to different levels of IGS protection, enhanced information may in principle alleviate concerns about consumer protection within Member States. However, it is highly unlikely that policyholders are capable of understanding and processing all relevant information, particularly with regard to cross-border insurance business. Moreover, additional information does not address the issue of the differential consumer protection between different Member States and the fragmented IGS landscape within the EU as such, i.e. the lack of IGS in many Member States.


4.3.Subsidiarity analysis and legal basis


In its sentence of 4 December 1986 (Case 205/84), the European Court of Justice gave four reasons why policyholders need special protection:


1. insurance is a highly particular service because it is linked to future events, the occurrence of which is uncertain at the time a contract is concluded;
2. an insured person may find himself in a very precarious position if he does not obtain payment after filing a claim for compensation;
3. it is very difficult for a person seeking insurance to assess the terms of a contract and the outlook for the insurer’s future financial position;
4. insofar as insurance has become a mass phenomenon, it is just as essential to protect the interests of third parties.


Although action at Member State level could in principle contribute to address some aspects of the problems that have been identified, it would also leave some important aspects untouched.


In particular, Member States acting on their own would not be able to appropriately address the problems due to the coexistence of inconsistent features in the mechanisms set up to protect policyholders. It follows, that in accordance with the principles of subsidiarity and proportionality as set out in Article 5 TFEU, the objectives of the proposed action cannot be sufficiently achieved by Member States and can therefore be better achieved by the EU. Relevant proposals will not go beyond what is necessary to achieve the objectives pursued. Only EU action can ensure that all policyholders and beneficiaries acquiring insurance policies in the EU benefit from equal and comprehensive protection in the event that an insurance undertaking defaults, which also ensures a level playing field and thereby promotes further integration within the Internal Market.


The legal basis for EU action in the insurance field is to be found in the Treaty provisions related to free provision of services. According to Article 3 of the EU Treaty, the EU pursues the objective of an Internal Market characterised by the free movement of goods, persons, services, and capital. Article 26 of the Treaty of the Functioning of the European Union (TFEU) further states that the Internal Market shall constitute an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured in accordance with the provisions of the TFEU Treaty. Any follow-up action is likely to be based on Article 53 (2) of the TFEU which is the legal basis to adopt EU measures aimed at achieving the Internal Market in financial services.


4.4.IGS outside Europe


A number of countries outside the EU have already established IGS. In North America schemes for life and non-life insurance have been established in Canada and in the USA. Already in the 1970s the USA established distinct IGS, covering the life and non-life sector respectively, in each State. Before doing business in another State, insurance undertakings must be licensed in that State and must adhere to its IGS. Sub-section 4.6 of the Oxera report provides an overview of the main operational characteristics of the US IGS system.liv


In the Asia-Pacific region, Japan and Korea have established national schemes, covering both life and non-life insurance. Regarding Japan, the Non-life Insurance Policyholders Protection Corporation and the Life Insurance Policyholders Protection Corporation were established in 1998.lv The OECD reports the existence of IGS for life and non-life insurance also in Malaysia, Taiwan, Singapore and the Philippines.lvi

5.OBJECTIVES OF AN EU ACTION


Taking into account the domestic and the cross-border context, potential future EU action on IGS protection should pursue the following objectives:


MAIN OBJECTIVES


5.1.objective 1: Ensure an even and comprehensive protection of policyholders


EU action on IGS should ensure an adequately high and even protection of policyholders, sufficiently reducing the risk that the non-payment of claims by insurers will mean substantial losses passed on to policyholders or taxpayers. Relevant action should therefore ensure the following:


1. that protection mechanisms are present in all Member States to protect policyholders and that their resources are adequate in all Member States;
2. that the geographical scope of protection schemes does not maintain or produce loopholes in the protection of policyholders in any Member State;lvii
3. that other design features of protection schemes do not maintain or produce loopholes in policyholder protection in any Member State. This is particularly important as the geographical scope is strongly intertwined with other design features: when these are not sufficiently homogeneous, loopholes in the protection of policyholders in Member States can arise in spite of a harmonised geographical scope throughout the EU.


SUPPORTING OBJECTIVES

 


5.2.objective 2: avoid competive distortions

A harmonised framework on IGS protection at EU level should also aim at contributing towards a level playing field between insurance companies and improving competitive neutrality of business conducted by domestic undertakings and incoming EU insurers who operate under the freedom to provide services or who provide insurance via branches.


5.3.objective 3: reduce adverse incentives


EU action should ensure that the design features of the protection mechanisms minimise the risk of moral hazard for policyholders, insurers and supervisors/public authorities. It has been argued that the existence of a safety net in the form of an IGS, may lead consumers to be less inclined to assess the financial situation of the insurer that they contract with and to make a prudent selection. However, given the difficulty for consumers to correctly assess risk related information it can be argued that the introduction of a protection mechanism will not provide for the wrong incentives. Similarly a harmonised framework on IGS should prevent tax-payers from ultimately bearing the costs of an undertaking's mismanagement by introducing a legal framework which is financed by the undertakings themselves and that does not incentivise excessive risk-taking.


When there is a safety net to protect the interests of policyholders, supervisors might feel less pressured to carry out their supervision. The design of the protection mechanism should therefore also ensure that potential moral hazard problems in relation with supervision are minimised.


5.4. OBJECTIVE 4: ENSURE COST EFFICIENCY

EU action on IGS must strike the right balance between the benefits to policyholders and the costs linked to the protection offered. This means that both welfare costs of protection as set-up costs need to be minimised. In the end, an IGS that is not cost efficient will lead to higher costs for policyholders.

Minimise welfare costs of protection


From a societal point of view, the effects associated with the introduction of protection mechanisms in the case of insurance failure are to a large extent distributional.lviii If a protection mechanism exists, the losses in the case of failure are shifted from the policyholders concerned to a larger population. In other words, the protection funds will absorb an amount of losses that is equal to the losses that would hit consumers (or taxpayers) in the absence of a protection mechanism.lix


Under this distributional angle, the argument that the introduction of a protection mechanism will lead to excessive costs for society must be seen in the right perspective: the costs - in terms of financial volume - of the default of an insurer remain by and large the same, but resources to absorb them are paid by different groups of individuals and thus might entail welfare costs of different degrees.lx Welfare costs are very difficult to estimate quantitatively, but they can be assessed qualitatively in the light of the allocation of losses caused by an insurance failure. Furthermore, welfare costs can be determined by induction when looking at the resources needed by the protection mechanism chosen: the higher the resources mobilised, the higher the possibility of welfare costs. Finally, welfare costs can also be qualitatively analysed in terms of possible adverse incentives produced by IGS.


Minimise set-up and operational costs of protection


Welfare losses represent the main costs for society that can stem from the creation of a protection mechanism. However, when a protection mechanism is set up, it generally also entails set-up and operational financial costs. These costs must also be taking into account in assessing whether the benefits of protecting policyholders' claims outweigh the possible costs linked to the protection offered. The objective of minimising set-up and operational costs also includes the objective of respecting the existing supervisory structure as much as possible as well as to adequately taking into account European rules on state aid. As competitive distortions between insurance undertakings operating in the same Member State may also be generated when protection is linked to implicit or explicit Member State support, state-supported funding mechanisms that can create competitive distortions should be avoided.


5.5.objective 5: ensure market confidence and stability

EU action on IGS should finally aim at enhancing market confidence and furthering the stability of the EU internal market in insurance services. By increasing consumer confidence in insurance undertakings and products, an IGS may contribute to promote consumer demand and finally enhance the stability of the financial system.

6.ANALYSIS OF AVAILABLE POLICY OPTIONS


In order to enhance readability, Figure 2 shows all available policy options that are analysed in this IA. An extended list of policy options is set out in Annex A.


Figure 2 - Option tree


As the option tree shows, the first (1 to 4) group of alternative options requires the selection of a preferred option before identifying the next group of alternative options. When certain options are not mutually exclusive, this is indicated in the analysis. The second group of options (5 to 11) represent different design features of an IGS scheme. They do not follow from one another, but they are closely intertwined and therefore need to be put in a common context. The policy options discussed below will be analysed in terms of their compliance with the objectives identified above. The following score system has been used for the assessment of the options: from slightly positive (+) to strongly positive (+ + +); from slightly negative (–) to strongly negative (– – –); no or negligible impact: void.

6.1.The nature of a possible EU action


- Option 1.1: No action


Preserving the status quo implies a continuation of the coexistence of very different national approaches to policyholders’ protection. These differences generate uneven and inappropriate levels of policyholders' protection in several Member States and may hinder the harmonious development and functioning of the EU Internal Market for insurance services.


- Option 1.2: Further reduce the probability of failure of insurance undertakings


Failures can be prevented by strengthening the risk management system in insurance undertakings or by enhancing prudential supervision, particularly through an increase of the solvency requirements.


1.2.1. Strengthening the risk management system


The risk management system encompasses the whole range of processes present in the operational activity of insurance undertakings that aim at ensuring that an insurance undertaking is able to correctly and professionally manage its risks. However, practice suggests that even a very solid risk management system cannot fully exclude the risk of default. Solvency II substantially strengthens the system of governance in general and the risk management system (and function) in particular. Introducing further legal constraints in this regard would most likely be opposed by stakeholders because of the additional costs and administrative burden that would follow from this.


1.2.2. Enhanced prudential supervision and higher solvency requirements


If designed in an appropriately risk-sensitive way, solvency requirements can mitigate a potentially excessive risk-taking behaviour by insurance undertakings, limit the probability that they fail and therefore protect (up to a certain extent) the economy and society from the negative consequences linked to their failure.


It is however impossible to set solvency requirements at a level which is high enough to absorb all losses. Capital requirements would in such a case be so high that insurance undertakings would no longer be able to offer their services at a price which is affordable for consumers. In a worst case scenario this would lead to the implosion of the whole insurance market. Statistical evidencelxi shows that solvency requirements are optimally designed when they are sufficient to absorb losses of insurance undertakings in all cases except for those exceptional circumstances that would require too much capital. Solvency II requires insurance undertakings to hold sufficient capital to cover their obligations over a 1-year time horizon subject to a 99.5% VaR confidence level. This ensures that failure of an insurer occurs no more often than once in every 200 cases.


- Option 1.3: Introduce a protection of policyholders after failure of an insurance undertaking has occurred


1.3.1. Improved transparency under a caveat emptor approachlxii


Enhancing the information available to policyholders (about the existence of IGS and the level of IGS protection provided) should enable them to make a more informed choice between insurers. It would allow them to make their choice not only on the basis of the price offered in the market, but also on the basis of the intrinsic quality of the offer (in terms of the risk of failure of the insurer).


However, the adoption of such a caveat emptor approach is likely to be ineffective because policyholders are unaware of important risk-related information and are incapable to correctly process important but complex risk-related information (problem drivers 1 and 2). If the EU were to increase transparency, it would still be very unlikely that the great majority of policyholders would fully understand and be capable to appropriately process the complex risk-related information they receive.


In conclusion, a transparency measure would not overcome the shortfalls of the existing status quo, with the consequence of maintaining uneven or inappropriate levels of policyholder protection in several Member States. The Oxera report argues that adopting a caveat emptor approach might not be acceptable, especially when substantial losses or a large number of claimants are involved.


1.3.2. Legal priority for consumers in winding-up


Notwithstanding common principles contained in the Winding-up Directive 2001/17/EC as recasted by the Solvency II Directive 2009/138/EC, winding-up procedures vary substantially between Member States. In addition, winding-up proceedings inherently include the risk of an insufficient amount of assets for the protection of policyholders, giving rise to uncertainty over whether policyholders will be compensated. Moreover, winding-up proceedings are in any case complex and very expensive processes taking a long time. They can, therefore, hardly provide an effective and immediate protection for policyholders.


- Option 1.4: Explicit unlimited guarantee from public authorities


An explicit unlimited guarantee from public authorities has the obvious drawback of using (potentially a very large amount of) taxpayers money and thereby affecting public finances. It would entail a sub-optimal allocation of insurance failure losses leading to a reduction in welfare, resulting from a too-large redistribution of failure losses. Moreover, a guarantee from public authorities can also create a moral hazard behaviour (insurers might be less inclined, for example, to limit their risks through reinsurance contracts).


- Option 1.5: Explicit limited guarantee (Insurance Guarantee Scheme)


In the event that insurers fail, a guarantee scheme/fund can absorb insurers' losses up to its financial endowments.


Statistical evidence suggests that high levels of security for consumers can be best achieved by combining (lower) capital requirements with a guarantee scheme rather than having (higher) solvency requirements without a guarantee scheme. A guarantee scheme somehow bundles the protection from losses in excess of insurers' capital (tail risk)lxiii thereby reducing their variability (and as a consequence the funding needed to absorb them) if the number of insurers participating in the scheme is sufficiently large. Furthermore, a guarantee scheme may spread the funding needed to provide protection against these excess losses among a higher number of consumers and therefore provide the same level of protection at a lower cost. These ideas are represented in Error: Reference source not found, which shows how the 'centralisation' of the tail risk of insurance undertakings facilitates a consistent distribution of excess losses (losses in excess of solvency requirements plus excess capital, if any) for the guarantee scheme.


Another important argument in favour of IGS is that they, if properly designed, may reduce the problems of a suboptimal allocation of insurance failure losses. Without IGS in place either the policyholders of the defaulted insurer or the totality of taxpayers absorb losses causing a reduction in social welfare in both cases. An IGS that covers losses, includes the entire community of policyholders to absorb them. This can minimize the allocation problem of insurance failure losses and therefore maximise social welfare. It is also argued that IGS may contribute to the development of competitive markets. In other words, IGS can be seen as a 'smooth exit mechanism for incompetent insurers from the market'.lxiv


As recalled in sub-section 4.4 of the Oxera report, IGS also have the advantage of: (1) being able to guarantee a speedy payment to policyholders; (2) minimise and possibly bring to zero the loss incurred by policyholders; (3) introduce an element of predictability and certainty on the effects of the failure of an insurance undertaking for its policyholders.


The most commonly raised argument against introducing IGS is the potential incentive to a moral hazard behaviour that IGS may create for policyholders, insurers and supervisors. As indicated above, there is sufficient evidence suggesting that this argument is not as strong as it might seem.


Similarly, the Oxera report notes that although there is very little evidence on how the introduction of IGS can influence the proper/inappropriate allocation of economic incentives, it may be concluded that the proper design of IGS features can in general address and appropriately manage potential moral hazard effects.


Regarding set up and operational costs, the Oxera report notes in section 5 that in the absence of failure, financial costs associated with running an IGS are minimal, and that when failures occur, operational costs are small/negligible compared with the actual resources needed to provide the guarantee. According to the Oxera report they can be estimated to be around 0.5% or less of the funding endowments.


In the 2008 public consultation, there were split views in relation to IGS related EU actions. Some respondents favoured them, while others were in favour of maintaining the status quo or improving transparency. In the consultation there were also split views regarding the possibility to mitigate the possible moral-hazard drawbacks of IGS through an appropriate design, so as to finally obtain IGS that improve competition and the functioning of the insurance market. Respondents in favour of IGS stressed how IGS would be effective in solving the problems/consequences identified in this IA: increased consumer protection, increased consumer confidence, increased financial stability, level playing field between insurers. They also stressed that EU action on IGS could encourage the development of the single market. In contrast, respondents against IGS stressed the difficulty to sufficiently minimise the constraints such as costs or adverse economic incentives, so that drawbacks would outweigh benefits.


Table 1 contains an evaluation of the arguments discussed above. The preferred policy option is therefore Option 1.5 (Explicit Limited Guarantee / IGS).

Table 1 - Summary of policy options' evaluations – The nature of a possible public authority intervention
Objectives

EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
1.1 No action
1.2.1 Strengthening internal risk management+----------
1.2.2 Higher Solvency requirements+----------
1.3.1 Improve transparency- -
1.3.2 Legal priority during winding -up+------- 
1.4 Explicit unlimited Guarantee++++--------------
1.5 Explicit limited Guarantee (IGS)++++--- --++


6.2.the guarantee size (IGS funding needs)


It is important to bear in mind that IGS are designed to cover the most extreme losses that occur with a very low probability. Error: Reference source not found illustrates the choice of IGS scheme size/funding in terms of coverage of risk of failure. The vertical red line in Error: Reference source not found shows the cut-off point up to which a chosen level of IGS funding will be able to protect policyholders from losses.


The target fund of an IGS,lxv is influenced by many parameters, among which two appear to be the most important: the probability of default (PD) of insurance undertakings and the level of targeted security for policyholders. As has been set out above, the PD for insurers oscillates between 0.1% and 0.5% depending on market conditions. Besides the probability of default of insurance undertakings, IGS funding needs/financial endowments are mostly influenced by the level of security provided to consumers: the higher the security provided by an IGS, the higher the required IGS financial endowments/funding needs. A key decision is therefore the level of security that an IGS is expected to provide to policyholders.


In practical terms, the level of security provided to policyholders is determined in relation to the part (statistically, the percentile)lxvi of the IGS loss distribution that the IGS financial endowments can cover.lxvii The percentile (level of security) chosen should not only provide a high level of security for consumers but also be financially realistic: i.e. it should have the potential to achieve the objective of a sufficiently high protection of policyholders, without requiring excessive resources.

In order to identify an appropriate level of protection offered by IGS, the coverage levels of existing national schemes have been analysed, and past cases of large insurance failures have been examined, estimating how many resources would be needed to protect policyholders against similar failures across the EU. It appears from this analysis that existing IGS protect consumers from losses up to a percentile that ranges between the 75th and the 99th.lxviii On the basis of available evidence, three funding levels are considered: 75%, 90%, and 99%.


The following list of policy options can be drawn up with regard to the level of IGS financial endowments, taking into consideration both the probability of default of insurers and the level of security for consumers:lxix


- Option 2.1: No action (harmonization) at EU level
- Option 2.2: Harmonization at EU level
- Sub-option 2.2.1: Low risk, low security (PD=0.1%, percentile=75%)
- Sub-option 2.2.2: Low risk, medium security (PD=0.1%, percentile=90%)
- Sub-option 2.2.3: Low risk, high security (PD=0.1%, percentile=99%)
- Sub-option 2.2.4: High risk, low security (PD=0.5%, percentile=75%)
- Sub-option 2.2.5: High risk, medium security (PD=0.5%, percentile=90%)
- Sub-option 2.4.6: High risk, high security (PD=0.5%, percentile=99%)


While option 2.1 is inconsistent with the objective of providing a high and even level of protection to policyholders in all Member States, the choice between the various sub-options in option 2.2 clearly depends on a cost-benefit analysis.


The benefit must be considered in terms of the security provided to consumers, which is expressed in percentiles, with higher percentiles meaning higher security. In addition, Error: Reference source not found and 50 provide a more tangible indication of the security offered to policyholders in terms of the biggest failure that IGS financial endowments under the various policy options can cope with in the life and non-life sector.


