The asset portfolio of Dutch insurance
companies and Solvency II
University of Groningen, Faculty of Economics and Business
Abstract: With Solvency II
framework especially for insurance companies is coming into force. The requirements and capital charges , which are determined by the diverse modules of the calculation, influence likely the investment behavior of insurance compani
diversification and asset
estimate, the asset portfolio is a key to lower the required capital surplus. In this research is studied if there is a relationship between Solven
relatively risky investments (shares) to safer assets (bonds and other fixed
securities) as an anticipation on the upcoming regulation, also taking the current funding ratio into account. This is examined for Dutch insuran
insurers as well as property casualty insurers based on the data from the annual reports 2009 and the Nederlandsche Bank.
Keywords: Solvency II, asset portfolio, regulation, ex ante preparation, investment risk, property casualty insurance company, life insurance company, funding ratio
The asset portfolio of Dutch insurance
companies and Solvency II
Marjanne Vos
University of Groningen, Faculty of Economics and Business
Master thesis Accounting
Version: 1 August 2011
With Solvency II becoming effective on January 1st 2013, a tightened regulation framework especially for insurance companies is coming into force. The requirements and capital charges , which are determined by the diverse modules of the calculation, influence likely the investment behavior of insurance compani
diversification and asset-liability management are becoming part of the overall estimate, the asset portfolio is a key to lower the required capital surplus. In this research is studied if there is a relationship between Solvency II and the shift from relatively risky investments (shares) to safer assets (bonds and other fixed
securities) as an anticipation on the upcoming regulation, also taking the current funding ratio into account. This is examined for Dutch insuran
insurers as well as property casualty insurers based on the data from the annual reports 2009 and the Nederlandsche Bank.
Solvency II, asset portfolio, regulation, ex ante preparation, investment risk, property ualty insurance company, life insurance company, funding ratio
The asset portfolio of Dutch insurance
companies and Solvency II
University of Groningen, Faculty of Economics and Business
, a tightened regulation framework especially for insurance companies is coming into force. The requirements and capital charges , which are determined by the diverse modules of the calculation, influence likely the investment behavior of insurance companies. Since market risk, liability management are becoming part of the overall-risk estimate, the asset portfolio is a key to lower the required capital surplus. In this cy II and the shift from relatively risky investments (shares) to safer assets (bonds and other fixed-interest securities) as an anticipation on the upcoming regulation, also taking the current funding ratio into account. This is examined for Dutch insurance companies, for life insurers as well as property casualty insurers based on the data from the annual reports
Solvency II, asset portfolio, regulation, ex ante preparation, investment risk, property ualty insurance company, life insurance company, funding ratio
The asset portfolio of Dutch insurance
companies and Solvency II
Author Marjanne Vos
Student number S1972065
Email mjanne.vos@gmail.com
University University of Groningen
Faculty Faculty of Economics and Business Master Master of Science in Accountancy Supervisor University of Groningen Prof. Drs. Dirk Swagerman Supervisor Ernst & Young Drs. John van Teijen RA
Preface
It is remarkable how in the last three years the worldwide financial markets have acted and how the occurrence of the last financial crisis and the impact of it has hit banks, insurers and pension funds. No wonder that regulators all feel obligated to protect companies from the volatility of the market by posing new requirements to financial strength and solvency.
The case of the insurance industry is not very different from the banking sector. As of January 1st 2013 insurance companies in the European Union have to comply with the new regulation framework of Solvency II. The influence on the control environment of an insurer is significant, since it concerns not only a total balance sheet approach for calculation of the solvency requirement but also governance and disclosure requirements.
Actually this research is inspired by the ideas of John van Teijen, an expert in auditing insurance companies, who suggested that the Solvency II framework in relation with asset allocation could be a interesting subject. After a fast start with this plan and plenty of time for developing the framework of this research at Ernst & Young – for which I am very thankful – the tempo of the statistical part and literature was considerably lower.
I would like to thank in the first place the two supervisors, Dirk Swagerman and John van Teijen, for their patience, their comments and suggestions and feedback rounds. I am also very thankful for Ernst & Young, giving me the opportunity to start with an internship for my thesis and offering me a part-time job at the FSO department. Furthermore I would like to thank my parents for remembering I still have to finish my thesis, although I am still in Lyon (France), and for always being supportive in the choices I make.
This thesis forms the end of my Master of Science in Accountancy, but also leads the start of the Executive Master in Accountancy. I have enjoyed writing this final thesis and I hope you all enjoy reading it.
Lyon (France), 5 August 2011 Marjanne Vos
Contents
Preface 3
Contents 4
1. Introduction 5
1.1 Relevance 5
1.2 Problem analysis and research questions 6
1.3 Conceptual framework 7
1.4 Structure 8
2. Institutional framework (Solvency II directives) 9
2.1 Solvency II 9
2.2 Solvency Capital Requirement under Solvency II 10
2.3 Vision on Solvency II 11
3. Literature 12
3.1 Portfolio Theory by Markowitz 12
3.2 Aspects of solvency, capital requirements and regulation 12 3.3 Asset allocation, portfolio management and diversification 14
4. Research design and methodology 16
4.1 Research questions and hypotheses 16
4.2 Design and methodology 16
4.3 Data selection / collection 19
5. Indications of early anticipation of insurers other than in asset management 20 5.1 Participation in the development of the framework 20 5.2 The presence of an internal team for Solvency II issues 20
5.3 Designing and revising of the ORSA 21
5.4 Solvency II and the risk management function 21 5.5 Solvency II and the asset management function 21
5.6 Remarks 22
6. Results 23
6.1 Descriptive statistics of the sample 23
6.2 Statistical results t-test 24
6.3 Statistical results regression analysis and Pearson correlation 25
6.4 The influence of the funding ratio 26
7. Conclusions and recommendations 28
7.1 Institutional framework of Solvency II 28
7.2 Other indications of ex ante preparation on Solvency II 28 7.3 The influence of Solvency II on the asset portfolio 2009 29
7.4 Overall conclusion and recommendations 30
8. Limitations and future research 31
8.1 Limitations of the research and consequences 31
8.2 Possibilities for future research 31
9. Literature 32
1
Introduction
After the financial crisis, regulation and more effective supervision on companies operating on the financial markets is a hot issue. All possible measurers are taken to protect enterprises from the volatility of the financial markets and stock market crashes. With the introduction of Solvency II from January 1st 2013 onwards, a tightened regulation framework especially drafted for insurers is coming into force. The current Solvency I framework has been adjusted for the last time in 2002. These rules have developed a variety of weaknesses due to changes in the market and development of government requirements and one particular of them is important for this research: the lack of a risk-sensitive approach to the determination of the solvency capital requirements. Solvency I calculates a factor-based capital surplus on the basis of the technical provisions (only liabilities), which is equal for all insurance companies. The asset allocation and influence of market risk is not a point of interest in this regulation framework.
In the new Solvency II directives, besides the risk deducted from the technical provision, market risk, credit risk and operational risk per insurer are also factors influencing the calculation of the Solvency Capital Requirement (SCR). The largest contributors to market risk, are equity, spread and interest rate risks. Also within the market risk module, the match of assets and liabilities (concerning the duration) is considered. This means that the asset portfolio, the diversification of the assets held and the asset-liability management will all have a significant impact on the determination of the required capital surplus of insurers. This is likely the cause for a change in asset allocation. The Solvency II guidelines have been presented on July 10th 2007, these have been accepted on April 22nd 2009 by the European Parliament and on May 5th 2009 by the European Commission.
The quantitative impact studies (QIS) which are conducted by EIOPA (the former CEIOPS) and the Nederlandsche Bank in cooperation with the insurers can be seen as a preparation of the insurance industry for the introduction of the new regulatory framework Solvency II. Because the first QIS have taken place since 2005 – despite that the definitive interpretation of the directives was still unknown – to measure and test the effects on the balance sheet of insurance companies, insurers have the possibility to anticipate by rebalancing their asset portfolio with less risky assets or improved diversification.
There are different trends to distinguish with regard to the composition of the portfolio. Some experts think insurance companies will adjust asset allocation to the new requirements, thereby maximizing returns for a certain level of risk. Others expect the asset allocation to remain broadly constant because the risk and return profile for an insurer will not change due to new requirements. As well as the asset portfolio can be adjusted, the insurance sector also has the chance to bridge the gap between the upcoming standards and Solvency I concerning internal control and risk management.
The subject of this study is to examine the relationship between the quantitative restrictions of the new Solvency II directives and the assets held in the portfolio of insurers. It is assumed that Solvency II will induce more conservative and prudent investment behaviour on account of the capital surplus factor as an increment of the solvency requirement.
1.1
Relevance
From the perspective of accountancy this subject has a connection with the specialty of internal control and risk management, but also with the main theories in the field of financial accounting. The regulation and standard-setting motives arise from the basic concepts of information asymmetry: adverse selection and moral hazard. Besides, the new government directives have a substantial impact on the audit procedures of the external auditor since Solvency II encourages the use of market-based values and requires other additional disclosures and it is the task of the external auditor to provide the requested assurance when auditing the financial statements.
This impact on auditing can also be seen in connection with the standard-setting aspect of financial accounting. With IFRS 4 phase II and IFRS 9 the accounting standards for insurance companies are also becoming more consistent with Solvency II requirements, although some differences in measurement and disclosure requirements remain.
In this sentence, this research is an addition to the effectiveness of solvency regulation by examining if the capital surplus requirements discourages investments in certain asset categories, for insurers in particular in stead of banks, and to prove that there is a significant ex ante effect originating from the future regulation. Most of the researches conducted are based on ex post data when regulation came already into operation.