It appears very clearly that IGS are not able to deal alone with the biggest failureslxx, but their capacity to do so increases when financial endowments are higher. Options 2.2.1 and 2.2.4 would only allow to compensate losses that arise from the failure of small insurance firms, in the order of the 16th biggest insurance undertaking and above in the respective Member State. Option 2.2.2 protects policyholders against the losses of medium size insurance undertakings, covering losses in the range of the 11th – 15th biggest insurance undertaking's failure, but not above that. Options 2.2.3 and 2.2.5 are quite similar in their coverage and they are providing higher protection against medium-size insurer failures than the previously described options. They cover up to the 11th – 6th biggest insurance undertaking in most Member States. Finally, option 2.2.6 (high risk/high protection) is protecting policyholders against the failure of one among the biggest five insurance undertakings in many Member States, and against the failure of the 6th-10th biggest insurers in others. Only in three countries (UK, SWE, LU) it would not cover the losses caused by failure of the 10th biggest insurance undertaking.


An analysis of the funding needs of an IGS should also take into account the annual costs that a certain funding may impose on society, in case resources are anticipated but losses do not eventually materialise. Error: Reference source not found indicates the share of annual premiums that correspond to each funding need. For example, the funding needs for option 2.2.3 (PD=0.1%, percentile=99%), taking into consideration a cost of capital of 6% (in line with Solvency II quantitative impact studies,) amounts to an annual cost of 0.08% of annual premiums.


Table 2 presents a summary evaluation of the various policy options related to the size of the guarantee offered by IGS. On the basis of the preliminary conclusions drawn by the Commission so far, it shows a tentative preference for option 2.2.3 which would ensure a high level of protection under normal market conditions while equally ensuring a sufficiently high level of protection in times of stress.

Table 2 - Summary of policy options' evaluations – The guarantee limit (IGS funding needs)
Objectives

EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
2.1 No action
2.2.1 Harmonization low risk, low protection (PD=0.1%, alpha=75%)++----+
2.2.2 Harmonization low risk, medium protection (PD=0.1%, alpha=90%)++----+
2.2.3 Harmonization low risk, high protection (PD=0.1%, alpha=99%)++++------+++
2.2.4 Harmonization high risk, low protection (PD=0.5%, alpha=75%)++----+
2.2.5 Harmonization high risk, medium protection (PD=0.5%, alpha=90%)++++-------++
2.4.6 Harmonization high risk, high protection (PD=0.5%, alpha=99%)++++--------+


6.3.tools for an EU action on IGS


- Option 3.1: No action


Leaving relevant action to Member States would mean preserving existing loopholes as well as uneven levels of policyholder protection in several Member States, unless all Member States would decide to coordinate and adopt IGS along appropriately coherent principles. Even if action at national level were taken, Member States acting on their own would not be able to address the problem of inconsistencies in the geographical scope and other design features of IGS.


- Option 3.2: A coordinated EU non-binding approach to IGS to be followed by Member States on a voluntary basis


Current shortfalls could also be corrected by means of soft law instruments, such as recommendations, communications, guidelines and codes of conduct. By adopting these tools, the Commission might indicate the IGS design features that it considers most appropriate. However, these instruments do not have any legally binding force and Member States would be asked to remedy the highlighted shortfalls on a voluntary basis.


Even in the case of positive reactions by Member States, it is difficult to foresee how non coordinated responses by Member States could effectively address the problems determined by the absence of a comprehensive and consistent framework for IGS in the EU.


- Option 3.3: Case-by-case legally binding interventions (infringements)


Unsatisfactory situations of IGS protection could be addressed through selective policy interventions that, in practice, could take the form of formal infringement proceedings against Member States so as to determine appropriate changes in national IGS or their set-up.


Although selective measures might occasionally close existing gaps in policyholders' protection in some Member States, it can be assumed that they would fail to effectively address the problems linked to the absence of a comprehensive and consistent framework for IGS in the EU. The 2008 public consultation clearly showed no support for case-by-case interventions.


- Option 3.4: A legally binding EU-wide approach to IGS


The introduction of a legally binding EU-wide approach to IGS is most likely the best way to provide an adequate remedy to the existing loopholes and inequalities in policyholder protection. Moreover it seems to be the most adequate and proportional tool to guard against the need for taxpayer involvement. In case of binding legislative measures on IGS at EU level, two possible legal instruments are available:


EU Regulation


Regulations are normative acts defined in Article 288 of the TFEU. They have general application, are binding in their entirety and directly applicable in all Member States, thus leaving the national authorities hardly any flexibility with regard to their implementation. Given the existing fragmented landscape on IGS and the absence of any EU coordination/harmonization in this field to date, it can be assumed that a Regulation would excessively restrict Member State action to implement an EU framework for IGS.


EU Directive


Another legal instrument provided for by Article 288 of the TFEU is that of the Directive. It has individual application, meaning that it is binding upon those to whom it is addressed. It requires Member States to achieve a certain result but, unlike a regulation, leaves them free to choose their own forms and methods. In the view of the Commission there are strong arguments for choosing the legal form of a Directive in case of binding measures on IGS, given the complexity of the issue and the required degree of flexibility with regard to the national implementation of each design feature.


Table 3 presents a summary evaluation of the various policy options related to the choice of the tool for EU action on IGS. In conclusion, and in view of the above considerations, the Commission prefers a legally binding EU-wide approach to IGS based on a Directive.


Table 3 - Summary of policy options' evaluations – Tools for a EU intervention on IGS
Objectives

EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
3.1 No action
3.2 A coordinated EU non binding approach-
3.3 Case-by-case intervention (infringements)++---+
3.4.1 EU Regulation++++-------+
3.4.2 EU Directive++++----+++


6.4.Minimum vs maximum harmonisation


Even if binding measures were to be introduced in the EU, there are still many ways in which an IGS can be designed and an analysis of the various options available is therefore necessary.


- Option: 4.1: No action


A coherent IGS framework at EU level would not be achievable without harmonising the scope of the action. Therefore, in case of no action, the objectives would not be fulfilled.


- Option: 4.2: Minimum scope of harmonisation


Harmonisation of the following design features seems necessary in order to ensure a minimum level of coherence and effectiveness at EU level:


- Level of centralisation: Should an IGS be created at national or at European level?
- Role: Should an IGS operate as a last resort protection mechanism or should it have a wider role?
- Geographical scope: Should an IGS operate on the basis of the home or host country principle (or on a combination of the two)?
- Policies covered: Which classes of insurance (life, non-life, etc.) should the IGS cover?
- Eligible claimants: Which policyholders/claimants (natural persons, legal persons, SME, etc.) are to benefit from IGS?
- Timing of funding: Should the IGS be funded ex-ante or ex-post (or a combination of the two)?
- Nature of scheme intervention: Should the IGS simply compensate losses or should it also be designed to secure the continuity of policies (portfolio transfer)?


During the 2008 public consultation most respondents said to be in favour of minimum harmonisation, though some preferred maximum harmonisation. The vast majority of respondents were in favour of harmonising the geographical scope. A large number of respondents were also in favour of harmonising policies covered and eligible claimants. Fewer respondents supported harmonisation of the nature of the intervention and the timing of the funding. It was also stressed that a too limited approach might put into question the relevance of an EU action.


In its advice, CEIOPS (2009b) recommends adoption of a minimum harmonisation approach in order to fill the gaps in the current protection of policyholders in the EU. In the 2009 informal stakeholders meetings, the CEA indicated that a majority of CEA members would be in favour of minimum harmonisation whilst a minority would be in favour of maximum harmonisation. Should the minimum harmonisation approach be chosen, at least the following design features should be harmonised: geographical scope, policy covered and eligible claimants. AMICE, EFRP and EFDI agreed with the CEA. FINUSE was of the view that more design features should be harmonised.


- Option 4.3: Maximum scope of harmonisation


Maximum harmonisation at EU level means that national legislation may not exceed the terms of EU legislation. In other words, provisions adopted at EU level have to be considered as exhaustive, leaving Member States no further room for manoeuvre with regard to the adoption of supplementary procedures. A possible set of additional design features to be considered in this case might comprise the following:


- Pooling/separation of funding: Should the IGS funds be pooled (or kept separated) between classes of insurance activity (life , non-life, etc.)?
- Compensation limits and reductions in benefits: Should there be restrictions (and if so, which ones) in IGS payments per claimant or per policy?
- Exclusions from eligible claimants: Should specific situations be excluded from protection by an IGS?
- IGS contributions: How should insurance undertakings contribute to the IGS fund?
- Ownership, management and administration: How are IGS set up, managed and administered?
- Advertising/information requirements: How can insurance undertakings communicate IGS-related information to their customers?


Although maximum harmonisation is be better suited to enhance completion of the Internal Market, there seems to be a great deal of reluctance among Member States to adopt such an approach.


Table 4 presents a summary evaluation of the three policy options dealing with the level of harmonisation of the various IGS design features. In view of the existing differences relating to IGS protection in Member States, and given the wide consensus among stakeholders, the preferred option is that of minimum harmonisation.


Table 4 Summary of policy options' evaluations – Scope of a possible EU level IGS binding intervention
Objectives

EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
4.1No action (no harmonization of the scope)
4.2 Minimum Harmonisation++--++
4.3 Maximum Harmonisation++++------+


Conclusion: On the basis of the evidence provided in this IA, the White Paper will propose to introduce a Directive in order to ensure that IGS exist in all Member States and that they comply with a minimum set of design features as proposed under option 4.2. lxxi

7.EXPECTED ECONOMIC AND SOCIAL IMPACT OF RETAINED POLICY OPTIONS


As there is currently no legislation on IGS at EU level, information on the economic and social impact of an EU action in this field remains rather abstract and general. The impact mainly depends on the way in which the specific design features are implemented. For more reliable information it is therefore necessary to continue the analysis based on a set of pre-defined policy options.


At the current stage, the Commission is interested in collecting feedback from stakeholders on its White Paper. Therefore preferences expressed on specific IGS design features are meant as preliminary ones, which remain open to the feedback of stakeholders.


Annex B presents the analysis carried out so far by the Commission on the IGS specific design features mentioned in option 4.2, and illustrates the considerations which have been taken into account by the Commission when setting out preliminary preferences. Conducting an open dialogue with stakeholders on these preliminary preferences will allow the Commission to monitor and possibly update its evaluation of the various policy options when drafting follow-up measures, and to assess and possibly confirm whether the retained options on IGS design features satisfy the main objectives set out in this IA. A further impact assessment will therefore accompany follow-up measures, which will analyse in detail the possible combinations of IGS design features, and choose in a more definitive way the optimal vector of features for an IGS solution at EU level.lxxii


On the basis of the analysis contained in Annex B, the Commission's preliminary preferences with regard to the IGS design features mentioned under the minimum harmonization approach as set out under option 4.2 are the following:

- Level of centralisation: the Commission prefers introducing an IGS in all Member States because this is consistent with the existing national micro-prudential supervisory framework;
- Role: the Commission believes that the role of an IGS should be that of solely acting as a last resort protection mechanism in order to avoid as much as possible moral hazard problems in the behaviour of insurance undertakings and possible state aid issues;
- Geographical scope: in the Commission's view, the home state principle is the preferable policy option, especially because of its consistency with the existing supervisory framework;
- Policies covered: the Commission prefers to cover life policies and selected non-life policies as this strikes the right balance between ensuring a sufficiently large and solid protection of consumers on the one hand, and limiting costs on the other hand;
- Eligible claimants: the Commission believes that covering natural persons and selected legal persons (including SME) is the best way to strike the right balance between ensuring a sufficiently large and solid protection for consumers on the one hand, and cost efficiency on the other hand;
- Timing of funding: the Commission has a preference for of ex-ante funding which could be complemented by ex-post funding where necessary. This will ensure the immediate availability of funds while limiting costs to industry and consumers;
- Nature of scheme intervention: the Commission strongly encourages portfolio transfer where it is reasonably practicable to do so and justified in terms of costs and benefits. However, when all other means are exhausted, IGS should at least compensate losses of policyholders and beneficiaries.


The table below summarises the preferred set of tentative policy options as examined in this IA.


Table 5 - Summary of policy options' evaluations – Retained policy options
Objectives

EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
1.3.5 Explicit Limited Guarantee (IGS)++++--- --++
2.2.3 Harmonization low risk, high protection (PD=0.1%, alpha=99%)++++------+++
3.4.2 EU Directive++++----+++
4.2 Minimum Harmonization++--++
5.2 An IGS in all MS++++----++
6.2 Last resort protection+++-----++
7.2 Home state principle+++-----+++
8.5 Life and selected non-life policies++++++------+++
9.3 Natural and selected legal persons (incl. SME)++++------+++
10.4 Combination of ex-post and ex-ante funding++++---------++
11.2 Portfolio Transfer++++------+++


The expected economic and social impact of these retained policy options is presented in the following .


7.1.Impact on policyholders


Action taken at EU level can be expected to benefit policyholders by increasing their protection in the event that insurance undertakings are unable to fulfil their commitments. On the other hand, insurance undertakings are expected to pass a part of their contributions on to consumers which most likely will result in an increase of their premiums.


This point can be illustrated by a (theoretical) example: If, (compared to a situation where no IGS existed), IGS are established at the level of each Member State, based on the home state principle and covering life and non-life policies, (up to the 99th percentile and based on a scenario of PD=0.1%), this would require EU policyholders to contribute to the creation of funds amounting to a total of 13 billion EUR (see Table 6). This currently corresponds to 1.24% of annual gross written premiums. Applying this target level over, for instance, a 10-year time horizon would translate into an annual contribution of 0.124% of gross written premiums by each contributing undertaking/policyholder.lxxiii


These funds should be considered as additional premiums. Policyholders are paying to insure themselves against the possibility that their insurance undertaking defaults. The payments provided by policyholders can be considered to be roughly equivalent to the expected value of the losses they would avoid in case their insurance undertaking defaults. The financial costs for policyholders can be computed considering a cost of capital of 6% (in line with Solvency II quantitative impact studies). For an IGS with a funding endowment of 1.24% of annual premiums, this would translate into financial costs of 0.08% of annual premiums.


7.2.Impact on insurance undertakings


EU action on IGS will affect insurance undertakings in different ways, depending on whether they operate in Member States already having an IGS or not. In those cases where no IGS has been established so far, insurance companies might face – in all circumstances in the case of ex-ante funding and whenever insurance undertakings fail in case of ex-post funding – financial costs due to the introduction of the IGS if they are not able to pass their IGS contributions entirely onto consumers.


These funds, unlike the case of policyholders, constitute a financial cost (and not an anticipation of funds) for insurance undertakings, as losses hitting insurance undertakings in case of default only depend on capital (and not on premiums paid).


7.3.Impact on taxpayers


The introduction of IGS in all Member States can be expected to benefit taxpayers as there will be less need in the future to use their money in the case of default of an insurance undertaking. Based on the practical example set out above this would save taxpayers money up to 13 billion EUR (see Table 6) upon a timeframe, for example, of 5 or 10 years. It is important to bear in mind that EU action on IGS will affect taxpayers in Member States in different ways, depending on whether they are resident in a Member State already having an IGS or not.


7.4.Impact on existing IGS schemes


EU action will affect existing IGS to the extent that the framework established at EU level deviates from the national IGS framework already in place. Main impacts may, in particular, include:


IGS funding needs: The size of existing IGS funds may be affected. A preliminary rough calculation of the respective amounts (only if positive, i.e. an increase in funds), can be derived from the last column of Error: Reference source not found. DK, for example, would roughly need to raise its endowments by 217.20 million EUR. A further analysis for existing ex-ante funded IGS is also performed in Table 3.12 of the Methodological report where the impact on IGS funds is analysed in terms of the implied change in the level of security provided to policyholders.


Geographical scope: Currently, only MT (life and non-life) and NO (non-life) are operating an IGS on a host country principle basis, while LV (life and non-life), UK (life and non-life), PL (life) operate their IGS on a home plus host country principle basis. These schemes would need to modify their geographical scope and start operating on a home country principle basis.


Nature of scheme intervention: Currently only six IGS (DE life, FR life, UK life, FR non-life, NO non-life and ES) operate portfolio transfer, while the other existing IGS limit themselves to paying compensation to claimants.


Policies covered: Unless a precise scope of non-life policies considered is defined, it is not possible to draw definitive conclusions on the expected impact. However, it is likely that existing IGS in DK, IE, and NO would have to extend their scope of protection to life insurance products.


Eligible claimants: As the precise scope of selected legal persons has not yet been defined, no definitive conclusions on the expected impact can be drawn. However, it is likely that existing IGS in LV and PL for life insurance and in IE, LV and MT for non-life insurance have to extend the scope of their eligible claimants.


Timing of funding: As the extent of ex-ante funding has not yet been defined, it is not possible to draw definitive conclusions on the expected impact. However, it is likely that existing IGS in PL life, UK life and non-life, IE non-life and NO non-life would have to introduce an ex-ante funding element in addition to their current ex-post funding arrangements.


7.5.Impact on Small or Medium Enterprises (SME's)


Action taken at EU level can be expected to benefit SME's by increasing their protection in the event that insurers are unable to fulfil their commitments. On the other hand, introducing IGS protection for SME's throughout the EU will have an impact on SME's, as insurers will pass a part of their contributions on to SME's which will result in an increase of their premiums. These funds should be considered as additional premiums which they are paying to insure themselves against the possibility that their insurance undertaking defaults. The payments provided by each SME can be considered roughly equivalent to the expected value of the losses they would avoid in case their insurance undertaking defaulted. Moreover, the impact on SME will depend to whether they are already protected or not by an existing IGS in the various national frameworks.


7.6.Impact on supervisory authorities


Supervisory authorities in Member States that do not have yet an IGS in place, might need to be involved in their set-up and possibly also start managing them, while those authorities that already manage IGS, would have to ensure that their scheme is compliant with the proposed design features. A more precise analysis of the impact on supervisory authorities will be considered in the impact assessment accompanying the follow-up measures once the precise set of IGS design features will be definitively decided.


7.7.International impact


As a general rule, third country insurance undertakings that provide or want to provide insurance services in the EU must have their branches authorised in at least one EU Member State.lxxiv This means that after the authorisation these branches of third country insurance undertakings fall within the scope of this IA. The impact on third country insurers can therefore be expected to be the same as for EU-EEA insurers.


7.8.Environmental impact


Environmental impacts are expected to be marginal.


7.9.Impact on financial stability


The retained policy options are expected to bring benefits to financial stability, as they ensure that failures up to a relevant size do not produce threats to financial stability (since IGS are able to absorb them).


7.10.Impact on the economy


The retained policy options are expected to bring two main benefits to the economy. First, a level playing field will be created that avoids competitive distortions between domestic and non-domestic insurers. Second, the possibility of sub-optimal allocation of losses on policyholders and taxpayers will be reduced. This should, in turn, have positive effects improving the economy's growth path.


7.11.Impact on social welfare


The retained policy options may improve social welfare for three reasons:

- increased protection of policyholders;
- less use of taxpayers' money;
- insignificant welfare losses while redistributing insurance default losses.


Increased protection of policyholders is the most important impact on social welfare resulting from the introduction of a harmonised framework of IGS protection at EU level. Protecting policyholders – who in general are highly risk averse – from uncertainty and financial losses, is expected to increase social welfare substantially.