It is remarkable that most studies in the scope of solvency and regulation are conducted within the banking sector. The comparison between banks and insurers can be made: the capital and solvency requirements lead to the holding of a capital surplus for banks as well as insurance companies. In this case this research is an addition to the one of Hellman et al. (2000) by investigating if the capital surplus requirements discourage investing in certain asset categories, particularly for insurers in stead of banks, and to prove that there actually is an desirable effect caused by the regulation.
Davis (2000) and Eling et al. (2009) suggest that solvency requirements have influence on portfolio allocation, specifically on the risk-level on the portfolio. Davis (2000) studied therefore the allocation of pension fund portfolios. Eling et al. (2009) developed a new solvency model, recognizing that solvency requirements can lead to a suboptimal asset allocation in the insurance industry. For these suggestions this research tries to find empirical proof on an ex-ante base and only for Dutch insurers.
1.2
Problem analysis and research questions
To investigate the relationship between the Solvency II framework and the asset portfolio of Dutch insurers, the following main question in this research is posed:
• Is the Dutch insurance industry already anticipating on the new quantitative restrictions as part of the new Solvency II directives (effective in 2013) and can we find significant shift in the investment portfolio during 2009 from relatively risky investments to safer assets caused by this ex ante preparation?
An affirmative answer to this question gives an indication of the ex ante impact of requirements posed by governments. Prudent investment behavior regarding the asset portfolio of insurers can thus be induced by directives. A negative answer does not mean there is no ex ante preparation at all, it just cannot be found in the investment portfolio due to several reasons (for example because the impact of the capital charges is still uncertain).
In order to find an answer on the main question, a division has been made into several research questions to determine the ex ante preparation and the influence on the asset portfolio:
1. What is exactly implied by the institutional framework, Solvency II, concerning capital surplus requirements?
2. Are there any indications, aside from the investment portfolio, which give evidence of ex ante preparations by insurers?
3. Has there been a clearly recognizable shift in the portfolio during 2009 from relatively risky investments (equities) to safer assets (bonds and other fixed-interest securities) related to the new Solvency II directives?
1.3
Conceptual framework
The concepts arising from the main research question are the allocation of the investment portfolio and regulation. The theoretical framework which my research includes is the association between the regulation from the European authorities and the influence on the asset portfolio of Dutch insurers measured in risks (before the regulatory framework has come into force). This perspective is supported by the concept of economic consequences: each directive on the specialty of accounting (including risk management) influences the course of action of the business when it comes to compliance with the relevant framework of regulation.
Besides the concept of information asymmetry has a role; managers are being influenced by both their own profit seeking behaviour, both the maximization of the firm value from the goal of the shareholders. This means they are willing to take more risk - to maximize their own utility - than accepted by stakeholders (including also policyholders) is expected; this is part of the agency theory. Regulation on the field of solvency is thus also meant as a sort of protection for the interests and motives of stakeholders from the concept of information asymmetry.
Finally the theories regarding asset allocation and portfolio management are part of this framework. The optimum investment portfolio has only market risk to be considered since all specific business risks are eliminated by diversification within the portfolio. This research assumes that companies are composing their investment portfolio according to the utility theory (or maximization of expected wealth) with the restriction that the risk level cannot exceed a certain level, to comply with the solvency requirements. Asset-liability management (ALM) studies are here an important starting point since with this is endeavoured to establish a balance between assets and liabilities for the long term. The influence of the regulation on the risk exposure on the asset-side of the balance is in this research studied.
Two incentives are important in this study. The first is the incentive of managers and shareholders to gain as much profit as possible to assure the management bonus and dividend distribution. The second one is the incentive to lower the technical provision since regulation poses capital requirements on the value of the insurance contracts.
The above can be summarized in a schematic figure, which explains the complete conceptual framework of this study:
INVESTMENT PORTFOLIO - Portfolio theory (Markowitz)
REGULATION - Solvency II Framework - Prudent person rules
(PPR) RISK APPETITE
- Moral hazard and information asymmetry - Utility and ruin theory
LIABILITIES - ALM studies INCENTIVE: - Low technical provision INCENTIVE: - High profit
Of course this research has its limitations since it is a broad subject. The goal and scope of this thesis is to prove the relationship between Solvency II directives and the investment portfolio of insurance companies before the institutional framework is effective. Therefore the research only focuses on the first pillar of the Solvency II framework. This is the pillar concerned with the quantitative restrictions which influence the capital charge also by the market risk module. The directives are not subject of the debate in this thesis. The second limitation is the categorization of the asset portfolio. In this research the only categories that are distinguished are bonds and stocks. Financial instruments and real estate are not counted for. To assign a risk profile to the asset portfolio, the general assumption is used that bonds have a lower risk profile than stocks. A conservative approach to investing, prefers bonds to stocks. This will be further analyzed in chapter four and six.