With regard to taxpayers, the options chosen have a significant positive impact in terms of saving taxpayers' money. As taxpayers can be assumed to be in part low risk averse, guarding against the need of taxpayers' involvement can be interpreted as a source of increased social welfare.


7.12.Administrative burden


Administrative burden cannot be assessed with precision at this stage, but will be subject to burden measurement under the standard cost model in the impact assessment accompanying any follow-up measure. In any case, the preferred options are not expected to lead to any significant administrative burden, especially because they are in line with the existing structure of supervision.

8.FOLLOW-UP MEASURES – MONITORING AND EVALUATION


The IA clearly provides evidence supporting the need for a legally binding EU solution on IGS protection based on minimum harmonization in order to ensure that IGS exist in all Member States and that they comply with a minimum set of design features.


The Commission, while drafting follow-up measures, will monitor and update its assessment of the various policy options linked to the proposed EU solution for IGS. In particular, the Commission will carefully evaluate the feedback received and take it into account when coming forward with a legislative proposal. The Commission will sum up the contributions received by the first half of 2011. An impact assessment will then be conducted and the Commission will put forward a legislative proposal on insurance guarantee schemes which will be presented to the Council and to the European Parliament.
ANNEXES








A. LIST OF POLICY OPTIONS


1 The nature of a possible EU action

- Option 1.1: No action

- Option 1.2: Further reduce the probability of failure of insurance undertakings

- Sub-option 1.2.1: Strengthening the risk management system

- Sub-option 1.2.2: Enhanced prudential supervision and higher solvency requirements

- Option 1.3: Introduce a protection of policyholders after failure of an insurance undertaking has occurred

- Sub-option 1.3.1: Improved transparency under a caveat emptor approach

- Sub-option 1.3.2: Legal priority for consumers in winding-up

- Option 1.4: Explicit unlimited guarantee from public authorities

- Option 1.5: Explicit limited guarantee (Insurance Guarantee Schemes)

2 The guarantee size (IGS funding needs)

- Option 2.1: No action (harmonization) at EU level

- Option 2.2: Harmonization at EU level

- Sub-option 2.2.1: Low risk, low security (PD=0.1%, percentile=75%)

- Sub-option 2.2.2: Low risk, medium security (PD=0.1%, percentile=90%)

- Sub-option 2.2.3: Low risk, high security (PD=0.1%, percentile =99%)

- Sub-option 2.2.4: High risk, low security (PD=0.5%, percentile=75%)

- Sub-option 2.2.5: High risk, medium security (PD=0.5%, percentile=90%)

- Sub-option 2.2.6: High risk, high security (PD=0.5%, percentile=99%)

3 Tools for an EU action on IGS

- Option 3.1: No action (Only spontaneous action at Member States level)

- Option 3.2: A coordinated EU non-binding approach to IGS to be followed by Member States on a voluntary basis

- Option 3.3: Case-by-case legally binding interventions (Infringements)

- Option 3.4: A legally binding EU-wide approach to IGS

- Sub-option 3.4.1: EU Regulation

- Sub-option 3.4.2: EU Directive

4 Minimum vs. maximum harmonisation

- Option: 4.1: No action

- Option: 4.2: Minimum scope of harmonisation

- Option: 4.3: Maximum scope of harmonisation

5 Level of IGS centralisation (Single EU-wide scheme vs. national schemes)

- Option 5.1: No action (harmonisation) at EU level

- Option 5.2: An IGS in all Member States

- Option 5.3: A single EU-wide IGS replacing (where relevant) national schemes

- Option 5.4: An EU-wide IGS that covers only policies written and sold cross-border via branches and/or free provision of services, plus national schemes covering domestic insurance activity;

- Option 5.5: An EU-wide IGS that covers only insurers who are part of a group supervision regime (including subsidiaries) plus national schemes for all other relevant policies (domestic and cross-border)

- Option 5.6: Complement existing IGS with a 28th regime

- Option 5.7: Introducing an IGS in all Member States complemented by a system of mutual support between national IGS.

6 Role of an insurance guarantee scheme

- Option 6.1: No action (harmonisation) at EU level

- Option 6.2: IGS as a last resort protection mechanism

- Option 6.3: Preventing failure and providing last resort protection

7 Geographical scope

- Option 7.1: No action (harmonisation) at EU level

- Option 7.2: Home country principle

- Option 7.3: Host country principle.

- Option 7.4: Home plus host country principle.

- Option 7.5: Home country principle with lead supervisor

8 Policies covered

- Option 8.1: No action (harmonisation) at EU level

- Option 8.2: Protection of life policies only

- Option 8.3: Protection of non-life policies only

- Option 8.4: Protection of both life and non-life policies

- Option 8.5: Protection of both life and selected non-life policies

9 Eligible claimants

- Option 9.1: No action (harmonisation) at EU level;

- Option 9.2: Natural and legal persons;

- Option 9.3: Natural and selected legal persons (including SME's);

- Option 9.4: Natural persons only;

10 Timing of funding

- Option 10.1: No action (harmonisation) at EU level;

- Option 10.2: Ex-post funding

- Option 10.3: Ex-ante funding

- Option 10.4: Combination of ex-post and ex-ante funding

11 Nature of scheme intervention

- Option 11.1: No action (harmonisation) at EU level

- Option 11.2: Portfolio transfer

- Option 11.3: Compensation of claims


A number of options will not be specifically dealt with in the White Paper but may become relevant for a legally binding EU solution on IGS at a later stage. Some or all of them will then be addressed in a separate Impact Assessment. These options include the following:


- Pooling (or not) of funding between classes of insurance activity;
- Compensation limits and other reductions in benefits (restrictions to IGS payments);
- Compensation limits and other reduction in benefits – per customer or per policy;
- Exclusions from eligible claimants;
- Allocation of contributions among insurers;
- Capping the level of contributions that can be raised in any time period;
- Ownership, management and administration; and
- Advertising/information requirements.

POSSIBLE CONTENT OF A LEGALLY BINDING EU SOLUTION FOR IGS FOLLOWING A MINIMUM HARMONISATION APPROACH: PRELIMINARY ANALYSIS OF THE OPTIONSlxxv


Level of IGS centralisation (Single EU-wide scheme vs. national schemes)


- Option 5.1: No action (harmonisation) at EU level


Maintaining the status quo implies a continuation of the existing fragmented landscape for IGS. This may hinder, for the reasons explained in this IA, the correct functioning of the Internal Market for insurance services, by creating conditions of uneven and insufficient policyholder protection in several Member States.


In the 2008 public consultation, there were split views with regard to keeping the status quo. Some respondents preferred this, while others were in favour of EU actionn.


- Option 5.2: An IGS in all Member States


The creation of an IGS in all Member States is consistent with the existing national micro-prudential supervisory framework.


Funding needs for the EU when opting for a home country principle national scheme in each Member State are those presented in Table 6.


Table 6 - Funding needs for the EU with national (home country principle) IGS (m €)


Source: Methodological report, Table 3.8. For a full analysis and figures for each Member State see sub-section 3.2 or the MR.


If IGS cover both life and non-life (i.e. total) insurance, the Member State (see Table 3.2, MR) with the highest funding need is the UK. The Member State with the lowest funding need is LV. As a percentage of written premiums, the Member State with the highest funding need is SE, while the Member State with the lowest funding need is BG.


If IGS only cover life insurance (see Table 3.4, MR), the Member State with the highest funding need is the UK. The Member State with the lowest funding need is LV. As a percentage of written premiums, the Member State with the highest funding need is SE, while the Member State with the lowest funding need is LV.


If IGS only cover non-life insurance (see Table 3.6, MR), the Member State with the highest funding need is DE. The Member State with the lowest funding need is LT. As a percentage of written premiums, the Member State with the highest funding need is DE, while the Member State with the lowest funding need is HU.


The 2008 public consultation showed that if an EU action were to be taken, there would be support for introducing an IGS in all Member States.


In the 2009 informal meetings with stakeholders, CEA, AMICE, FINUSE and EFDI showed support for this option.


- Option 5.3: A single EU-wide IGS replacing (where relevant) national schemes


The creation of a single EU-wide IGS that covers all relevant policies written and purchased within the EU would overcome the problems stemming from the existence of various different national legal frameworks. However, a single EU-wide IGS would not be consistent with the existing national micro-prudential supervisory framework.


Funding needs for the EU when opting for a single EU-wide IGS are presented in Table 7.


Table 7 – Funding needs for the EU with a single EU-wide scheme and comparison with funding needs under national IGS (m €)


L

Source: Methodological report, Table 4.29. For a full analysis and figures for each Member State see sub-sections 4.5 and A5.4 of the MR.


If a single EU-wide scheme were established, funding needs would change compared to the case of an IGS in each Member State in a way which is very dependent on the level of security chosen. If the level of security chosen is high, funding needs decrease with a global saving of funds. If instead the level of targeted security is low, funding needs increaselxxvi.


The impact in each Member State on IGS funding needs when opting for a single EU-wide IGS instead of national schemes based upon the home country principle for total insurance (life and non-life), as well as life and non-life insurance separately are shown in Tables 20, 22 and 24 of Annex 5 to the MR respectively.


The 2008 public consultation showed that there was no support for introducing a single EU-wide IGS.


- Option 5.4: An EU-wide IGS that covers only policies written and sold cross-border via branches and/or free provision of services, plus national schemes covering domestic insurance activity


The creation of an EU-wide IGS that covers only cross-border business, i.e. policies written and sold cross-border via branches and/or FPS, can address the specific problems that arise in the cross-border context whilst maintaining national flexibility when it comes to purely domestic business.


In practice, however, such a solution is likely to create a number of complications. First of all, an EU-wide IGS for cross-border business would not be consistent with the existing national micro-prudential supervisory framework. Furthermore, insurers with cross-border business would need to take part in both the cross-border scheme and their national scheme. Uneven protection levels between and within Member States would also continue, especially if domestic and cross-border business protection were different.


Overall, the funding needs for the EU under this option are the same as under option 5.2. The funding needs specific to domestic national IGS and to the single EU-wide IGS covering cross-border insurance activity are presented in Table 8.


Table 8 - Funding needs for the EU with domestic national schemes supplemented by an EU-wide IGS covering cross-border (branches and freedom to provide services) activity (m €)


Source: Methodological report, Table 4.19. For a full analysis and figures for each Member State see sub-sections 4.3 and A5.2 of the MR.


While the great majority of Member States would not contribute much of their funding endowments to the EU-wide cross-border IGS, a few of them would have to contribute a very large share. The Member State with the highest contribution would be LU (96.24%), followed by IE (57.67%) and MT (43.32%)lxxvii.


The funding needs in the EU as a whole when opting for a single EU-wide IGS covering cross-border insurance activity (branches only) complemented by national IGS in all Member States for domestic business are overall the same as under option 5.2. The funding needs specific to domestic national IGS and those for the single EU-wide IGS covering cross-border insurance activity are shown in Table 9.


Table 9 - Funding needs for the EU with national schemes for domestic activity supplemented by an EU-wide IGS covering all cross-border (branches only) activity (m €)


Source: Methodological report, Table 4.24. For a full analysis and figures for each Member State see sub-sections 4.4 and A5.3 of the MR.


While a majority of Member States would not contribute much of their funding endowments to the EU-wide cross-border IGS, a few of them would have to contribute a significant share. The Member State with the highest contribution would be EE (32.62%), followed by IE (12.21%) and CY (12.56%)lxxviii.


- Option 5.5: An EU-wide IGS that only covers insurance undertakings that are part of a group supervision regime (including subsidiaries) plus national schemes for all other relevant policies (domestic and cross-border)


It would be possible to set up an EU-wide IGS that covers only those insurers who are part of a strengthened group supervision regime – yet to be established - or that covers only systemically important insurers (including their subsidiaries). However, this is likely to create the same complications of option 5.4. In addition, there is the question-mark over the financing of such a scheme, as this would mean that Member States supervising a large group would have to compensate for the failure of the entire group throughout Europe. Finally, in highly concentrated markets it could be very difficult for the remaining insurers to set up a national scheme.


The 2008 public consultation showed that an EU wide guarantee fund for insurance undertakings which are part of a group supervision regime was supported by some respondents.


In the 2009 informal meetings with stakeholders, the EFRP showed support for this option.


- Option 5.6: Complement existing IGS with a 28th regime


Existing national IGS could be complemented by a 28th regime. Whilst options 5.3 and 5.4 would replace national schemes for cross-border insurance activity, a 28th regime would simply complement national IGS. Depending on its design, a 28th regime might not only add additional complexity to the system, but could also cause the same complications that arise under option 5.4.


The funding needs EU wide for Option 1.6 depend on the operational characteristics of the national IGS and of the 28th regime. Given certain assumptions, the overall funding needs might be the same as under Option 5.2.


- Option 5.7: Introducing an IGS in all Member States complemented by a system of mutual support between national IGS


With a mutual support system between national IGS, any scheme that lacks sufficient funds would be financially supported by all the other schemes. To ensure that the potential costs are transparent and predictable, such a system would require an agreed fund-raising mechanism setting out the proportion each IGS is contributing, and under which circumstances.


Such a mechanism might create moral hazard problems and Member States that have not experienced any insurance failure may feel that they are subsidising failing insurers (and their customers) in other Member States.


The funding needs for this option depend on the operational characteristics of the national IGS and on the characteristics of the mutual support. Given certain assumptions the funding needs for the whole EU might be the same as under Option 5.2.


In its advice, CEIOPS (2009b) argues that the main advantages of this option are a mitigation of the funding problems particularly in small national markets because of the concentration in those markets and a broader distribution of the losses. CEIOPS also indicates that if there are national differences between the scopes of coverage of the various national IGS, the mutual support system should be limited to the scope harmonised across the EU. CEIOPS observes that this may limit the benefits from the creation of a mutual support system. Finally, CEIOPS also notes that it would not be fair if a national IGS that is funded ex ante supports the IGS of another Member State which is funded ex post, at least if the financial difficulties of the latter could have been avoided through ex ante funding.


Table 10 presents a summary of the arguments which make the Commission come forward in the White Paper with the solution of an IGS in all Member States as a preliminary preferred solution.


Table 10 Summary of policy options' evaluations – Level of IGS centralisation (Single EU-wide scheme vs. national schemes)
Op. ObjectivesConstraints
EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
5.1 No action at EU level
5.2 An IGS in all MS++++----++
5.3 Single EU-wide IGS++++------------
5.4 EU-wide IGS cross-border activity + national schemes++++------------
5.5 EU-wide IGS lead-supervisor + national schemes+++------------
5.6 28th regime+++-----------
5.7 IGS in all MS + system of mutual support+++++------------+


Role of an insurance guarantee scheme


- Option 6.1: No action (harmonization) at EU level


The understanding of the Commission is that all existing IGS perform the role of a last resort mechanism.


In the 2009 informal meetings with stakeholders, FINUSE and representatives of the DGS sector supported the view of leaving decisions on the role of an IGS to Member States.


- Option 6.2: IGS as a last resort protection mechanism


The role of an IGS as a last resort protection mechanism is to protect policyholders, but not to prevent a crisis or to stop an insurance undertaking from getting into financial difficulties or becoming insolvent. This is the job of the supervisory authority and of other prudential regulatory tools such as solvency requirements.


In its advice, CEIOPS (2009b) recommends that IGS are set up as a last resort protection mechanism.


In the 2009 informal meetings with stakeholders, CEA, AMICE, CEIOPS and EFRP were concerned that extending the role of an IGS to preventing the failure of an insurance undertaking would create competitive distortions and increase moral hazard. They also considered it unfair that industry might be called upon to help a competitor that would eventually stay in the market. CEA, AMICE, CEIOPS and EFRP were in favour of limiting the role of IGS to providing last resort protection.


- Option 6.3: Preventing failure prevention and providing last-resort protection


IGS may also be that of intervene to prevent the failure of an insurance undertaking. The IGS would guide the insurance undertaking through its financial difficulties and ensure that it stays in business.


Giving an IGS this wider role presents important disadvantages. Indeed, it creates distortions of competition and increases moral hazard. It may also be considered unfair to ask other insurance undertakings to help a competitor stay in business.


It should be remembered that, compared to the banking sector, there is less cause for concern in the insurance sector over liquidity problems (leading to failure). Consequently, introducing a 'preventing failures' function in the insurance field does not seem as valuable as it might be in the banking sector.


Table 11 presents a summary of the arguments which make the Commission support the view that the role of an IGS should be that of solely acting as a last resort protection mechanism.


Table 11 Summary of policy options' evaluations – Role of an IGS
Op. ObjectivesConstraints
EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
6.1 No action at EU level
6.2 Last resort protection+++-----++
6.3 Prevention of failure and last resort protection++++-----------+


Geographical scope


- Option 7.1: No action (harmonization) at EU level


Some national IGS are based on the home country principle (for life insurance: BG, DE, ES, FR, RO; for non-life insurance: DK, ES, FR, IE, RO), while others (for life insurance: MT; for non-life insurance: MT, NO) are structured around the host country principle, and some others (for life insurance: LV, PL, UK; for non-life insurance: LV, UK) are structured under the home plus host country principle.


Preserving the status quo would maintain existing conditions of insufficient and uneven protection of policyholders in several Member States.


- Option 7.2: Home country principle


The main advantage of the home country principle (see Endnote Error: Reference source not found) is its consistency with the supervisory framework, since the home country supervisor is responsible for prudential regulation, including solvency requirements, and for starting the winding-up process. This is of particular importance if the IGS serves to facilitate the transfer of portfolio. Moreover, the home country principle is the principle followed for guarantee schemes in the banking and securities sector. Finally, an important consideration supporting the home country principle is that the administration of an IGS is closely linked with rules regarding insolvency and liquidation, which are under the responsibility of the home Member State. Funding needs are as assessed under option 5.2.


The Oxera report states that the main advantages of the home country principle are: correspondence with the current supervisory structure, ease and efficiency in handling cases, acceptance by the insurance industry, alignment with the approach followed for DGS and ICS, ease of treatment of insurance default cases and administrative feasibility. The main disadvantages identified in the report are: the possible uneven protection of consumers within a Member State (if the level of protection is not harmonised), unlevel playing field between domestic and non-domestic insurers (if the level of protection is not harmonised), incentives to moral hazard of public authorities in case of preponderant cross-border (compared to domestic) activity of domestic insurers.


In the 2008 public consultation, the vast majority of respondents showed a preference for the home country principle, as this approach is consistent with the EU supervisory framework. Most respondents were however not in favour of the idea that subsidiaries should also participate in and be covered by the IGS of the Member State in which the group supervisor is located (in case of a group support regime).


In its advice, CEIOPS (2009b) expresses a preference for the home country principle.


In the 2009 informal stakeholders meetings, all participants supported the home country principle.


- Option 7.3: Host country principle


The main advantage of a host country principle (see Endnote Error: Reference source not found) is that it ensures that there is no uneven policyholder protection in all Member State, thus preventing any possible distortions of the level playing field between insurers competing in the same country.


Adopting a host state principle nonetheless has its drawbacks. First, it duplicates administrative costs as it requires insurers with cross-border business to take part in two or more IGS. Second, IGS intervention might be difficult in practice: the authorities that operate the scheme would not be the ones that conduct and supervise the winding-up proceedings, and this is likely to cause difficulties.


If one opts for a national IGS in each Member State based on the host country principle, the funding needs are as shown in Table 12. Funding needs would be reduced by some 0-1% compared with home country based national schemes in each Member State.


Table 12 - Funding needs for the EU with national host country principle IGS, compared with national home country principle IGS (m €)


Source: Methodological report, Table 4.14. For a full analysis and figures for each Member State see sub-sections 4.2 and A5.1 of the MR.


Opting for a national IGS in each Member State according to the host country principle instead of the home country principle is estimated to have important distributional (between Member States) effects (see sub-section 4.2 of the MR and Table 4 of Annex 5 to the MR). While in the great majority of Member States funding needs change only slightly (between -5% and +5%), they change considerably for a few Member States. In particular, the most important increase in funding needs is in NO (14.78%) as the Norwegian insurance market is covered for some 17% by branches of insurance undertakings based in other EU-EEA countries. The biggest decrease in funding needs is in EE (-30.00%) as some 33% of the activity of Estonian insurance undertakings takes place via branches in other EU Member States.


The picture changes somewhat if IGS protection is limited to cover life policies. While for most Member States the variation in the funding needs is relatively modest, the situation becomes important for a few Member States. In particular, the most important increases in funding needs are in LV (43.40%), LT (13.78%) and CZ (13.31%) as in these Member States the market is covered to a large extent by branches of insurances established in other EU-EEA countries. The most important decreases in the funding needs are in EE (-53.39%) and Cyprus (-12.87%) as from these Member States a part of life insurance policies are sold in other EU Member States via branches.


The picture changes again when IGS protection is limited to non-life policies. While for most Member States the variation in the funding needs is relatively modest, the change becomes important for a few Member States. In particular, the most important increase in funding needs is NO (72.36%) as in NO the market is covered for more than half by branches of insurance undertakings based in other EU-EEA countries. The most important decreases in funding needs are in IE (-35.54%) and DK (-16.71%) as a part of non-life insurance policies are sold via branches from these Member States in other EU Member States.


The Oxera report states that the main advantages of the host country principle are: an even level of protection of all policyholders in all Member States and a level playing field between insurers competing in the same Member State. The main disadvantages are: possible double payments of insurers with cross-border business (if there is an IGS under the home state principle in their Member State of origin), misalignment with the supervisory structure, possible difficulties in the treatment of insurance undertakings' defaults, and difficult acceptance by the insurance industry and by supervisors.


In the 2008 public consultation only very few respondents showed support for the host country principle.


- Option 7.4: Home plus host country principle


The home plus host country principle can bring an effective solution to the problems of an isolated country. When adopted at the EU level, however, it does not provide substantial additional benefits compared to the home country principle (in the case of a sufficient harmonisation of the IGS design features). The benefits of this regime are far outweighted by the drawbacks/complexity added to the system.


- Option 7.5: Home country principle with lead supervisor


The Oxera report states that the main advantages of a national IGS with a lead supervisor are the following: an even consumer protection across Member States in the event of failure of an entity belonging to the group, the neutrality of the IGS with respect to the decision to enter a Member State via a branch or a subsidiary. The disadvantages identified in the Oxera report are: uneven consumer protection in Member States for insurance undertakings authorised and supervised in that Member State, unlevel playing field between insurers competing in the same Member State, incentive for moral hazard behaviour of public authorities in case of preponderant cross-border (compared to domestic) activity of domestic insurers.


In conclusion, the Commission believes at this stage that the adoption of IGS on the basis of the home country principle provides the most benefits to consumers and minimises the problems of implementation. This solution is therefore put forward as the Commission's preferred option. Table 13 presents a summary of the arguments supporting this conclusion.


Table 13 Summary of policy options' evaluations – Geographic Scope
Op. ObjectivesConstraints
EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
7.1 No action at EU level
7.2 Home state principle+++-----+++
7.3 Host state principle++++---------+
7.4 Home + Host principle+++++----------++
7.5 Home state principle with lead supervisor perimeter+++--------++


Policies covered


- Option 8.1: No action (harmonization) at EU level


In those countries that have already set up an IGS, the scope of coverage is rather heterogeneous: BG, DE, and PL only cover life policies, while ES, FR LV, MT, RO and UK protect both life and non-life policies. Finally, DK, IE and NO only protect non-life policies.


Preserving the status quo means that uneven protection of policyholders within and between Member States is maintained.


- Option 8.2: Protection of life policies only


The collapse of a life insurer can often cause very severe financial hardship for large groups of consumers. It is therefore advisable to include life policies in an IGS in order to provide a high level of protection to retail consumers.


In the 2009 informal stakeholders meetings, CEA reported that a majority of its members considered were in favour of covering only life policies as the practical relevance of covering non-life policies was lower. AMICE and EFDI supported this view.


It should however be noted that even if the average loss to policyholders is generally smaller in the case of a non-life insurer going into default, there are instances where losses to individual policyholders and third party claimants may well exceed that of a typical life insurance product.


- Option 8.3: Protection of non-life policies only


The severe consequences which may result for policyholders from failure of a life insurer make this option not advisable.


- Option 8.4: Protection of both life and non-life policies


Since substantial losses can be passed on to the holders of both life and non-life policies, policyholders will receive a more complete and appropriate protection if the EU acts to protect both types of policy – albeit in different ways and under different rules. However, doubts exist, also in view of the comments of some stakeholders, on whether this full coverage is entirely justified.


In its advice, CEIOPS (2009b) recommends that IGS cover both life and non-life policies.


In the 2009 informal stakeholders meetings, CEIOPS and FINUSE supported the protection of both life and non-life policies.


- Option 8.5: Protection of both life and selected non-life policies


The Oxera report argues that there may be reasons to exclude particular classes of non-life insurance from protection and to include only liability insurance, compulsory insurance and retail policies. The Oxera report also states that if the IGS protection is limited to natural persons only, a case can be made for excluding certain policies (e.g. marine, aviation and transit) from the scope of IGS protection as they cover commercial risks only.


In the 2009 informal stakeholders meetings, the EFRP explained that if the IGS is a genuinely last resort protection measure, life and a selection of non-life policies should be covered.


Table 14 presents a summary of the reasons why the Commission believes that preference should be given to covering life and selected non-life insurance policies.


Table 14 Summary of policy options' evaluations – Policies covered
Op. ObjectivesConstraints
EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
8.1 No action at EU level
8.2 Life policies only++++-----++
8.3 Non-life policies only++------+
8.4 Life and non-life policies++++++-------++
8.5 Life and selected non-life policies++++++------+++


Eligible claimants


- Option 9.1: No action (harmonization) at EU level


Existing IGS restrict the eligibility of claimants in various uncoordinated ways. In life insurance, while a majority of Member States (DE, ES, FR, MT, RO) provide coverage to basically all policyholders, two Member States (BG, UK) only provide coverage to natural persons and SME, and two other Member States (LV, PL) only provide coverage to natural persons. In non-life insurance, some Member States (ES, FR, RO) provide coverage to all policyholders. Some other Member States (IE, LV, MT) only protect natural persons, and three Member States (DK, NO, UK) protect both natural persons and SME.


Preserving this situation is not advisable as it would maintain uneven levels of protection within and between Member States for various classes of policyholders.


- Option 9.2: Natural and legal persons


Covering all natural and legal persons might be excessively expensive. It may also not be fully justified because of the main objective of IGS, i.e. the protection of retail customers.


- Option 9.3: Natural and selected legal persons (including SME's)


In order to reduce funding needs, eligibility could be restricted to those claimants who meet certain criteria.


One possibility might be to exclude large corporate policyholders from protection of non-compulsory insurance policies. Not only are these policyholders better equipped to assess the financial soundness of insurers, but they also have access to a network of insurance brokers who can scan the market and find insurers with the skills, capacity and financial strength to underwrite the risk. Finally, large corporate policyholders can also diversify their risks by purchasing policies with various insurance companies or seek other forms of protection.


Another possibility could be to limit coverage to natural persons and SME's. In that case, particular care would have to be taken in defining an SME.


The Oxera report indicates that eligible claimants should be consumers and possibly small businesses.


In the 2009 informal stakeholders meetings, FINUSE expressed a view in favour of protecting natural persons and SME's. FINUSE acknowledged however that it would be difficult to give a proper definition of SME.


- Option 9.4: Natural persons only


One possibility is to restrict IGS protection to natural persons only. However, this might raise concerns about inadequate protection for legal persons that resemble retail customers.


In its advice, CEIOPS (2009b) recommends that eligible claimants should be at least all natural persons, and that Member States should be allowed to extend the scope of coverage to other claimants.


In the 2009 informal stakeholders meetings CEA and EFDI expressed preference for protecting natural persons only as the main objective of IGS is consumer protection and not company protection. EFDI suggested that Member States should be given the possibility to introduce additional cover for SME's at national level.


Table 15 presents a summary of the reasons why the Commission believes that eligibility of natural and selected legal persons (including SME's) should be retained as the preferable option.


Table 15 Summary of policy options' evaluations – Eligible claimants
Op. ObjectivesConstraints
EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
9.1 No action at EU level
9.2 Natural and legal persons++++-------++
9.3 Natural and selected legal persons (incl. SME)++++------+++
9.4 Natural persons only++-----+


Timing of funding


- Option 10.1: No action (harmonization) at EU level


Currently, the majority of existing schemes covering life insurance are funded ex-ante or involve a sizeable element of ex-ante funding (BG, DE, ES, FR, LV, MT, RO). Exceptions are UK and PL. In non-life insurance, Member States with an important component of ex-ante funding are DK, ES, FR, LV, MT and RO, while the UK, NO and IE have ex-post funded IGS.


The lack of any IGS harmonisation at EU level would imply maintaining a situation in which there are considerable differences between schemes in terms of when contributions are collected. These differences have an impact on the protection of policyholders in Member States.


In the 2009 informal stakeholders meeting, CEA, EFDI and AMICE expressed a preference for no harmonisation at EU level with regard to the timing of funding. On the other hand, CEIOPS argued that harmonisation of the timing of funding was important.


- Option 10.2: Ex-post funding


In an ex-post funded scheme, resources remain with the contributing institutions until a failure occurs, and levies are paid to the scheme only once losses arise. It follows that set-up and operational costs are limited. Ex-post funding is more subject to moral hazard as failed institutions never contribute to the IGS.


The Oxera report states that the main advantages of ex-post funding are: the very low set-up and administrative costs, the lower cost for insurance undertakings, the possibility that collected funds are tailored on actual default losses. The main disadvantages are: the difficulty to ensure a prompt pay-out to policyholders, the fact that failed insurance undertakings do not contribute to the loss caused by their failure, that funds are collected in a possibly more pro-cyclical way, the fact that it might in the end not be possible to collect funds from the insurance industry due to their weak general conditions.


- Option 10.3: Ex-ante funding


In a pre-funded scheme, funds are raised in anticipation of possible future failures, with resources transferred to, and managed by, the IGS via a system of levies on industry. The first advantage therefore is the fact that money is readily available to protect consumers should a failure occur. Moreover, ex-ante funding is less subject to moral hazard problems because insurers that become insolvent will have already contributed to the IGS.lxxix Finally, ex ante funding is more likely to avoid the pro-cyclicality associated with ex-post funded schemes. It is obvious that set-up and operational costs tend to be higher here than in the case of ex-post funding.


The Oxera report states that the main advantages of ex-ante funding are: that funds are in principle more quickly available to the IGS, that failed insurance undertakings contribute to the loss of their failure, that funds are collected in a possibly less pro-cyclical way. The main disadvantages are: the higher set-up, administrative and operational costs and the possibility that collected funds are insufficient (if not complemented by ex-post funding).


In its advice, CEIOPS (2009b) recommends that IGS should be required to make payments as soon as practicable after claims have been assessed. However CEIOPS does not express itself clearly in favour of ex-ante or ex-post funding.


In the 2009 informal stakeholders meeting, CEA stated that the insurance industry saw no merit in ex-ante funding.


- Option 10.4: Combination of ex-post and ex-ante funding


When part of the IGS funding is ex-ante and part is ex-post, some of the funds would be immediately available to the IGS without imposing too high ex-ante costs / mobilization of funds on industry and consumers.


In the 2009 informal stakeholders meeting, CEIOPS favoured a combination of ex-ante and ex-post funding.


In conclusion, the Commission believes that an appropriate combination of ex-ante and ex-post funding is preferable. Table 16 presents a summary of the arguments that support this conclusion.


Table 16 Summary of policy options' evaluations – Timing of funding
Op. ObjectivesConstraints
EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
10.1 No action at EU level
10.2 Ex-post funding++-------
10.3 Ex-ante funding++++----------+
10.4 Combination of ex-post and ex-ante funding++++---------++


Nature of scheme intervention (portfolio transfer/compensation)


IGS can work in different ways. In the first scenario, IGS secure the continuity of the policies by, for instance, facilitating their transfer to a solvent insurer or taking direct charge of them (portfolio transfer). In the second scenario, IGS compensate policyholders or beneficiaries for their losses if an insurance undertaking becomes insolvent (compensation of claims).


- Option 11.1: No action (harmonization) at EU level


In life insurance, existing IGS are split between those that provide compensation (BG, LV, MT, PL, RO) and those that ensure portfolio transfer (DE, FR, ES, UK). In non-life insurance the majority of IGS provide compensation (DK, IE, LV, MT, RO, UK), while portfolio transfer is ensured in FR, NO and ES.


The absence of any harmonisation at EU level concerning the nature of the intervention of IGS schemes would imply maintaining a situation of uneven levels of policyholder protection in Member States.


In its advice, CEIOPS (2009b) recommends to leave flexibility to Member States regarding the question whether IGS should only deal with compensation of claims or portfolio transfer.


- Option 11.2: Portfolio Transfer


From a policyholder protection point of view, continuity of insurance cover may be more advantageous than compensation, particularly in those cases where policyholders would otherwise find it difficult to get equivalent cover (on similar terms) with an alternative insurer.


The Oxera report notes that from a consumer protection point of view and in order to limit wider market impacts, continuity may be preferable, particularly for life insurance policyholders. In the case of non-life insurance, the arguments for continuity may be less relevant, since contracts are generally short-term. Nevertheless there may be instances where there could be benefits for an IGS to secure continuity of non-life policies e.g. where policies are ‘non-standard’ or the failed insurer has a significant share of the market and it is difficult for policyholders to find alternative cover quickly at the same price if supply is restricted.


- Option 11.3: Compensation of Claims


All the funding needs presented above for the various options of the IGS design features are based on the assumption of portfolio transfer. If IGS provide compensation (only) both for life and non-life insurance, the funding needs for the whole EU under option 5.2 are those presented in Table 17.


Table 17 - Funding needs for total insurance for the EU with national IGS providing compensation only and comparison with the portfolio transfer case (m €)


Source: Methodological report, Table 4.31. For a full analysis and figures for each Member State see sub-sections 4.6 and A5.5 of the MR.


If IGS provide compensation (only) for life insurance only, the funding needs for the whole EU under option 5.2 are those presented in Table 18.


Table 18 - Funding needs for life insurance for the EU with national IGS providing compensation only and comparison with the portfolio transfer case (m €)


Source: Methodological report, Table 4.35. For a full analysis and figures for each Member State see sub-sections 4.6 and A5.5 of the MR.


Finally, if IGS provide compensation (only) for non-life insurance only, the funding needs for the whole EU under option 5.2 are those presented in Table 19.


Table 19 - Funding needs for non-life insurance for the EU with national IGS providing compensation only and comparison with the portfolio transfer case (m €)


Source: Methodological report, Table 4.37. For a full analysis and figures for each Member State see sub-sections 4.6 and A5.5 of the MR.


The total funding needs tend to be lower (some 7%) because not all policies need to be protected, but only those that actually lead to a real claim against the insurer.


Table 20 presents a summary of the reasons why the Commission believes that portfolio transfer is the preferred option.


Table 20 Summary of policy options' evaluations – Nature of scheme intervention
Op. ObjectivesConstraints
EffectivenessCostsIncentivesEase Imp.Cost-effectiveness
OptionsHigh policyholders protectionEven policyholders protectionResources neededCorrect allocation of lossesSet-up and operational costsConsumers moral hazardInsurers moral hazardSupervisors / public authorities moral hazardConsensus among stakeholdersAdministrative burden
11.1 No action at EU level
11.2 Portfolio Transfer++++------+++
11.3 Compensation of claims++----++








B. METHODOLOGICAL REPORT ON THE DERIVATION OF IGS LOSS DISTRIBUTIONS

See separate document








C. SUPPLEMENTARY TABLES TO THE METHODOLOGICAL REPORT


1. Additional statistics on distribution of losses and balance sheets of insurers


Figure 0.3 - Insured losses from natural catastrophes and man-made disasters throughout the world


Source: European Central Bank (European Central Bank 2009)


Figure 0.4 - Stock prices in the Euro area


Source: European Central Bank (European Central Bank 2009)


Figure 0.5 - Investment-grade corporate bond spreads in the Euro area


Source: European Central Bank (European Central Bank 2009)


Figure 0.6 - Write-downs and losses at selected insurance companies (since beginning 2007, total of USD 261.2 billion)


Source: Schich (2009)


Figure 0.7 – Expected default frequency for the euro area insurance sector


Source: European Central Bank (European Central Bank 2009)


2. Additional descriptive statistics


Table 0.21: Estimated average number of defaults per year and average time between defaults.
Average number of defaults
per year (PD=0.1%)
Average number of defaults
per year (PD=0.5%)
Years between defaults
(PD=0.1%)
Years between defaults
(PD=0.5%)
AT0.080.3913.02.6
BE0.160.786.41.3
BG0.040.2025.05.0
CY0.040.1827.85.6
CZ0.050.2619.23.8
DE0.502.522.00.4
DK0.190.975.21.0
EE0.020.1052.610.5
ES0.291.463.40.7
FI0.040.1828.65.7
FR0.391.942.60.5
GR0.080.4012.52.5
HU0.050.2420.84.2
IE0.351.742.90.6
IS0.010.0683.316.7
IT0.241.224.10.8
LI0.060.3215.93.2
LT0.030.1435.77.1
LU0.361.782.80.6
LV0.020.1145.59.1
MT0.040.2223.34.7
NL0.301.503.30.7
NO0.130.677.51.5
PL0.080.4112.32.5
PT0.080.4112.22.4
RO0.040.2123.84.8
SE0.211.034.91.0
SK0.040.1828.65.7
SL0.020.1052.610.5
UK0.432.142.30.5
EU4.1520.740.240.05
EEA4.3621.790.230.05
EU avg0.291.463.40.7
EU-EEA avg0.281.423.50.7

Note: based on average probabilities of default and ignoring correlation. EU and EEA averages are weighted by number of insurers in each country in 2007.