Finally the decision is made to only collect data from the Dutch insurance industry and only for life insurance companies and property and casualty insurers. Since the annual reports of 2010 were not published when the research started, the research is based on the data of the Nederlandsche Bank and annual reports of 2009. This means that the situation of January 1st 2009 is compared with the balance on January 1st 2010.
1.4
Structure
The research starts by explaining the institutional framework of Solvency II in the second chapter. The main reasons for a new framework and the approach is presented. The specific parts of the directives and the ways that these directives can influence the asset portfolio by setting capital charges is also stated. The third chapter starts brief with an introduction to the portfolio theory of Markowitz. The second category of literature relevant for this subject highlights the aspects of solvency, capital requirements and government regulation. The last part concerns literature on the subject of asset allocation, portfolio management and diversification. The concepts as stated in section 1.3 can also be found in chapter three and are discussed in the literature.
Chapter four starts with the research design and methodology. The research questions are described and a brief introduction to the statistical analysis is given, including a summary of the formulas used. The methods for sample selection and data collection forms the final section of this chapter.
The next chapter answers the first research question. It is a short qualitative analysis of indications of anticipating insurers found in the annual reports, categorized in the following factors:
1. Participation in the development of the framework; 2. Composition of an internal Solvency II team;
3. Design or revision of the Own Risk and Solvency Assessment (ORSA); 4. Solvency II as point of significant interest to risk and asset management;
In chapter six the statistical part of this research is conducted, starting with the descriptive statistics of the sample. The further statistical research starts by proving that the insurers favor bonds in stead of stocks. The second part answers if a disposal of stocks (assets with a high risk profile) automatically leads to an acquisition of bonds (assets with a low risk profile). Finally the influence of the current funding ratio is studied.
The conclusions can be found in chapter seven, in the first sections the conclusions for each research question are summarized and in the last section an overall conclusion is drawn. Recommendations supporting the conclusion can also be found in this chapter. Chapter eight follows with the limitations and possibilities for further research, based on the limitations of this research.
2
Institutional framework (Solvency II directives)
Although Solvency II will become effective not earlier than January 2013, there already have been five quantitative impact studies (QIS) conducted by the European Insurance and Occupational Pensions Authority (EIOPA). In the most recent study,QIS5, almost 70 percent of the European insurers were participating. Aim of these studies is to estimate the impact of the directive as well as fine-tuning the standard formula and encouraging the preparedness of the insurance industry and financial authorities. In this chapter, first is briefly explained the most important concepts and aspects of the Solvency II directives and the consequences for the insurance industry. Next, to keep relevance with the subject, the focus lies only on the first pillar of Solvency II: quantitative restrictions. These restrictions are the calculation of the technical provision, the Minimum Capital Requirement (MCR) and the Solvency Capital Requirement (SCR). The latter is determined by the level of risk minus the diversification effects, whereas the MCR functions as an absolute bottom level. It will be clear that only the SCR has a potential to influence insurers in determining their asset portfolio. The conclusion of this chapter is an answer to the first research question:
• What is exactly implied by the institutional framework, Solvency II, concerning capital charges?
2.1
Solvency II
2.1.1 Causes for new directives
As the current Solvency I framework has had its last adjustment in 2002, a variety of changes in our economic environment have taken place. The shortcomings due to changes in the market and development of government requirements and the development of a new framework for the banking sector (Basel II) made it clear that a new framework was needed.
The most important issues for designing new regulations are the following: 1. Lack of risk sensitivity
The current Solvency I framework only takes the liability-driven risk into account, based on an insurers technical provision. The asset allocation is completely ignored, although this does influence the risk profile of an insurer. Other risks that are not included are for example market risk, credit risk and the operational risk. Besides, the current framework does not have many qualitative requirements regarding risk management and governance
2. Interferences for a efficient internal market
Due to Solvency I the minimum requirements are well-known in the insurance industry, but these can be adjusted by additional regulation imposed by local regulators. The differences among countries within the European Union for solvency requirements have shown to be the consequence.
3. Suboptimal regulations for supervision
Solvency I focuses mainly on the juridical entities or business units although the possibility exist that these entities form a group. The danger of seeing each business unit as an object for supervision has also the consequence that firm-wide risks are not observed. It also leads to a gap between the way business units are controlled and the way the authorities supervise these units.
4. No international convergence
The current Solvency I framework does not operate conform an economic risk based approach, while IASB and other standard-setting institutions already do. The fair value part of the regulation is also missing in Solvency I.
2.1.2 Approach
Since the Solvency II regulation is broadly analyzed by investors, auditors and regulators, detailed information can be found everywhere. A general approach, a brief description, will be presented here. The Solvency II framework consists of three pillars:
1. Pillar 1: quantitative restrictions / capital requirements 2. Pillar 2: qualitative restrictions / general requirements 3. Pillar 3: disclosure requirements
The first pillar, area of interest in this research, focuses on the technical provisions and calculation of the solvency and minimum capital requirement (SCR and MCR) and poses rules for investments. The Solvency II framework is an equivalent of Basel II: “it addresses risk from the perspective of quantitative requirements, supervisory process, market transparency and disclosure.” 1
The second pillar contains regulations and requirements regarding the internal control environment and governance policy of an insurance company, which will be monitored by the supervisory authorities. The third pillar focuses on the disclosures of the insurer. Diverse aspects of an insurers’ solvency, governance, risk management have to be published and some specific capital management issues have to be included.
2.2
Solvency Capital Requirement under Solvency II
Structure of the Solvency Capital Requirement (SCR) can be found in appendix A. SCR consists of basic SCR + operational risk SCR. Basic SCR is divided in five other categories. The most important component from the calculation of basic SCR for my research study is the market risk module. According to EIOPA “the market risk module shall reflect the risk arising from the level or volatility of market prices of financial instruments which have an impact upon the value of the assets and liabilities of the undertaking.”2 Market risk module of QIS5 contains seven types of risk:
• Foreign exchange risk (Mktfx)
• Property risk (Mktprop)
• Concentration risk (Mktconc)
• Interest rate risk (Mktint)
• Equity risk (Mkteq)
• Spread risk (Mktsp)
• Illiquidity risk (Mktill)
As said in the introduction: large contributors to market risk are equity, spread and interest rate risks. This is likely the cause for a change in asset allocation. There are different trends to distinguish with regard to the composition of the portfolio. Morgan Stanley expects fund managers to maintain low allocations to equity and property. Besides experts think insurance companies will adjust asset allocation to the new requirements, thereby maximizing returns for a certain level of risk. Deutsche Bank takes a different point of view. They expect the asset allocation to remain broadly constant
1 Solvency II, Interpreting the key principles, published by Ernst & Young LLP, EYG no. EG0016, 2008
2 Consultation Paper no. 69: Draft CEIOPS’ advice for Level 2 implementing measures on Solvency II, art. 109b Equity risk
In the report from EIOPA about QIS 3 it appeared that the market risk contributes to about 70% of the basic SCR of life insurance companies. For property and casualty insurance companies, this is almost 75%. The diversification effects in the portfolio to decrease the influence of the market risk lies, according to EIOPA, between the 14 and 36%. In the research of ORTEC Finance the prove is given that the solvency ratio of an insurance company can change dramatically from almost 400% under the directives of Solvency I to only 122% when calculated conform the Solvency II directives in QIS4.
To give an example why insurance companies have an incentive to adjust the strategic asset allocation: for hedge-funds a risk charge of 45% is applied and for shares this is 32%. This risk charge is based on the expected returns, the volatility of the market price and the risk profile of the asset. The risk charge is later translated into the capital charge. Within the Solvency I framework, the portfolio allocation and diversification did not have any impact on the calculation of the required capital charge.
2.3
Vision on Solvency II
The new regulation framework coming into force starting 2013 has a significant impact on the operation environment of insurance companies. With a fundamentally different method to determine the capital requirement, it is demanded that insurers are already anticipating. The QIS are an excellent way to start. The risk sensitive approach and influence of the complete balance sheet on the capital requirements will cause insurance companies to revise the risk management system and a structured data recording.
For Dutch insurance companies the impact of the new directives may be less than in the rest of Europe. Almost all insurers operating on the Dutch market tend to hold already more capital than required. Still it is important to realize that a well-diversified portfolio is rewarded and that the market risk for life insurance companies is the most important component of the basic solvency capital requirement calculation.
The risk of insurance companies becoming insolvent is adequately reduced by the new framework. With the equity market crash of 2001-2003 and the latest financial crisis 2008-2009 it appeared that many insurance companies cause themselves a lot of trouble by still having the perspective to take a high risk and reward structure. This is not the case under Solvency II. Although it may influence the efficient operation of the insurers by posing regulation on the profit aspect of the investments, the reasons why a new framework is necessary are very clear.
3
Literature
Scientific research relevant for my study will be categorized by subject. The first category of literature contains the ‘traditional’ portfolio theory by Markowitz. The second highlights aspects of solvency, capital requirements (or charges) and (government) regulation. The third category concerns asset allocation, portfolio management and diversification. Although most research is conducted in the banking sector and among pension funds, the used concepts and outcomes can be used for the insurance sector as well. The goal of setting capital requirements by regulators, for example, is the same for the banking sector as well as the insurance industry. The calculation follows a different pattern but the aim of the directives stays the inducement of prudent behavior and reduce the risk of insolvencies.