Source: CEIOPS data, own elaboration

Table 0.22: Losses of historical selected defaults.
CountryBiggest failureYear of defaultSectorTotal l oss (m€)Total loss
(as % of total premiums)
Number of policyholders / claims
ROMetropol2003Composite2.90.2%8 427 (3 573 paid)
FREuropavie2000Life0.40.0%N.A.
DEMannheimer2003Life100.00.1%344 000
IEICI1985Non life315.08.1%N.A.
ESReunión1992Non life35.40.1%N.A.
FRInternational Claims Services SA1999Non life10.20.0%260
UKIndependent Insurance2001Non life738.00.8%190 000
UKChester Street2001Non life146.50.2%N.A.
DKPlus Forsiking A/S2002Non life12.40.2%N.A.

Source: Oxera report(Oxera 2007) and CEIOPS updates (CEIOPS 2009b; CEIOPS 2009a)


Table 0.23: Estimated average and largest losses under different default probability scenarios (m €).
LifeNon-Life
Average Loss PD=0,1%, (LGD=15%)Average Loss PD=0,5% (LGD=15%)Largest loss (LGD=15%)Average Loss PD=0,1%, (LGD=15%)Average Loss PD=0,5% (LGD=15%)Largest loss (LGD=15%)
AT8.7343.642'117.931.658.24379.80
BE25.22126.125'491.502.8914.43479.11
BG0.030.157.440.030.164.28
CY0.412.04107.480.050.268.21
CZ0.984.91312.040.281.4166.06
DE114.78573.8910'662.7937.30186.483'335.71
DK17.7188.572'707.701.517.56273.26
EE0.080.3836.980.020.085.49
ES24.74123.702'418.827.5137.561'372.05
FI5.5627.821'705.751.185.92326.19
FR178.44892.2229'584.4825.21126.053'505.44
GR1.145.72207.780.251.2737.12
HU0.793.96120.900.050.2610.86
IE22.12110.584'099.662.0110.07436.71
IS0.020.119.860.100.4929.88
IT58.37291.8415'157.574.8924.471'086.66
LIN.A.N.A.N.A.N.A.N.A.N.A.
LT0.080.3918.940.020.124.88
LU11.4957.431'627.130.532.67130.87
LV0.010.065.050.030.145.69
MT0.190.9738.780.090.4414.26
NL39.95199.748'880.7612.3961.972'499.79
NO11.9259.603'947.701.175.85260.47
PL2.5612.79981.340.522.62243.93
PT6.0430.221'303.300.753.74244.26
RO0.120.5932.870.100.4826.30
SE28.73143.634'735.418.0540.272'018.13
SI0.311.53115.490.221.0984.02
SK0.341.7290.300.070.3729.23
UK305.101'525.5027'864.4215.5377.672'533.59
EU854.034 270.14120 432.62123.16615.7819 161.88
EU-EEA865.974 329.85124 390.18124.42622.1219 452.23

Note 1: Numbers in Italic refer to estimates based on approximate market structure

Note 2: losses are calculated assuming a Loss Given Default of 15%

Source: Methodological report, Table A4.1 and own elaboration.


Table 0.24: Updated calculation of average and stressed Probabilities of Default (PD).
Rating Grade (S&P)Probability of default over one year (S&P)Number of Leading European Insurance groups in each rating class, by year
In 2008
(during financial crisis)

Average
(1981-2008)

20052006200720082009
AAA0000000
AA+0000000
AA0.43%0.02%22331
AA-0.40%0.03%57765
A+0.31%0.05%65886
A0.21%0.06%66339
A-0.58%0.08%66555
BBB+0.18%0.16%01110
BBB0.59%0.28%11100
BBB-0.71%0.28%00022
BB+1.14%0.68%00000
BB0.63%0.89%00000
Average0.404%0.065%
Adjusted average (to account for unrated companies)0.100%

Note: Average PD is calculated as weighted average of average historical PD over period 1981-2008) weighted by number of companies in each rating class over last 5 years. Average PD in 2008 is calculated as weighted average of observed default rates during 2008 weighted by number of companies in each rating class in 2008.

Source: CEIOPS, Standard&Poor's, Oxera report, own elaboration


Table 0.25: Updated table of IGS characteristics (Part I)


Table 0.26: Updated table of IGS characteristics (Part II)


Notes: Belgium is not included as reported by Oxera as the Belgian IGS only has one participant; (a.1) only in case of a long term life insurance continuation; (c.1) maximum payout for any single insurer capped to MTL 1mil (around Eur 2'329'000); (c.2) policyholders 90%, third party claimants 100%; (c.3) individual claims are unlimited but there is a total payout limit of 700m euro; (c.4) 100% is for residential property and compulsory liability insurance; (d.1) levies are raised for costs expected during the next 12 months; (f.1) home state for protection and host state for contribution; (f.2) participation of foreign branches not required and not permitted; (f.3) all contracts not covered by a home scheme need to be covered by the scheme; (f.4) unless branches of EU insurer protected to an equivalent level; (f.5) mandatory for insurers providing insurance which is mandatory by law or regulation; (f.6) required to participate but protected only if wound up under Irish law; (f.7) branches not protected but required to contribute for non-life risks located in Spain; (f.8) the fund covers claims arising under a contract protecting a a risk situated in Malta or originating a commitment in Malta; (i.1) Public ownership and management, but formally a private right corporation; (j.1) The fund responds to the National Supervisory Authority but it is not foreseen that it will receive staff in case of a default; (j.2) Privately managed, with a board appointed by regulator; (j.3) Managed by representatives chosen by industry and vetted by supervisor. Subject to supervision of insurance supervisor.

Sources: CEIOPS update to the Oxera report (CEIOPS 2009b; CEIOPS 2009a)


Table 0.27: Estimated funds available to existing IGS (m €).
Estimated funds availableSector
BG*0,70Life
DE640,00Life
DK40,30Non life
ES1 331,00Life +Non life
FR569//250Life//Non life
IE*26,48Non life
LV0.8//2.8Life//Non life
MT2.33//2.33Life//Non life
NO*16,04Non life
PL*39,03Life
RO17.10//84.50Life//Non life
UK*1 766//316Life//Non life

Note 1: * – ex-post funded scheme

Note 2: Funds available for schemes with ex-ante payment are based on figures reported by Oxera and CEIOPS. Funds available for schemes with ex-post payment are estimated (numbers in italics) based on average fund size of ex-ante schemes with respect to Gross Premium Written

Source: Oxera Report (Oxera 2007), CEIOPS update (CEIOPS 2009b; CEIOPS 2009a), own elaboration


Table 0.28: Estimated funding needs as a share of gross premiums written in each country.
PD=0.5%PD=0.1%
α=75%α=90%α=99%α=75%α=90%α=99%
AT0.32%0.98%4.74%0.04%0.17%1.18%
BE0.42%1.28%6.16%0.06%0.22%1.54%
BG0.08%0.21%0.88%0.01%0.04%0.22%
CY0.34%1.08%5.46%0.05%0.18%1.36%
CZ0.18%0.51%2.27%0.03%0.09%0.57%
DE0.52%1.33%5.21%0.08%0.25%1.32%
DK0.56%1.42%5.54%0.09%0.26%1.41%
EE0.10%0.47%3.68%0.01%0.06%0.85%
ES0.37%0.95%3.80%0.06%0.18%0.96%
FI0.45%1.67%10.00%0.05%0.26%2.43%
FR0.51%1.38%5.88%0.07%0.25%1.49%
GR0.19%0.51%2.20%0.03%0.09%0.55%
HU0.17%0.41%1.57%0.03%0.08%0.40%
IE0.28%0.74%3.11%0.04%0.13%0.79%
IS0.18%0.64%3.70%0.02%0.10%0.90%
IT0.35%1.02%4.74%0.05%0.18%1.19%
LI0.00%0.00%0.00%0.00%0.00%0.00%
LT0.13%0.38%1.67%0.02%0.07%0.42%
LU0.58%1.37%4.92%0.09%0.26%1.25%
LV0.07%0.23%1.08%0.01%0.04%0.27%
MT0.24%0.80%4.28%0.03%0.13%1.06%
NL0.30%0.85%3.78%0.04%0.15%0.95%
NO0.32%1.24%7.91%0.04%0.18%1.90%
PL0.12%0.44%2.51%0.01%0.07%0.61%
PT0.23%0.73%3.67%0.03%0.12%0.91%
RO0.11%0.27%1.01%0.02%0.05%0.26%
SE0.74%2.21%10.51%0.10%0.38%2.63%
SI0.14%0.55%3.47%0.02%0.08%0.83%
SK0.14%0.45%2.41%0.02%0.07%0.60%
UK0.43%1.14%4.62%0.07%0.21%1.17%
EU0.43%1.17%4.92%0.06%0.21%1.24%
EU-EEA0.43%1.17%4.95%0.06%0.21%1.25%

Note: estimates based on the home state principle, under different scenarios for the ‘over the cycle’ probability of default and levels of protection.

Source: Methodological report, Table 3.2 - CEIOPS (CEIOPS 2008), CEA (CEA 2009), own elaboration.

Table 0.29: Ranking group, in terms of market share, of largest company in the Life insurance sector covered by IGS under different funding scenarios.
PD=0.1%
α=75%
PD=0.1%
α=90%
PD=0.1%
α=99%
PD=0.5%
α=75%
PD=0.5%
α=90%
PD=0.5%
α=99%
AT....2..21
BE..43432
BG..........1
CY....3....1
CZ....3..31
DE....4442
DK....3332
EE....2....1
ES..44442
FI333332
FR..44442
GR..44442
HU....2..32
IE....2..22
IS..........1
IT..43442
LI......332
LT..........2
LU..33333
LV222221
MT222222
NL....2..22
NO433331
PL....4441
PT..43432
RO332222
SE443433
SI....3..31
SK..43431
UK444443

Legend: 1 = company size rank is between 1 and 5; 2 = company size rank is between 6 and 10; 3 = company size rank is between 11 and 15; 4 = company size rank is below 15; .. = not defined using current data

Note: funding needs estimated under home state principle and different scenarios for mean “over the cycle” probability of default and coverage level

Source: Methodological report, Table 3.2, CEIOPS (CEIOPS 2008), CEA (CEA 2009), own elaboration.


Table 0.30: Ranking group, in terms of market share, of largest company in the Non life insurance sector covered by IGS under different funding scenarios
PD=0.1%
α=75%
PD=0.1%
α=90%
PD=0.1%
α=99%
PD=0.5%
α=75%
PD=0.5%
α=90%
PD=0.5%
α=99%
AT....2..21
BE..44442
BG....3..33
CY..........2
CZ..33332
DE444442
DK..33332
EE....2..21
ES..44441
FI432332
FR444441
GR....4..42
HU....3332
IE333332
IS............
IT443431
LI..33332
LT....3..32
LU..33332
LV333332
MT443432
NL..22222
NO..33332
PL....3441
PT..44442
RO..33332
SE443432
SI..32331
SK..43431
UK..44442

Legend: 1 = company size rank is between 1 and 5; 2 = company size rank is between 6 and 10; 3 = company size rank is between 11 and 15; 4 = company size rank is below 15; .. = not defined using current data

Note: funding needs estimated under home state principle and different scenarios for mean “over the cycle” probability of default and coverage level

Source: Methodological report, Table 3.2, CEIOPS (CEIOPS 2008), CEA (CEA 2009), own elaboration.


Table 0.31: Ratio of gross maximum losses to GDP under different loss scenarios.
PD=0.5%PD=0.1%
α=75%α=90%α=99%α=75%α=90%α=99%
AT0.0153%0.0468%0.2276%0.0021%0.0080%0.0568%
BE0.0354%0.1068%0.5127%0.0048%0.0183%0.1282%
BG0.0010%0.0026%0.0108%0.0001%0.0005%0.0027%
CY0.0114%0.0360%0.1817%0.0015%0.0060%0.0452%
CZ0.0048%0.0134%0.0595%0.0007%0.0024%0.0150%
DE0.0310%0.0791%0.3105%0.0048%0.0147%0.0788%
DK0.0449%0.1143%0.4467%0.0070%0.0213%0.1134%
EE0.0012%0.0058%0.0455%0.0001%0.0008%0.0105%
ES0.0150%0.0386%0.1538%0.0023%0.0071%0.0390%
FI0.0119%0.0436%0.2620%0.0014%0.0067%0.0636%
FR0.0497%0.1356%0.5768%0.0073%0.0244%0.1457%
GR0.0029%0.0080%0.0343%0.0004%0.0014%0.0087%
HU0.0045%0.0111%0.0423%0.0007%0.0021%0.0108%
IE0.0602%0.1621%0.6787%0.0089%0.0294%0.1717%
IS0.0027%0.0096%0.0554%0.0003%0.0015%0.0135%
IT0.0176%0.0517%0.2404%0.0025%0.0090%0.0603%
LIN.A.N.A.N.A.N.A.N.A.N.A.
LT0.0015%0.0043%0.0190%0.0002%0.0008%0.0048%
LU0.1707%0.4057%1.4575%0.0276%0.0777%0.3711%
LV0.0008%0.0024%0.0115%0.0001%0.0004%0.0029%
MT0.0202%0.0669%0.3559%0.0026%0.0108%0.0879%
NL0.0368%0.1045%0.4664%0.0053%0.0185%0.1174%
NO0.0138%0.0533%0.3401%0.0015%0.0079%0.0818%
PL0.0034%0.0121%0.0698%0.0004%0.0019%0.0171%
PT0.0166%0.0520%0.2603%0.0022%0.0087%0.0648%
RO0.0009%0.0022%0.0085%0.0001%0.0004%0.0022%
SE0.0452%0.1355%0.6448%0.0062%0.0233%0.1614%
SI0.0051%0.0197%0.1250%0.0006%0.0029%0.0301%
SK0.0029%0.0096%0.0509%0.0004%0.0016%0.0126%
UK0.0744%0.1954%0.7948%0.0112%0.0358%0.2014%
EU0.0366%0.0988%0.4164%0.0054%0.0179%0.1052%
EU-EEA0.0361%0.0976%0.4141%0.0054%0.0176%0.1045%

Note: funding needs estimated under home state principle and different loss scenarios for mean “over the cycle” probability of default and loss frequency.

Source: Methodological Report, Table 3.2, Table 3.2; Eurostat; own elaboration

3. estimates of Coverage and protection of cross border flows

Based on the approximate estimation of bilateral trade flows presented in the Annex on Cross Border Insurance Activity, it is possible to estimate the possible amounts of losses which could be passed on to policy-holders and claimants across borders. As in the Methodological report, due to the absence of more detailed data, a proportionality assumption is going to be used to attribute quotas of Exposure At Default to imported and exported flows of premiums.


Table 0.32: Estimated shares of domestic and imported premiums not covered, at least partially, by any IGS.
LifeNon LifeTotal Insurance
Domestic activity+ estimated importedShare of domestic activityEstimated share of total importDomestic activity+ estimated importedShare of domestic activityEstimated share of total importDomestic activity+ estimated importedShare of domestic activityEstimated share of total import
AT100%100%99%99%100%45%100%100%88%
BE100%100%98%92%100%22%98%100%71%
BG4%0%99%99%100%45%66%66%78%
CY100%100%95%96%100%33%98%100%78%
CZ99%100%93%98%100%36%99%100%87%
DE5%0%98%98%100%27%49%48%74%
DK100%100%99%1%0%47%76%75%94%
EE100%100%99%95%100%32%97%100%50%
ES4%0%99%1%0%45%2%0%88%
FI100%100%99%99%100%45%100%100%89%
FR4%0%99%1%0%49%3%0%95%
GR100%100%98%98%100%35%99%100%81%
HU100%100%99%99%100%45%100%100%93%
IE99%100%85%18%0%41%94%97%75%
IS100%100%99%99%100%45%99%100%65%
IT100%100%94%96%100%31%99%100%85%
LI99%100%99%45%100%45%99%100%98%
LT99%100%93%96%100%33%98%100%83%
LU100%100%97%90%100%30%88%100%84%
LV0%0%0%0%0%0%0%0%0%
MT0%0%0%0%0%0%0%0%0%
NL100%100%99%99%100%45%100%100%80%
NO100%100%97%55%100%0%88%100%22%
PL4%0%99%99%100%45%24%22%94%
PT100%100%99%99%100%45%100%100%95%
RO4%0%99%1%0%45%2%0%81%
SE100%100%99%99%100%45%100%100%91%
SI100%100%99%99%100%45%100%100%79%
SK100%100%99%99%100%45%100%100%93%
UK0%0%0%0%0%0%0%0%0%
EU avg26%25%62%56%57%23%35%34%54%
EU-EEA avg27%26%63%56%58%19%35%34%53%

Note: Import flows calculated using approximate estimate of bilateral flows based on proportionality assumptions. All exports exiting a country with a home principle IGS are considered covered (at least partially), all imports entering a country with a host state principle IGS are considered covered (at least partially).

Source: Table 2.1, Annex on cross-border insurance activity in the EU-EEA, Tables 2.6-2.8; own elaboration.


Table 0.33: Estimates of net losses after intervention of existing IGS under different loss scenarios (m €).
PD=0.5%PD=0.1%
α=75%α=90%α=99%α=75%α=90%α=99%
AT41.56126.82616.395.6321.59153.91
BE118.45357.711 717.4416.1761.22429.37
BG0.000.052.410.000.000.09
CY1.825.7428.980.240.967.21
CZ6.0617.0975.760.873.0319.08
DE113.311 281.866 899.740.000.001 273.98
DK61.72219.19973.910.007.97217.20
EE0.190.907.110.020.121.64
ES0.000.00288.010.000.000.00
FI21.3378.33470.362.4812.01114.23
FR122.761 749.6010 109.650.000.001 942.11
GR6.5818.0977.660.963.2519.61
HU4.5111.2342.790.712.1110.88
IE87.77281.051 261.370.0029.26299.25
IS0.401.438.270.050.222.02
IT272.42800.063 717.4037.98138.92932.40
LIN.A.N.A.N.A.N.A.N.A.N.A.
LT0.431.225.440.060.221.37
LU63.97152.01546.0810.3429.11139.03
LV0.000.000.000.000.000.00
MT0.000.0014.770.000.000.14
NL209.47594.482 652.5029.96105.03667.70
NO22.97135.12947.640.006.41215.76
PL0.000.00178.050.000.0014.00
PT27.0984.78424.363.6014.24105.62
RO0.000.000.000.000.000.00
SE149.65448.612 135.3420.5477.07534.33
SI1.766.8043.210.201.0110.40
SK1.605.2727.970.200.866.91
UK0.001 911.0214 163.420.000.002 033.31
EU1 312.458 151.9146 480.12129.96507.988 933.77
EU-EEA1 335.828 288.4647 436.03130.01514.619 151.55

Note 1: Loss scenarios as per table 3.2 of methodological report. IGS fund sizes as per Table 0 .27. Home principle loss distribution is used for all countries.

Note 2: Countries with an IGS in place are indicated in grey

Source: Table 0 .27, Methodological Report, Table 3.2.