3.1
Portfolio theory by Markowitz
The foundations of portfolio theory (often called ‘Modern Portfolio Theory’) were laid by Markowitz (1952) to fill the gap of theory about the effects of diversification and the trade-off of risk and return in a selected portfolio. His theory is part of the scientific theories of rational behavior under uncertainty. In explaining rational investment behavior uncertainty is an essential aspect. Markowitz theory is mainly based on two elements: expected value and variance, where variance is the measure of risk.
Markowitz distinguishes two phases in the process of selecting and composing a portfolio under risk and uncertainty, starting by observation and ending with the definitive portfolio choice on a given time. For explaining investment behavior Markowitz considers the following rule: “the investor does (or should) consider expected return a desirable thing and variance of return an undesirable thing, within a finite number of alternate probability distributions.” This is the “expected returns - variance of returns” rule. The rule that states that an investor only maximizes return is rejected since this rule does not lead to portfolio diversification, which is reasonable and observed as investment practice.
This mean-variance analysis is further studied when (expected) utility theory as a more complex measure was introduced. Markowitz describes in 1976 the foundations, arguments and differences between the theories and explains why mean and variance are essential data to maximize expected value of a portfolio. Markowitz argues that mean-variance analysis is a much more convenient and more economical way of calculating (by minimizing standard deviation of return for different expected returns) than appointing the maximum expected value of utility functions. In addition, the necessary inputs are easier to find. The last argument Markowitz gives is that “real investors these days usually seem more comfortable with the idea of examining risk-return tradeoffs than with psychoanalyzing their utility function and letting the computer pick a portfolio that maximizes its expected value.”
But the adequacy of the answer given by mean-variance analysis has been doubted by supporters of the utility theory and analysis. Markowitz argues that an investor does not lose much when selecting its portfolio on mean-variance analysis since utility analysis is less convenient (and economical). This is also explained in a 1984 paper of Markowitz: when portfolios are put in order by using expected utility as well as mean-variance, both orders obtained are nearly identical. Markowitz theory has also been empirically proved when an efficient portfolio is selected among an infinite number of alternate distributions and when the effect of leverage is included.
3.2
Aspects of solvency, capital requirements and regulation
There has been a lot of research conducted concerning capital requirements. In the research of Hellman et al. (2000) is examined if an increase in capital requirements can compensate for the disadvantageous effects of liberalization. They pose that the competition between banks caused by liberalization of the markets undermines the prudent behavior of banks from the perspective of moral hazard. Capital requirements reduce the motive to ‘gamble’ (for example by risky and/or speculative investments). Kane
(1989) and Cole et al. (1995) formulate this in the following words: “banks choose a risky asset portfolio that pays out high profits or bonuses if the gamble succeeds but leaves depositors, or their insurers, with the losses if the gamble fails”3
.
The conclusion of Hellman et al. (2000) is that capital requirements are not always as effective as previously thought. As additional measures to support the capital requirements the researchers suggest regulation in the form of asset-class restrictions and exposure rules if this limits the banks to invest in speculative asset categories. Beside the support of capital requirements, this also offers a possibility to restrict the problem of moral hazard. Bhattacharya et al. (1998) also state that risk-sensitive capital requirements are potentially useful devices to handle the problem of moral hazard.
Opposite conclusions are drawn by Munch and Smallwood (1980) who has obtained empirical evidence about the effects of solvency regulation in the property and liability insurance industry. The researchers assume that a firm chooses its own probability of insolvency. The conclusion is that the requirements regarding the minimum amount of capital actually diminish the number of insolvent companies. Furthermore is concluded that, based on the observations, the traditional model of insolvency is supported where insolvency is “the unlucky outcome of value-maximizing risk-taking”. According to the research of Lamm-Tennant (1989) maximization of returns is also the most important goal of the investments held. Here is also concluded that a process of asset liability management is an essential contribution to the profitability and solvency of life insurance companies.
The research of Myers and Read (2001) points out how capital surplus should be allocated. In the case of a property casualty insurer, the capital requirement should not only depend from asset portfolio, but also from the several business lines within the company. The correlation and diversification effects across the lines of business should be taken into consideration. High surplus requirements with a minimum risk of the insurer becoming insolvent have the consequence that the costs are passed on the customers, which is an undesirable effect. Although the different business lines influence the calculation of capital surplus allocation, Myers and Read (2001) find that reducing asset risk is the best possibility to reduce the capital surplus requirements, which supports this research.
Eling et al. (2009) define a new approach to solvency regulation and assessment. In stead of risk-based capital standards with a minimum capital requirement, the researchers set minimum requirements for investment performance. Key item in this newly developed model is the solvency line: “the solvency line determines minimum investment requirements in terms of expected return and standard deviation of the rate of return on the insurer’s investment portfolio.” This means that insurers with sufficient or higher equity capital have more possibilities in selecting their portfolio. The incentive of management to hold no more capital than by regulators required (as in the case of Solvency II) can lead to a suboptimal asset portfolio with low-risk profile, which can be a threat for the profitability of the company.
An advantage of this model is that the contrary is possible: depending from insurers equity capital and risk-level of their liabilities, insurers can profit from assets with higher risk and return. Another advantage is that adjusting a portfolio to meet the requirements is easily accomplished (assuming all assets are traded in a liquid market) and the costs are probably less than those to raise additional capital. In the paper the model has been tested on a German non-life insurer. There is no empirical evidence that this model is also applicable to life insurance companies as well. However, the model of Eling et al. (2009) is an interesting subject with multiple fields of applications, for insurance companies as well as for regulators on the financial market. The general view is that solvency regulation encourages companies to operate inefficiently, but the risk of insolvency is actually reduced (Myers and Read. (2001), Munch and Smallwood (1980), Bijapur et al. (2007).)
3Hellmann, T.F., Murdock, K.C., Stiglitz, J.E. (2000). Liberalization, moral hazard in banking, and prudential regulation: are
Solvency regulation can also be seen in a different perspective: not only as a way to reduce the risk of becoming insolvent, but also as a role of regulation to address the agency problem – specifically for risk-taking strategies – between customers (policyholders) and shareholders. The main problem exists since shareholders can create more value with a risky asset strategy, but this reduces the amount to be invested. The rationale is that an increased risk of insolvency lowers the premiums customers are willing to pay, which will lead to a reduction of the investment. Filipovic et al. (2009) conclude that solvency regulation does contribute to a solution for the inefficiencies leading from the agency problem, for example by limiting the possible investment strategies.
The last aspect of solvency which is important to mention, is that the classification of solvency depends from the detection or calculation method and the accounting principles used. Two detection methods of solvency classification are tested by Grace et al. (1998): risk-based capital and the financial analysis tracking system – as used by the NAIC. The power of both methods to identify troubled insurers is not identical and even combining both does not lead to a better indication of weak insurers (companies with a higher risk of becoming insolvent). An important issue with both formulas is that they rely on accounting data which can be manipulated by the accruals and thus the accounting principles used. To obtain a higher accuracy regarding the financial strength of insurers, other information should also be collected, for example external judgement of an expert.
The outcome of monitoring solvency and identifying weak insurers also depends from the accounting principles or reporting standards used. The research of BarNiv (1990) in the property and liability insurance sector of the United States had given empirical evidence that statutory accounting principles (SAP) and market value and cash-flow based accounting principles (MVA) can predict insolvencies with greater accuracy than GAAP (generally accepted accounting principles). The tendency to move to a more fair value estimation of assets and liabilities is thus also supported in the scope of monitoring solvency.
3.3
Asset allocation, portfolio management and diversification
In Davis (2001) two alternative approaches of regulation on the asset portfolios are subject of the research. These two which can enhance diversification in a portfolio of a life insurer or pension fund are prudent person rules (PPR) and quantitative restrictions (QR). The conclusion states that PPR function better than QR in the case of pension funds, except in certain specific circumstances, like emerging markets. Davis (2001) remarks that PPR can be considered for life insurers, particularly for those operating in developed countries in a competitive market.
Also in Bijapur et al. (2007) research is conducted in the area of regulation of asset portfolios. An econometric model is formulated to test the hypothesis that quantitative restrictions (QR) impose costs on the company in connection with the suboptimal returns on investments. A statistically significant influence is found on the portfolio returns of life insurance companies in seven OECD-countries caused by QR, varying between 0,2 – 0,3%. The flexibility of prudent person rules (PPR) allows the company to obtain a more diversified portfolio and to reduce the market risk (and thus the company can obtain a higher return for a given level of risk).
In stead of an econometric model, Milne (2002) investigates the influence of capital requirements and risk standards as posed by Basel I on the portfolio allocation of banks. In this research is stated that when a bank on a certain point reaches the boundary of minimum capital requirements, the bank can always reduce its exposure by adapting its portfolio (as long as the investments are traded on a liquid market). Only when adjustment of the portfolio is relatively expensive or impossible, the portfolio choice will be influenced by the risk standards determined by the (financial) authorities. If this conclusion is also representative for life insurers, this means that the asset portfolio remains unaffected to influence of upcoming directives - assuming all assets are traded on a liquid market.
Milne assumes that violation of the capital regulations is judged ex post (“incentive based”). This perspective forms a contrast with the conventional view of capital requirements as ex ante enforcement which are taken into account in decisions concerning strategic portfolio choices. Milne (2002) draws the conclusion that the impact of capital requirements, where the constraints are risk-weighted, on portfolio allocation depends on the liquidity of the assets held. The threat of sanctions posed on violation of the capital regulations leads to risk-averse behavior in banking, especially on the short-term horizon. If this case applies to insures, a reduction of portfolio risk can be found for example by a shift in strategic allocation: a reduction of exposure to stocks and an increase of exposure to (government) bonds.