Table 0.34: Estimates of losses exported to other countries after intervention of existing home principle IGS, under different loss scenarios (m €).
PD=0.5%PD=0.1%
α=75%α=90%α=99%α=75%α=90%α=99%
AT0.150.442.160.020.080.54
BE6.1118.4688.630.833.1622.16
BG0.000.000.000.000.000.00
CY0.391.246.260.050.211.56
CZ0.020.040.190.000.010.05
DE1.3113.6072.660.000.0013.51
DK3.9112.9556.270.000.8212.83
EE0.060.302.320.010.040.54
ES0.000.000.000.000.000.00
FI0.933.4220.530.110.524.98
FR6.7239.72209.270.000.0043.62
GR0.000.000.000.000.000.00
HU0.000.000.000.000.000.00
IE50.72162.18727.490.0016.99172.67
IS0.000.010.060.000.000.02
IT3.249.5244.220.451.6511.09
LIN.A.N.A.N.A.N.A.N.A.N.A.
LT0.000.010.030.000.000.01
LU61.57146.30525.569.9528.01133.80
LV0.000.000.000.000.000.00
MT0.481.588.420.060.262.08
NL0.000.000.000.000.000.00
NO0.712.7417.450.080.414.20
PL0.000.000.030.000.000.01
PT0.260.814.030.030.141.00
RO0.000.000.000.000.000.00
SE0.000.000.000.000.000.00
SI0.000.000.020.000.000.01
SK0.000.020.080.000.000.02
UK0.000.000.000.000.000.00
EU135.86410.581 768.1611.5151.89420.47
EU-EEA136.57413.331 785.6711.5952.30424.69

Note 1: Loss scenarios as per table 3.2 of methodological report. Losses exported calculated proportionally to export flows illustrated in tables 2.9 and 2.10 of Annex on Cross Border insurance activity in the EU-EEA. A quota of IGS funds proportional to the share of exports is used to reduce losses i.e. it is assumed that losses are equally distributed between domestic and cross-border activities).

Note 2: Countries with an existing home IGS in place are indicated in grey

Source: Table 0 .27; Annex on cross border insurance activity in the EU-EEA, Tables 2.9 and 2.10; Methodological report, Table 5 of Annex 5.


Table 0.35: Estimates of losses imported by any EU country under different loss scenarios, net of protection offered by existing home state principle IGS and existing host state principle IGS (m €).
PD=0.5%PD=0.1%
α=75%α=90%α=99%α=75%α=90%α=99%
AT1.123.4214.850.090.423.52
BE4.5413.8959.990.371.6914.19
BG #0.020.070.320.000.010.08
CY0.110.321.410.010.040.33
CZ1.213.7116.090.100.463.81
DE #18.0253.68231.041.526.9255.34
DK #1.755.3423.180.150.675.50
EE0.030.080.330.000.010.08
ES #3.6911.2848.970.311.4011.61
FI0.401.215.240.030.151.24
FR #17.5450.73214.701.547.0451.82
GR0.451.386.000.040.171.42
HU0.280.863.720.020.110.88
IE #4.0011.9050.570.531.6311.98
IS0.010.040.160.000.000.04
IT18.9958.13251.921.557.1459.64
LI0.320.974.190.030.120.99
LT0.130.411.760.010.050.42
LU0.632.2410.620.010.202.44
LV #0.000.000.000.000.000.00
MT0.000.000.000.000.000.00
NL5.0315.3866.790.421.9115.84
NO0.005.6674.340.000.0010.44
PL #0.000.000.000.000.000.00
PT1.233.7616.300.100.473.87
RO0.080.231.020.010.030.24
SE1.915.8525.380.160.736.02
SI0.090.271.160.010.030.27
SK0.120.361.570.010.040.37
UK #0.000.000.000.000.000.00
EU81.35244.501 052.937.0031.33250.93
EU-EEA81.68251.161 131.627.0231.46262.40

Note 1: Countries with an host IGS in place are indicated in grey; currently their whole fund is reduced with the average domestic losses; # indicates countries with a home state principle IGS

Note 2: Net losses scenarios as per Table 0 .34 (A quota of IGS funds proportional to the share of exports is used to reduce losses exported). Losses imported calculated proportionally to import flows illustrated in table 2.9 and 2.10 of Annex on Cross Border insurance activity in the EU-EEA. Losses imported by each country are reduced by amount of any host-state principle IGS present there. (The funds of the IGS are reduced by the average amount of expected losses generated by domestic companies).

Source: Table 0 .27; Annex on cross border insurance activity in the EU-EEA – Tables 2.9 and 2.10; Methodological Report - Table 5 of Annex 5.


4. Preliminary analysis: comparison of potential losses stemming from Aspis group insolvency with results of model used in the Methodological Report.


Aspis group, a large insurance group making up roughly 13% of the total insurance sector in Greece, has seen its license to trade in the insurance sector revoked in September 2009. In the run-up to the revocation, EPEIA (the Greek regulator) asked Aspis group to provide a financial resurrection plan for the first time in 2008. In 2009 EPEIA asked the company to provide a total of 237 Eur million in cash guarantees which, after negotiations with the company, were reduced to 203.5 Eur million. A final request for a financial resurrection plan was advanced by EPEIA in September 2009 and license was revoked upon failure by Aspis to produce such plan.


The amount of guarantees requested by the regulator can be interpreted as the best estimate of the capital gap which needs to be filled to allow the company to continue operations.

Using the terminology of the IGS IA Methodology Report, this corresponds to the concept of expected losses used in the in case of a `portfolio transfer` intervention.


By using a rather strong proportionality assumption, the part of this amount referring to losses in the non-motor lines of business3 can be estimated as 140 Eur million (see Table 0 .36)4.


The funding needs for an IGS, calculated based on Greek market data and different assumptions on default probabilities, are presented in Table 0 .37 (taken from MR Table 3.2).

As it can be seen, these numbers are much lower than the currently expected loss of 140 Eur million.

This can be explained by considering that, in terms of the probability distribution of losses generated by the model employed in the IGS Methodology Report, an insolvency of this size5 can be interpreted as an exceptional event, which should have happened in the Greek market with a probability lower than 0.5% or α above 99.5% (see Figure 0 .8)6.


Therefore, even in case an IGS holding funds between 0.55% and 2.2% of total gross premiums written in the Greek market had been in place, the losses would still have required a State intervention in order to grant protection to all claimants and policy holders.


Nonetheless, if such an IGS had been in place, it would have held resources corresponding to roughly 15% to 50% of the total intervention necessary, resulting in a marked reduction of taxpayer-financed state involvement.


Table 0.36: Estimated loss according to MR model assumptions and implied LGD based on estimated EAD
VariableSourceUnitCalc.Value
Total GR Exposure at default (excl. motor) (2007)MR table 2.2m€A9 495
Total GR Gross Premiums Written (excl. motor) (2007)MR table 2.2m€B3 537
Total GR Gross Premiums Written (incl. motor) (2007)MT table 4.3m€C5 141
Company Total Gross Premiums Written (2009)internal communicationm€D694
Market share of CompanycalculationE=D/C13.50%
Share of Motor in Greek marketcalculationF=(C-B)/C31.20%
Share of company without motor on market without motor stays constant due to proportionality assumptions
Share of EAD attributed to company is equal to market share attributed to company
Amount of EAD attributed to companycalculationm€G=A*E1282
Estimated loss using MR model = EAD*LGD(15%)calculationm€H=G*15%192
Estimation of gap by Greek supervisor (incl. motor)internal communicationm€I203.5
Estimation of gap by Greek supervisor (excl. motor)calculationm€J=I*(1-F)140.0
Implied LGD (excl. Motor)calculationK=J/G10.92%


Table 0.37: IGS funding needs for Total Insurance sector under Home state principle for different confidence levels and default probabilities; funding needs in absolute values and as a share of total gross premiums written
α PD = 0.5%PD=0.1%
75%90%99%75%90%99%
GRFunding needs (m€)6.5818.0977.660.963.2519.61
Share of Premiums0.19%0.51%2.20%0.03%0.09%0.55%


Figure 0.8: Position of the losses generated by the Aspis default on the estimated loss distribution function for the total insurance sector in Greece, under home state principle and two different probabilities of default


5. References


CEA, 2009. European Insurance in Figures (2007 Data) dataset. CEA Statistics Home Page. Available at: http://www.cea.eu/index.php?page=statistics [Accessed October 5, 2009].


CEIOPS, 2009a. Annex 2 to: CEIOPS Input to the EC work on Insurance Guarantee Schemes, Brussels: Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS). Available at: http://www.ceiops.eu/media/files/publications/submissionstotheec/CEIOPS-DOC-18-09%20_Input_to_EC_work_on_IGS-approved_clean_.pdf [Accessed October 8, 2009].


CEIOPS, 2009b. CEIOPS Input to the EC work on Insurance Guarantee Schemes, Brussels: Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS). Available at: http://www.ceiops.eu/media/files/publications/submissionstotheec/CEIOPS-DOC-18-09%20_Input_to_EC_work_on_IGS-approved_clean_.pdf [Accessed October 8, 2009].


CEIOPS, 2009c. CEIOPS' Second Bi-annual Financial Stability Report, Brussels: Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS). Available at: http://www.ceiops.eu/media/files/publications/reports/Fin-Stability-autumn-2009/CEIOPS-second-bi-annual-Financial-Stability-Report-2009-20091210.pdf.


CEIOPS, 2008. Statistical Annex 2007 to the report: Financial Conditions and Financial Stability in the European Insurance
and Occupational Pension Fund Sector 2007-2008. Risk Update, Brussels: Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS). Available at: http://www.ceiops.eu/media/files/publications/reports/CEIOPS-Autumn-Report-2008.pdf [Accessed October 8, 2009].


European Central Bank, 2009. Financial Stability Review, Frankfurt am Main, Germany: European Central Bank.


Oxera, 2007. Insurance guarantee Schemes in the EU: comparative analysis of existing schemes, analysis of problems and evaluation of options, Brussels: European Commission DG Internal Market and Services. Available at: http://ec.europa.eu/internal_market/insurance/docs/guarantee_schemes_en.pdf [Accessed July 3, 2009].


Schich, S., 2009. Insurance companies and the financial crisis. OECD Journal: Financial Market Trends, 2009(2).


Standard and Poor's, 2009. 2008 Annual Global Corporate Default Study and Rating Transitions, Standard and Poor's. Available at: http://www.google.co.uk/url?sa=t&source=web&ct=res&cd=1&ved=0CAgQFjAA&url=http%3A%2F%2Fwww.valuation.co.il%2Fdata%2Fwacc%2FSnP-Default_Transition_and_Recovery_2008.pdf&ei=_QmzS7SXMZuosQaxupHhAQ&usg=AFQjCNFPfsXxdf9Z_1B7hQdXzDNNRp_d3g&sig2=Jjj7KrW2n6CVOGBKSOLqzA.








D. ANNEX ON CROSS-BORDER INSURANCE ACTIVITY IN THE EU-EEA


1. IGS Coverage


Table 0.38: Summary of geographic scope of existing IGS.
LifeNon-Life non-motorNon-Life motor
Geographic scopeGeographic scopeGeographic scope
HomeHostHomeHostHomeDomestic
ATx
BEx
BGxx
CYx
CZx
DExx
DKxx
EEx
ESxxx
FIx
FRxxx (1)x
GRx
HUx
IExx
IS
ITx
LIx
LTx
LUx
LVxxxxx
MTxxx
NLx
NOx
PLxxx
PTx
ROxx
SEx
SIx
SKx
UKxxxxx

Notes: (1) only for companies selling mandatory insurance

Source: Methodological report, Table 3.13 - Oxera report (Oxera 2007), CEIOPS updates (CEIOPS 2009), Agreements and Conventions related to the implementation of the 4th Motor Directive (Council of Bureaux 2009).


Table 0.39: Coverage of life insurance by existing national IGS in EU-EEA countries


Note and legend: The table shows the cases in which an IGS covers life insurance activity, taking into consideration the existence of possible cross-border interactions between Member States. Countries listed on rows export insurance services to countries listed on columns.

White cell:no IGS coverage.
Small black dot:exports are covered by an IGS, but domestic insurance activity in the importing country is not covered.
Big black dot:exports are not covered by an IGS, but domestic insurance activity in the importing country is covered.
Black cell:both exports and domestic activity in the importing country are covered by an IGS.

Source: Oxera report (Oxera 2007), CEIOPS updates (CEIOPS 2009), own graphical presentation

Table 0.40: Coverage of non-life insurance by existing national IGS in EU-EEA countries


Note and legend: The table shows the cases in which an IGS covers non-life insurance activity, taking into consideration the existence of possible cross-border interactions between Member States.

Countries listed on rows export insurance services to countries listed on columns.

White cell:no IGS coverage.
Small black dot:exports are covered by an IGS, but domestic insurance activity in the importing country is not covered.
Big black dot:exports are not covered by an IGS, but domestic insurance activity in the importing country is covered.
Black cell:both exports and domestic activity in the importing country are covered by an IGS.

Source: Oxera report (Oxera 2007), CEIOPS updates (CEIOPS 2009), own graphical presentation


Table 0.41: Uneven protection of policyholders within Member States
Importing Member State
Home IGSHost IGSHome and Host IGSNo IGS
Exporting
Member State
Home IGS+
Host IGS-
Home and Host IGS+
No IGS-


Legend: + = The cross border activity is more protected than the domestic one

- = The cross border activity is less protected than the domestic one

Note: Uneven policyholders' protection due to IGS design features other than geographic scope are also possible (not considered in the Table).

Source: own elaboration.

Table 0.42: Life insurance – Cases of uneven protection of policyholders within each Member State.


Legend: + / - = The cross border activity is more / less protected than the domestic one

Source: Oxera report (Oxera 2007), CEIOPS updates (CEIOPS 2009), own graphical presentation

Table 0.43: Non-life insurance – Cases of uneven protection of policyholders within each Member State.


Legend: + / - = The cross border activity is more / less protected than the domestic one

Source: Oxera report (Oxera 2007), CEIOPS updates (CEIOPS 2009), own graphical presentation

Table 0.44: Life insurance - coverage limits and deductions across Member States


Table 0.45: Non-life insurance - coverage limits and deductions across Member States


2. Cross Border Activity


2.1. Exports

Official data by CEIOPS provides a breakdown of export flows of insurance by business line (life, non-life and composite companies) and by channel, in terms of gross premiums written. (CEIOPS 2008)


A breakdown into life and non-life business lines is obtained by attributing premiums of composite companies based on a comparison with CEA data (CEA 2009), which applies a life/non-life classification to data provided voluntarily by members of the national associations of insurers.

Data in non-life insurance is reduced in each country proportionally to the share of motor insurance in order to obtain an estimate of total premiums written in all non-motor sub-lines of non-life insurance.


Table 0.46: Exports of insurance services towards other EEA countries, total insurance sector (m€).
(m €)(as a share of home activity)
Branches onlyBranches plus FPSBranches onlyBranches plus FPS
Calculations32+33/(1+5)(2+3)/(1+5)
AT45.4745.470.35%0.35%
BE1 052.621 409.733.77%5.06%
BG0.000.000.00%0.00%
CY12.0960.092.28%11.32%
CZ7.228.440.22%0.25%
DE1 060.281 496.050.73%1.03%
DK769.61887.454.20%4.85%
EE63.0063.0032.62%32.62%
ES0.000.000.00%0.00%
FI196.00205.284.17%4.36%
FR2 424.692 809.341.30%1.51%
GR0.000.000.00%0.00%
HU0.000.000.00%0.00%
IE4 881.6023 714.4011.78%57.24%
IS0.001.740.00%0.78%
IT281.52808.660.36%1.03%
LI1.552 777.480.06%99.27%
LT0.941.510.29%0.46%
LU851.3510 984.097.67%98.89%
LV7.597.593.38%3.38%
MT1.72196.640.38%43.32%
NL0.000.000.00%0.00%
NO6.24220.560.05%1.81%
PL1.001.000.01%0.01%
PT106.96108.900.93%0.94%
RO0.000.000.00%0.00%
SE0.000.000.00%0.00%
SI0.000.640.00%0.05%
SK2.673.470.23%0.30%
UK0.000.000.00%0.00%
EU11 766.3342 811.761.13%4.10%
EU-EEA11 774.1245 811.541.11%4.32%

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009)


Table 0.47: Exports of insurance services towards other EEA countries, Life business line (m€).
(m €)(as a share of home activity)
Branches onlyBranches plus FPSBranches onlyBranches plus FPS
Calculations32+33/(1+5)(2+3)/(1+5)
AT19.5619.560.27%0.27%
BE283.60498.881.28%2.25%
BG0.000.000.00%0.00%
CY5.0053.001.40%14.83%
CZ4.484.930.22%0.24%
DE158.37399.500.21%0.53%
DK73.2589.700.56%0.68%
EE63.0063.0053.39%53.39%
ES0.000.000.00%0.00%
FI196.00196.007.04%7.04%
FR314.58446.670.23%0.33%
GR0.000.000.00%0.00%
HU0.000.000.00%0.00%
IE3 215.0020 462.008.56%54.47%
IS0.000.000.00%0.00%
IT161.30463.870.26%0.76%
LI0.002 735.170.00%99.26%
LT0.000.000.00%0.00%
LU841.1610 415.008.33%103.19%
LV0.000.000.00%0.00%
MT0.245.470.11%2.55%
NL0.000.000.00%0.00%
NO0.000.000.00%0.00%
PL1.001.000.01%0.01%
PT92.0793.131.00%1.01%
RO0.000.000.00%0.00%
SE0.000.000.00%0.00%
SI0.000.410.00%0.09%
SK0.000.000.00%0.00%
UK0.000.000.00%0.00%
EU5 428.6133 212.110.71%4.37%
EU-EEA5 428.6135 947.280.70%4.66%

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009)


Table 0.48: Exports of insurance services towards other EEA countries, Life business line (m€).
(m €)(as a share of home activity)
Branches onlyBranches plus FPSBranches onlyBranches plus FPS
Calculations32+33/(1+5)(2+3)/(1+5)
AT25.9125.910.44%0.44%
BE769.02910.8513.48%15.96%
BG0.000.000.00%0.00%
CY7.097.094.09%4.09%
CZ2.753.520.21%0.27%
DE901.901 096.551.30%1.58%
DK696.36797.7513.62%15.60%
EE0.000.000.00%0.00%
ES0.000.000.00%0.00%
FI0.009.280.00%0.48%
FR2 110.112 362.674.28%4.79%
GR0.000.000.00%0.00%
HU0.000.000.00%0.00%
IE1 666.603 252.4043.12%84.15%
IS0.001.740.00%0.92%
IT120.22344.790.71%2.03%
LI1.5542.313.66%100.00%
LT0.941.510.77%1.24%
LU10.19569.091.00%56.12%
LV7.597.594.43%4.43%
MT1.48191.180.62%79.67%
NL0.000.000.00%0.00%
NO6.24220.560.27%9.42%
PL0.000.000.00%0.00%
PT14.8815.770.63%0.67%
RO0.000.000.00%0.00%
SE0.000.000.00%0.00%
SI0.000.230.00%0.03%
SK2.673.470.85%1.11%
UK0.000.000.00%0.00%
EU6 337.729 599.652.22%3.36%
EU-EEA6 345.509 864.262.20%3.42%

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009)


Figure 0.9: Total insurance sector, exports of insurance services to other EEA countries, by branches and via Free Provision of Services, in absolute terms (m €).


Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Figure 0.10: Total insurance sector, exports of insurance services to other EEA countries, by branches and via Free Provision of Services, as a share of total home activity.


Note: Countries ordered by exports of insurance services in absolute terms

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Figure 0.11 - Life insurance, exports of insurance services to other EEA countries, by branches and via Free Provision of Services, in absolute terms (m €).


Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Figure 0.12 - Non-life insurance, exports of insurance services to other EEA countries, by branches and via Free Provision of Services, in absolute terms (m €).


Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Table 0.49: Estimates of EAD corresponding to export flows within the EEA (m €)
Estimates of EAD corresponding to exports
LifeNon-LifeTotal
Exports via branchesExports via FPSTotal Life exportsExports via branchesExports via FPSTotal non-life exportsTotal exported via all channels
AT159.40.0159.448.60.048.6208.0
BE2 150.31 632.33 782.52 592.3478.13 070.46 852.9
BG0.00.00.00.00.00.00.0
CY38.0364.8402.814.10.014.1416.9
CZ14.41.415.94.01.15.120.9
DE1 612.12 454.54 066.63 222.9695.53 918.57 985.1
DK655.8147.3803.11 371.9199.71 571.62 374.7
EE271.90.0271.90.00.00.0271.9
ES0.00.00.00.00.00.00.0
FI2 611.90.02 611.90.038.138.12 650.0
FR2 741.01 151.03 892.07 193.9861.18 054.911 946.9
GR0.00.00.00.00.00.00.0
HU0.00.00.00.00.00.00.0
IE12 619.767 698.980 318.55 788.85 508.111 296.991 615.5
IS0.00.00.00.06.06.06.0
IT1 021.61 916.42 938.0230.5430.6661.13 599.1
LI0.020 841.120 841.14.5117.2121.720 962.8
LT0.00.00.01.20.71.91.9
LU6 381.672 633.379 014.935.81 960.91 996.781 011.5
LV0.00.00.08.50.08.58.5
MT1.531.633.03.6465.6469.2502.2
NL0.00.00.00.00.00.00.0
NO0.00.00.020.8714.4735.2735.2
PL2.50.02.50.00.00.02.5
PT403.14.6407.731.51.933.4441.1
RO0.00.00.00.00.00.00.0
SE0.00.00.00.00.00.00.0
SI0.01.91.90.00.40.42.3
SK0.00.00.04.21.35.55.5
UK0.00.00.00.00.00.00.0
EU30 684.8148 037.8178 722.620 503.110 643.231 146.3209 868.9
EEA30 684.8168 878.9199 563.720 528.311 480.832 009.2231 572.9

Source: Table 0 .46-Table 0 .48; Methodological Report, Table 2.2; CEIOPS (CEIOPS 2008)


Table 0.50: Ratios of estimated ‘exported’ EAD to GDP, Total insurance sector, exports via branches and via FPS (GDP in m €)
GDPTotal EAD home/GDPTotal EAD exported/GDP
AT270 782.425.55%0.077%
BE334 948.055.95%2.046%
BG28 898.61.43%0.000%
CY15 951.119.19%2.613%
CZ127 330.56.61%0.016%
DE2 428 200.041.75%0.329%
DK227 024.956.45%1.046%
EE15 626.63.91%1.740%
ES1 052 730.020.42%0.000%
FI179 536.025.06%1.476%
FR1 894 646.071.66%0.631%
GR226 437.04.12%0.000%
HU101 086.55.56%0.000%
IE189 751.284.78%48.282%
IS14 932.35.34%0.040%
IT1 546 177.427.28%0.233%
LI3 363.1627.92%623.317%
LT28 576.62.39%0.007%
LU37 465.8213.87%216.228%
LV21 111.01.30%0.040%
MT5 458.734.47%9.201%
NL568 664.061.36%0.000%
NO283 366.430.80%0.259%
PL311 001.76.61%0.001%
PT163 051.527.78%0.271%
RO124 728.51.14%0.000%
SE331 147.274.05%0.000%
SI34 568.210.12%0.007%
SK54 897.65.09%0.010%
UK2 044 133.0104.57%0.000%
EU12 363 930.052.69%1.698%
EU-EEA12 665 591.852.30%1.829%

Source: Table 0 .49; Eurostat, own elaboration


2.2. Imports


Official data lacks information on imports of insurance services via Freedom of Provision of Services (FPS) within the EEA and also lack information on the origins of flows of insurance services imported via branches. In order to obtain an estimate of the total imports in each country it is therefore be necessary to distribute the exports of each country across all importing countries.

Imports via FPS are attributed proportionally to the size of each country’s insurance market. This is justified by the fact that commerce of services within EU states seem to depend mostly on the relative size of their markets and on their legal and cultural similarities (Walsh 2006; Henk Kox & Arjan Lejour 2005), so that shares dependent on the size of markets should represent an acceptable first approximation of the real flows. The data obtained in this way has been compared with an alternative estimate based on shares of total imports via EEA branches, and estimated were closer than 10% for most countries, with the exception of import flows into Iceland and some of flows of life insurance imported from Ireland by most countries.

In order to obtain an estimate of the origin of flows of imports via branches, total exports via EEA branches are redistributed in all countries proportionally to their shares of total imports via EEA branches as reported in CEIOPS data. Total imports in each country estimated in this way slightly differ from total imports as reported by CEIOPS but differences seem to be contained in the vast majority of cases.

After an estimate of the of the origins of flows of imports is produced, an estimate of all bilateral trade flows in insurance services through the EEA can be obtained by summing these two quantities.

A summary of total estimated imports for each EEA country is presented in Table 0 .51 - Table 0 .53.

Table 0 .54 and Table 0 .55 present the estimated bilateral trade flows of imports among each EEA country.

Table 0.51: Imports of insurance services from other EEA countries, Total insurance sector (m€).
(m €)(as a share of home activity)
Branches onlyBranches plus FPSBranches onlyBranches plus FPS
Calculations66+76/(1+5)(6+7)/(1+5)
AT0.00369.190.00%2.84%
BE609.351 566.232.19%5.62%
BG0.007.990.00%2.26%
CY20.6435.103.89%6.61%
CZ298.99399.558.96%11.97%
DE2 108.866 003.871.46%4.15%
DK0.00593.690.00%3.24%
EE5.058.302.61%4.30%
ES0.001 216.180.00%2.85%
FI0.00131.540.00%2.80%
FR0.006 115.760.00%3.29%
GR32.44149.020.92%4.21%
HU0.0092.350.00%3.38%
IE1 493.232 112.943.60%5.10%
IS0.003.860.00%1.73%
IT3 732.106 431.714.76%8.20%
LI0.00104.120.00%3.72%
LT33.8543.8310.39%13.45%
LU92.40367.160.83%3.31%
LV23.9528.2710.67%12.60%
MT15.7227.473.46%6.05%
NL0.001 658.580.00%2.36%
NO1 806.092 239.9214.83%18.39%
PL0.00302.310.00%3.50%
PT0.00405.780.00%3.51%
RO0.0025.290.00%2.42%
SE0.00630.330.00%3.10%
SI0.0028.710.00%2.30%
SK0.0038.970.00%3.36%
UK2 340.1215 512.190.67%4.41%
EU10 806.7044 302.311.03%4.25%
EU-EEA12 612.7946 650.211.19%4.40%

Note: FPS bilateral flows are estimated based on proportionality assumptions, italics indicates numbers containing estimations

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Table 0.52: Imports of insurance services from other EEA countries, Life business line (m€)
(m €)(as a share of home activity)
Branches onlyBranches plus FPSBranches onlyBranches plus FPS
Calculations66+76/(1+5)(6+7)/(1+5)
AT0.00293.290.00%4.11%
BE115.871 011.320.52%4.56%
BG0.004.940.00%4.10%
CY13.0025.303.64%7.08%
CZ275.24358.8513.53%17.64%
DE925.083 988.401.23%5.31%
DK0.00539.750.00%4.09%
EE0.002.270.00%1.92%
ES0.00966.030.00%4.12%
FI0.00106.590.00%3.83%
FR0.005 555.600.00%4.07%
GR4.98108.110.20%4.32%
HU0.0083.070.00%4.12%
IE1 025.001 629.842.73%4.34%
IS0.001.400.00%4.09%
IT2 967.005 461.434.83%8.89%
LI0.00103.630.00%3.76%
LT28.1236.5213.79%17.90%
LU29.68293.670.29%2.91%
LV23.0025.1843.40%47.51%
MT3.0811.891.44%5.56%
NL0.001 088.850.00%4.12%
NO106.00511.201.08%5.20%
PL0.00277.680.00%4.12%
PT0.00375.280.00%4.08%
RO0.0017.090.00%4.11%
SE0.00534.810.00%4.12%
SI0.0018.250.00%4.12%
SK0.0034.930.00%4.12%
UK9.7812 579.310.00%4.12%
EU5 419.8335 428.250.71%4.67%
EU-EEA5 525.8336 044.480.72%4.67%

Note: FPS bilateral flows are estimated based on proportionality assumptions, italics indicates numbers containing estimations

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Table 0.53: Imports of insurance services from other EEA countries, Non-life business line (m€)
(m €)(as a share of home activity)
Branches onlyBranches plus FPSBranches onlyBranches plus FPS
Calculations66+76/(1+5)(6+7)/(1+5)
AT0.0075.900.00%1.30%
BE493.48554.918.65%9.72%
BG0.003.050.00%1.31%
CY7.649.804.41%5.66%
CZ23.7540.701.82%3.12%
DE1 183.782 015.471.70%2.90%
DK0.0053.940.00%1.05%
EE5.056.036.72%8.02%
ES0.00250.150.00%1.30%
FI0.0024.950.00%1.30%
FR0.00560.160.00%1.14%
GR27.4640.912.66%3.96%
HU0.009.280.00%1.30%
IE468.23483.1012.11%12.50%
IS0.002.460.00%1.30%
IT765.10970.284.50%5.70%
LI0.000.490.00%1.16%
LT5.737.314.71%6.00%
LU62.7273.496.19%7.25%
LV0.953.090.55%1.80%
MT12.6415.585.27%6.49%
NL0.00569.730.00%1.30%
NO1 700.091 728.7272.63%73.85%
PL0.0024.630.00%1.30%
PT0.0030.500.00%1.29%
RO0.008.200.00%1.30%
SE0.0095.520.00%1.30%
SI0.0010.460.00%1.30%
SK0.004.040.00%1.29%
UK2 330.342 932.885.04%6.34%
EU5 386.878 874.061.88%3.10%
EU-EEA7 086.9610 605.732.46%3.68%

Note: FPS bilateral flows are estimated based on proportionality assumptions, italics indicates numbers containing estimations

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Figure 0.13: Total insurance sector, imports of insurance services from other EEA countries, by branches and via Free Provision of Services, in absolute terms (m €).


Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Figure 0.14: Total insurance sector, imports of insurance services from other EEA countries, by branches and via Free Provision of Services, as a share of home activity.


Note: Countries ordered by imports of insurance services in absolute terms

Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Table 0.54: Approximate estimate of bilateral flows of trade within the EEA based on proportionality assumption, Life business line (m €)


Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


Table 0.55: Approximate estimate of bilateral flows of trade within the EEA based on proportionality assumption, Non-Life business line (m €)


Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


2.3. Trade openness


Based on the export and estimated import data presented in previous sections it is possible to calculate an index of openness to trade in the insurance sector for all EEA countries. The trade openness index, a standard indicator used in international trade economics, is defined as the share of imports plus exports over total production within any given sector.

Here, as we are interested in openness towards EU/EEA members, imports and exports considered are only those to and from other EU/EEA countries.


Figure 0.15: Trade openness index towards trade with other EEA members for all EEA Member States and overall for EU and EEA.


Source: CEIOPS (CEIOPS 2008) and CEA (CEA 2009), own elaboration


3. References


CEA, 2009. European Insurance in Figures (2007 Data) dataset. CEA Statistics Home Page. Available at: http://www.cea.eu/index.php?page=statistics [Accessed October 5, 2009].


CEIOPS, 2009. CEIOPS Input to the EC work on Insurance Guarantee Schemes, Brussels: Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS). Available at: http://www.ceiops.eu/media/files/publications/submissionstotheec/CEIOPS-DOC-18-09%20_Input_to_EC_work_on_IGS-approved_clean_.pdf [Accessed October 8, 2009].


CEIOPS, 2008. Statistical Annex 2007 to the report: Financial Conditions and Financial Stability in the European Insurance
and Occupational Pension Fund Sector 2007-2008. Risk Update, Brussels: Committee of European Insurance and Occupational Pensions Supervisors (CEIOPS). Available at: http://www.ceiops.eu/media/files/publications/reports/CEIOPS-Autumn-Report-2008.pdf [Accessed October 8, 2009].


Council of Bureaux, 2009. 4th Motor Insurance Directive Home Site. Available at: http://www.4directive.org/ [Accessed April 26, 2010].


Henk Kox & Arjan Lejour, 2005. Regulatory heterogeneity as obstacle for international services trade, CPB Netherlands Bureau for Economic Policy Analysis. Available at: http://ideas.repec.org/p/cpb/discus/49.html [Accessed April 27, 2010].


Oxera, 2007. Insurance guarantee schemes in the EU: comparative analysis of existing schemes, analysis of problems and evaluation of options, Brussels: European Commission DG Internal Market and Services. Available at: http://ec.europa.eu/internal_market/insurance/docs/guarantee_schemes_en.pdf [Accessed July 3, 2009].


Walsh, K., 2006. Trade in Services: Does Gravity Hold? A Gravity Model Approach to Estimating Barriers to Services Trade, Dublin, Ireland: Institute for International Integration Studies.


1 Directive 2009/138/EC of the European Parliament and of the Council of 25 November 2009 on the taking-up and pursuit of the business of Insurance and Reinsurance (recast), OJ L 335, 17.12.2009, pp. 1-155.

2 See Article 10 of Directive 2001/17/EC of the European Parliament and of the Council of 19 March 2001 on the reorganisation and winding-up of insurance undertakings, OJ L110, 20.04.2001, pp. 28-39.

3 All calculations in the IGS IA Methodology Report exclude motor insurance, as it is covered by its own separate arrangements.

4 Although the model employed in the IGS IA Methodology Report is not meant to be used for making predictions on individual default cases, the magnitudes in this real-life case appear to be in line with the assumptions used there:

- the amount of the expected losses implied by the assumption of the model used in the IGS IA Methodology Report, in case of a default of the size of Aspis (13% of the market, proportionally reduced to exclude motor insurance) is 192 Eur million, as illustrated in Table 0 .36, while the regulator seems to expect a loss of around 140 Eur million.

- reversing the argument: the Loss Given Default (LGD) implied by a final loss of 140 Eur million (excluding Motor) and by an estimation of the exposure at default based on proportionality assumptions is 10.92% (see Table 0 .36), while the loss given default assumed in the methodology report model is 15%.

5 A loss of 3.96% of gross premiums written in the domestic market (excluding Motor).

6 This probability is calculated based on the assumption that individual insurers have a default probability of 0.5% which is the maximum target probability of default under Solvency II. Under the assumptions that the probability of default of a single insurer is 0.1% the probability of such an event should be lower that 0.05% or an α of 99.95%.

i Because information is available for the entire EU-EEA area, this Impact Assessment (IA) has been developed not only for the 27 EU Member States but also for EEA countries: Norway, Iceland and Lichtenstein. In order to enhance readability, the term 'Member State' is used in this Impact Assessment as a synonym of 'country' and therefore it might also refer to EEA countries.

ii In the context of this IA consumers include policyholders, beneficiaries and, in the case of non-life insurance, third parties who may seek compensation. Therefore, the terms 'consumer' and 'policyholder/beneficiary' will be used interchangeably.

iii "Recommendation 5: The Group considers that the Solvency 2 Directive must be adopted and include a balanced group support regime, coupled with sufficient safeguards for host Member States, a binding mediation process between supervisors and the setting-up of harmonised insurance guarantee schemes.”, http://ec.europa.eu/internal_market/finances/docs/de_larosiere_report_en.pdf

iv Directive 1994/19/EC as amended by Directive 2009/14/EC on Deposit Guarantee Schemes and Directive 1997/9/EC on Investor Compensation Schemes.

v The various types of guarantee schemes, while all providing a certain level of consumer protection, have in part different objectives.

DGS are designed to compensate depositors for deposits at banks up to a specified limit if the bank is not in a position to repay them. The objective is twofold: from the consumer protection perspective, a part of the depositors' wealth is protected from losses due to bank failures; from a financial stability perspective, the confidence that deposits are protected reduces the likelihood of bank runs and thus contributes to preserving the stability of the financial system.

The ICS Directive applies instead to investment firms (including credit institutions) who provide investment services under the MiFID Directive (Directive 2004/39/EC on Markets in Financial Instruments). The ICS Directive provides for clients of investment firms to be compensated in two situations. Firstly, if an investment firm is unable to repay money owed or belonging to a client and held on his behalf in connection with investment services. Secondly, if an investment firm is unable to render a financial instrument belonging to the client and held, administered or managed on the client's behalf. However, the Directive does not cover reductions in the value of the investments, i.e. if the value of the investments' underlying assets decline, the value of the market declines or if an issuer or fund fails.

vi Nonetheless, reinsurance can produce contagion effects when insurance undertakings default. In order to analyse contagion, it is however necessary to dispose of firm level data which is not the available to the Commission at the moment. In the Oxera (2007) report it can be read how reinsurance policies are typically outside the scope of IGS protection.

vii The proceedings of the activities of the Working Group can be consulted on

http://ec.europa.eu/internal_market/insurance/guarantee_en.htm#docs

viii Oxera (2007), http://ec.europa.eu/internal_market/insurance/docs/guarantee_schemes_en.pdf.

The information presented in the Oxera report has recently been updated by CEIOPS (2009a),

http://www.ceiops.eu/media/files/publications/submissionstotheec/annex_2.pdf

ix CEIOPS is the Committee of European Insurance and Occupational Pensions Supervisors. It is composed of high level representatives from the insurance and occupational pensions supervisory authorities of EU Member States. The authorities of EEA countries also participate in CEIOPS. CEIOPS' website is: http://www.ceiops.org

x CEIOPS (2009b), http://www.ceiops.eu/media/files/publications/submissionstotheec/CEIOPS-DOC-18-09%20_Input_to_EC_work_on_IGS-approved_clean_.pdf

xi http://ec.europa.eu/internal_market/insurance/guarantee_en.htm#cons.

xii http://ec.europa.eu/internal_market/insurance/docs/guarantee/summary_en.pdf.

xiii http://ec.europa.eu/internal_market/insurance/guarantee_en.htm#whitepaper.

xiv http://www.europarl.europa.eu/comparl/tempcom/equi/default_en.htm.

xv Directive 2009/138/EC of the European Parliament and of the Council of 25 November 2009 on the taking-up and pursuit of the business of Insurance and Reinsurance (Solvency II).

http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:335:0001:0155:EN:PDF

xvi In financial mathematics and risk management, the most common quantitative measure of the risk of loss on a specific portfolio of assets is the so-called Value at Risk (VaR).

xvii About the absence of market discipline in the insurance market, see for example Eling M., Schmit J.T. (2008). Also, in Yasui T. (2001) one can read: 'The financial and managerial situation of insurance companies is much more technical and complex than that of ordinary companies. Non-professional policyholders can hardly be expected to verify the credibility of an insurance company sufficiently' and 'non-professional policyholders not only have limited ability to evaluate appropriately the financial soundness of insurance companies, but also they have little incentive to do so: because of the technical and complex nature of the financial situation of insurance companies, the cost of gathering sufficient information to make a wise decision is significantly high'.

xviii For a definition, see for example http://en.wikipedia.org/wiki/Risk_aversion.

xix For an in-depth complementary analysis of the risks faced by insurance undertakings, please see sub-section 4.1 of the Oxera report.

xx Generally speaking, when life insurance contracts are non-unit linked, investment/market risk is normally borne by the insurance undertaking. On the contrary, when life insurance contracts are unit-linked, investment/market risk is normally borne by policyholders. Looking at the split in life insurance business between unit-linked and non unit-linked activity in Member States, it appears that non-unit linked life insurance reserves are much higher (60-90% of total) than those of unit-linked ones. However, distinctions are in reality very difficult as in both unit-linked and non-unit linked products investment risk is de facto shared between insurers and policyholders. In the unit-linked sector, in fact, there are many insurance undertakings that offer guarantees to policyholders. They take a wide variety of forms including minimum returns, fixed annuity rates as well as contractual terms such as early or regular withdrawal of funds on terms that give policyholders valuable options. Thus, in these cases the insurance undertaking bears some of the market/investment risk and clear-cut distinctions are difficult to draw.

xxi The recent financial crisis has also shown the possibility of (loss) contagion from banks to insurers. Insurers tend to be, in fact, highly exposed to counterparty risk towards banks as they usually buy interest rate and equity derivatives from banks to hedge their market risk exposures.

Losses to life insurers can also derive from changes in policyholders’ behaviour. The recent financial crisis has shown how households (especially American) have, at a certain moment, accelerated the redemption of their investments in mutual funds. The same can in theory happen also with life insurers, although high redemption costs generally tend to discourage policyholders from doing so.

Losses to life and non-life insurers can also derive from a contagion effect when they belong to a financial conglomerate. A spreading of the lack of confidence could in fact be a reasonable reaction in case of exposure of both banks and insurances belonging to the same financial conglomerate to common management failings such as those stemming from high risk investment strategies and/or fraud.

xxii See sub-section 4.1.2 of the Oxera report for more details on these defaults and reference to various sources of information on failures of insurance undertakings. Besides, the Financial Services Compensation Scheme established in 2001 in the UK reports to have to date dealt with 30 insolvent insurance undertakings (28 non-life and 2 life insurers).

xxiii Once Solvency II will enter into force in 2012, it is expected to maintain the PD of EU insurance undertakings to 0.5% or less.

xxiv On the methodology used to estimate the EAD, see in the Methodological report (MR - Annex D to this IA) section Annex A2.

xxv For further details, see section 4 of Annex E: Supplementary tables to the methodological report

xxvi Error: Reference source not found is not a complete inventory of past failures in the EU. For a more detailed analysis of the losses generated by these and other cases of default of European insurers, see sub-section 4.5 of the Oxera report.

xxvii Both calculations are under the assumption of a 15% Loss Given Default (LGD) rate estimate. For an explanation of the reasons supporting the choice of LGD=15%, see in the MR sub-section A3.8.

xxviii In a skewed distribution such as the one of losses incurred by insurers, the average (expected value) is not, generally speaking, a statistics that should be considered as a correct indicator of risk.

xxix The main reason supporting the choice of a Vasicek model has been the very limited amount of information available to feed in the model. A Vasicek model is also used, for example, in the derivation of FIRB capital requirements under Basel II. For more details on the Vasicek model, see Annex A1 in the MR. On considerations specifically related to the appropriateness of the Vasicek model for estimating policyholders losses, see Annex A3 in the MR.

xxx See sub-section 3.4 of the MR. The very exceptional consequences of the recent default of the Aspis Pronia group in Greece are also compatible with the estimated loss distributions for Greece. They correspond in fact to the loss estimated with a PD = 0.5% at a confidence level slightly above 99.5% (see Error: Reference source not found).

xxxi If one also considers the diversification effect produced by the less granular nature of the entire EU market compared to national Member State markets, losses can be considered to be lower of some 20% compared to those indicated (see sub-section 4.5 of the MR). It should also be remembered that the Vasicek model is a single factor model and that it does not allow introducing differences across countries of the correlation between insurances.

xxxii A thorough analysis of the consequences for individual consumers requires detailed information on the distribution of individual policyholders' claims. The Oxera report only presents as an example the distribution of claims for limited parts of the German non-life insurance sector (Table 4.9, page 77). And a recent survey with national insurance associations has failed to provide the Commission with the necessary data on the distribution of individual policyholders' claims. As information on the distribution of individual policyholders' claims is therefore not available at the moment to the Commission, this analysis has not been possible in the IA.

xxxiii Finally, six countries have (only or also) special schemes that cover very specific classes of non-life insurance (BE, FI, DE, IT, PL and ES) For further details, see CEIOPS (2009a).

xxxiv The situation for the non-life motor insurance sector is, as shown in Error: Reference source not found, completely different, with almost the entire EU-EEA area covered by an IGS. A guarantee scheme for motor insurance is required, in fact, in every Member State by Directive 84/5/EEC (now recast in Directive 2009/103), even though only for the case of uninsured vehicles. Member States have nonetheless voluntarily extended over time to the case of defaulted insurance undertakings their already compulsory guarantee schemes for motor insurance. In conclusion, as IGS are today already present almost in every EU-EEA country and do not create substantial loopholes in the protection of policyholders, there is no apparent necessity to intervene at the EU level. For this reason, this IA focuses only on life insurance and non-life insurance (excluding motor). For further information on guarantee schemes for motor insurance, see http://www.4directive.org.

xxxv Endnote Error: Reference source not found applies. Furthermore, amounts for life and non-life are estimated proportionally to losses gross of IGS protection due to the difficulty to split IGS available resources when current IGS cover both life and non-life insurance.

xxxvi The statement does not take into consideration the possibility of an ex-post State intervention. It will be however shown in section 6 that this alternative is in general not preferable compared to setting-up an IGS. This is also proven by the fact that when Member States have experienced major defaults, they have, in general terms, preferred to introduce an IGS instead of keeping the existing situation (its is the case, for example, of DE and UK).

The statement does also not take into consideration the possibility that consumers are protected by means of a preferential treatment for consumers in liquidation procedures. It is however shown in section 6 that preferential treatment in liquidation is in general a less effective mean for protecting consumers than the setting-up of an IGS. It should also be noticed that some Member States that have experienced important insurance defaults have preferred to react setting up an IGS instead of immediately reinforcing or introducing a preferential treatment for consumers in liquidation procedures (it is the case of DE and UK).

xxxvii “Under the home state principle, the IGS covers policies issued by domestic insurers as well as by the branches of domestic insurers established in other EU-EEA Member States. In contrast, under the host state principle, the policies issued by branches of incoming EU-EEA insurers are covered by the local IGS.” (Oxera report, footnote 8). In compliance with this definition, in this IA policies sold cross-border under free provision of services are considered to be, as a general principle, covered in both cases of a home state principle based IGS and of a host principle based IGS by the IGS of the Member States where the insurance undertaking is authorised/established.

xxxviii See also Table 3.13 of the MR, section 2 of the Oxera report and CEIOPS (2009a).

xxxix For example, differences in the nominal amount covered by two IGS do not necessarily mean that policyholders are unevenly protected. A lower compensation limit in a less wealthy country might in fact provide higher relative protection than a higher compensation limit in a richer country.

xl For a presentation of the precise set of IGS design features that it has been possible to consider in the analysis, please refer to Table 3.13 in the MR. Some of the IGS design features not taken into account in the IA have anyhow been partially analysed in the MR. See sub-section 3.4 of the MR.

xli Cross-border activity means insurance services sold via Free Provision of Services (FPS) and through branches. Selling through subsidiaries does not enter into the definition of cross-border activity of this IA as sufficiently complete data on insurance groups is not available to the Commission at this moment.

xlii See Davies, J. Paul: Aviva to revamp European operations, FT.com, 21.10.2009,

www.ft.com/cms/s/0/419da410-be73-11de-b4ab-00144feab49a.html

xliii The estimates for each Member State indicate the losses that can be exported from insurers authorized in that Member State and that default. Estimated losses are to be considered as fully taking into account the effects of the existence of minimum solvency requirements for insurance undertakings in the EU (Solvency II in particular). See Tables 5, 7 and 9 of Annex5 to MR and Table 4.19 of the MR for further details on losses as share of premiums and for estimates of losses for life and non-life insurers only.

xliv Amounts for life and non-life are estimated proportionally to losses gross of IGS protection due to the difficulty to split IGS available resources when current IGS cover both life and non-life.

xlv In this calculation, domestic policyholders are considered to be possibly protected both by home and host principle based IGS. Amounts for life and non-life are estimated proportionally to losses gross of IGS protection due to the difficulty to split IGS available resources when current IGS cover both life and non-life.

xlvi For all other non-life insurance policyholders who represent the great majority, instead, the consequence of the failure of a non-life insurance undertaking is generally limited to the amount of prepaid but still not used premiums. Additional costs may however well arise from the need to arrange for replacement cover which may be difficult to obtain and may take time. The consequences of the failure may also affect third parties, as is the case for classes of liability insurance. The failure of an insurer may lead to the non-payment of claims for those policies, which will leave the injuring party exposed to the liability and the injured party without compensation. For further analysis, please refer to pages 74 to 77 of the Oxera report.

xlvii However, Yasui T. (2001) also correctly remarks that: 'it should be noted that the risk of bankruptcy contagion is likely to be smaller for the insurance sector'. The reason behind this is that bank deposits can be withdrawn in basically full amounts. In contrast, policyholders normally incur (heavy) losses due to cancellation deductions, so that policyholders can be expected to think twice before terminating their insurance contract. On how the opacity of the insurance industry tends to cause firm-specific information to spill over to the entire industry and result in an industry wide effect, see also Akhigbe A., Madura J. (2001).

xlviii For a general explanation of how insurance activity can foster economic growth, see CEA (2006). Furthermore, economic theory has also shown that negative shocks, and more in general uncertainty, can reduce growth in the absence of complete insurance markets. See for example Hansen G., Imrohoglu A. (1992).

xlix Lack of insurance cover may be particularly disruptive for those businesses that, in order to operate, have a legal obligation to be insured, as in the case of construction. See for example the case of HIH which failed in Australia and which had serious consequences on the construction activity in that country, presented in Impavido G., Tower I. (2009).

l There is, for example, evidence from the equity markets fall in 2001–03 that life insurers contributed to a downward spiral in markets when limited equity disposals by major insurers seeking to bolster balance sheets led to further declines in the market, requiring further disposals to prevent solvency margins from coming under pressure. In the current crisis, sales of equities and other instruments have been even more widespread.

The problem of insurers causing a downward spiral in financial markets is especially prone to show up when insurance undertakings undergo liquidity problems and therefore need to sell high volumes of assets on the financial markets. In general terms, insurers structurally have low exposure to liquidity risk because they are premium funded and not funded from wholesale money markets (as banks are). Furthermore, liquidity risk is also limited because claims are usually paid when a specified triggering event takes place rather than on demand, and because insurance undertakings' assets are predominantly marketable. There are, however, not negligible sources of liquidity risk that can come from:

- collateral calls in derivative business and securities lending (an issue at AIG in the recent financial crisis, for instance);
- market loss of liquidity on the trading of certain assets (which affected, for example, insurance undertakings with investments in Asset Backed Securities during the crisis);
- rising claims: it is possible - as experienced briefly by minor parts of the AIG group immediately after its rescue - that claims temporarily overwhelm available liquidity in what would be equivalent to a bank run.

In Yasui T. (2001) one can read: "A run could put … insurers in a serious liquidity crisis and possibly force them to go bankrupt." However it also notices that 'repayments of insurance products are usually made less quickly than bank deposits. Insurance companies should have more time to build liquidity for repayments so as to meet their obligations.'

li For further details, see Atkeson A. and Lucas R. E. Jr (1995).

lii For further details, see Sandmo A. (1998), and Varian H. (1980).

liii It must be recalled that it is very difficult (and at the moment there has not been any possibility) to provide direct evidence of any significant distortion in the competition in Member States on the basis of these considerations. This also because there are other factors, such as - for example - taxation, that certainly also have a very important impact on the price and demand for insurance services.

liv For more details on the life and health guarantee system in the USA, see also http://www.nolhga.com/; and for more details on the non-life guarantee system in the USA, see also http://www.ncigf.org/

lv For more information see Yasui T. (2001).

lvi See OECD (1999).

lvii See sub-section 3.2.2 for the various combinations of geographic scope (home and host state) that can produce a lack of policyholder protection in Member States.

lviii This statement is true for society. It is however not necessarily true for single categories of stakeholders, such as policyholders, taxpayers, insurers, etc. Section 7 on expected economic and social impacts will take account of this and consider costs for individual categories of stakeholders.

lix It follows that it is conceptually wrong to argue (at the level of the entire society) that if one creates a protection mechanism with a financial endowment able to absorb insurers' default losses they don't take place any more at the cost of establishing the protection mechanism. It is, in the same way, conceptually wrong to say that an implemented protection mechanism costs to society the amount of money given to its financial endowment.

lx For further details on this theoretical framework, see for example Smith W.T. (1996).

lxi Insurance undertakings' loss distributions tend to be skewed to the left, with very frequent small losses and very rare but also very high losses.

lxii It should be noted that a transparency policy option can also be envisaged in combination with other policy options examined under this and/or the following sub-sections. In those cases the policy option analysis will, by and large, appear as a combination of the various elements analysed under its components.

lxiii It is important to stress that not all losses suffered by insurers will hit the guarantee schemes, but only the part of the losses that exceeds the solvency requirements in place (plus excess capital, if any). The relevant concept of losses for IGS is therefore that of residual losses 'downstream' of insurers’ solvency requirements (plus excess capital, if any).

lxiv See Yasui T. (2001).

lxv How to reach and maintain the target fund is instead an issue related to IGS contributions.

lxvi On the concept of percentile, see also sub-section 2.2.3.

lxvii When the financial endowments are, for example, sufficient to cover the IGS loss distribution up to the, for example, 90th percentile, this by and large means that the level of security chosen avoids that losses are passed on to consumers in 90% of the cases possible. In other terms, it can (by and large) also be said that if the financial endowments cover the IGS loss distribution up to the 75th ,90th , 99th percentile, the IGS is expected to have not enough resources and therefore pass losses onto consumers only every 4, 10, 100 times that an insurer fails.

lxviii For the details of the analysis performed to identify this range of security levels, see sub-section 3.4 in the MR.

lxix A more precise analysis of whether these IGS funding needs have to be considered as minimum values (minimum harmonisation) or exact values (maximum values) is not necessary at this stage where the Commission is keeping the various options open for discussion with stakeholders (see end of this sub-section).

lxx Also the Oxera report notes that 'IGS can best deal with failures that do not involve potential losses that are large relative to the size of the market' and that 'large failures may need to be dealt with though other mechanisms'.

lxxi The list of IGS design features indicated under option 4.2 is not the only possible harmonization perimeter under a minimum scope of harmonization approach. It can, for example, also be envisaged that some of the design features listed under option 4.3.might be added under option 4.2 in the follow-up measures.

lxxii Annex C presents some (even more) preliminary analysis of each of the items from the maximum scope of harmonisation list of option 4.3. The analysis in Annex C is mainly focused on the need to harmonise or not harmonise at the EU level each of the design features. Its main purpose is to foster feedback from stakeholders as to whether the list of design features under option 4.2 should (or not) be enlarged and in which way. At the follow-up measures impact assessment stage, therefore, options related to these design features will be also analysed more in depth, and also thanks to comments and feedback received from stakeholders some of them might become as well the content of a EU legislative binding action on IGS. It might be the case, for example, for contributions to IGS and/or advertisement/information requirements.

lxxiii This means that for a home insurance premium of 500 EUR a year, there would be – for example for 10 years - a price increase of some 60 EUR-cents per year. In case of a life insurance premium of 1000 EUR per year, the price would go up for 10 years by some 1.24 EUR.

lxxiv The explicit request of authorisation for third-country branches is for life-insurer in Article 51 of Directive 2002/83/EC and for non-life insurers in Article 29a of Directive 73/239/EEC. The only exception is for Swiss non-life insurers as a specific agreement regulates their possibility to freely provide services cross-border in EU Member States (see http://ec.europa.eu/internal_market/insurance/solvency/index_en.htm)

lxxv It is important to draw the reader’s attention to the following. The various IGS design features are analysed in Annex B one at a time (the analysis on the level of IGS centralisation is for example presented before the analysis on geographic scope) in order to highlight the consequences of moving from one to another option available for each IGS design feature. Nonetheless, funding needs are always necessarily computed for a collection (vector) of options covering all design features and applying at the same time.

The particular vector of options used as a benchmark to compute funding needs in the IA can be found in Table 4.1 of the MR. In particular, funding needs have been computed for the home state principle option of the geographic scope design feature (unless differently indicated). This implies that, in general, comparisons between options are always made in this IA on the basis of the home state principle option, which constitutes the benchmark for comparisons. Presented funding needs in Annex B are to be intended gross of existing financial endowments where national IGS already exist. Net funding needs can be computed simply subtracting available funds in Member States that have IGS in place. Estimated available funds are presented in Error: Reference source not found. See also sub-section 3.3 of the MR.

lxxviThe decreased/increased funding needs compared to the home state national IGS case depend on the decreased concentration ()of the insurance market considered for the calculations: in fact, as the market share of each market participant is smaller in the EU than in each national market, the concentration of the insurance market decreases. This lower concentration entails a reduction in the probability of extreme losses and a higher probability of medium-high losses (see in the figure below how the loss distribution changes progressively from 0.1 to =0). The final effect on estimated losses depends on which of the two effects prevails. Consider for example the case of PD=0.1% and a 90th percentile of the IGS loss distribution. With this low PD, the effect of a reduction in the probability of extreme losses (a thinner tail, meaning smaller funding needs) is less important than the effect of a higher probability of medium-high losses (a fatter shoulder, meaning higher funding needs) in the portion of the loss distribution around the 90% percentile. It follows that funding needs increase. If funding needs are instead considered for the same PD (0.1%) but at the 99th percentile or at the same 90th percentile but for a higher PD (0.5%), the thinner tail effect prevails, with a reduction in funding needs.


lxxvii For a full set of figures please refer to sub-section 4.3 and Annex A5.2 of the MR.

lxxviii For a full set of figures please refer to sub-section 4.4 and Annex A5.3 of the MR.

lxxix This positive feature of ex-ante funded IGS can be reinforced by introducing ex-ante levies that are weighted by the risk of failure of the contributing insurance undertaking.

EN EN