Gatzert and Schmeiser (2008) conducted a research with the calculation of minimum safety levels and capital requirements according to two methods: value-at-risk and tail-value-at-risk. Various factors are taken into account to measure the influence on the safety levels and capital requirements, including the shift to an asset portfolio with more risky financial assets. The conclusion for non-life insurers with a recent shift from low-risk to high-risk is that additional capital is necessary to meet the requirements. But, although requirements are met, the costs of an insolvency in the non-life insurance industry differ substantially.
Concerning the asset allocation and portfolio management, a lot of theories are applicable. In section 3.1 the portfolio theory has been explained. The conventional model with regard to strategic asset allocation is the ‘ruin theory’ (collective risk theory or value-at-risk). This theory is, according to Sherris (2006) especially used when determining the risk standards within the directives. An alternative approach is the ‘utility theory’ (maximization of the expected profit). The latter theory can be used for optimization of the portfolio, according to Cummins and Nye. (1981). In this, the researchers also claim that since the eighties the ‘utility theory’ is used as an approach with investments by insurance companies.
When it comes to the investment portfolio, the high risk appetite is often founded by the concept of information asymmetry and the underlying aspects of adverse selection and moral hazard. After all there is also an incentive problem between shareholders and policyholders in the insurance industry. This is also the starting point in the research of Rauh (2006). He takes the point of view that management risk-shifting incentives dominate, corresponding to the concept of moral hazard. It appears that a correlation exists between the financial condition of pension funds and the amount of the asset portfolio invested in risky assets. Well-funded pension plans have an asset portfolio which contains more volatile assets compared to poorly funded plans. The firms the poorly funded plans appear to prefer safer assets in their portfolio. As Rauh (2006) states: “the higher the probability of bankruptcy, the safer the observed pension fund asset allocation.” This finding does not support the concept of moral hazard as was previously thought. For this research, it means that insurers with a lower funding ratio have a less risky asset portfolio than insurers with a sufficiently high funding ratio.
4
Research design and methodology
This research is mainly a quantitative analysis. By the data from annual reports over 2009 I examine if there is a significant ex ante influence from the directives of Solvency II. This influence can be recognized by measuring the shift in risk classification of investments, assuming that bonds have a lower risk profile in comparison with shares. If the future legislation has ex ante impact, I assume to find a shift in the portfolio: more bonds (safe assets), less shares (risky assets). The claim that bonds are more safe than shares is not always valid. The rating of junk bonds (or high-yield bonds) is below investment grade, which means Standard & Poor’s gave a rating below BBB-. This means the risk of default is higher and therefore a higher yield is paid. This research holds on to the general theory that when it comes to yields and dividend, bonds are less volatile compared to shares and a safer option to gain income from.
In this chapter the design of the empirical research is described by presenting the research questions and hypotheses. Furthermore, the way of selecting the sample of insurers and collecting the necessary data will be explained.
4.1
Research questions and hypotheses
The questions in this study are the following:
1. Are there any indications, aside from the investment portfolio, which give evidence of ex ante preparations by insurers?
2. Has there been a clearly recognizable shift in the portfolio during 2009 from relatively risky investments (equities) to safer assets (bonds and other fixed-interest securities) related to the new Solvency II directives?
The first question is an addition to the quantitative research. In the literature voluntary disclosure is a free choice to provide investors with accounting and other information to help the decision-making process of users of an annual report. Although in the asset portfolio may not have been a shift, insurers can already announce the impact of Solvency II in their annual reports and giving evidence that the company is anticipating on the upcoming directives.
Since the framework of Solvency II has been presented in July 2007, I expect that insurance companies have already assigned the impact of the new legislation on their organization. The next step is to reduce the impact by preparing the company and control environment on the new Solvency Capital Requirements by assessing and adjusting the asset allocation. The hypotheses for questions 2 and 3 (the quantitative part of this study) are:
- H0: the shift in the asset portfolio during 2009 for life insurance companies and property
casualty insurance companies is not the result of the ex ante impact of the new Solvency II directives
- H1: the shift in the asset portfolio during 2009 for life insurance companies and property
casualty insurance companies arises from the new Solvency II directives
4.2
Design and methodology
4.2.1 Evidence of ex ante preparation aside from the investment portfolio
The first question, to find if insurance companies are already preparing their selves for the new legislation, will be answered by a case-study of annual reports of financial year 2009. In the annual report companies give attention, besides the profit figures, to risk management, strategy and corporate governance. Several disclosures are required by legislation, especially International Financial Reporting Standards (IFRS). I
expect insurers to report on the impact of Solvency II on the company and which measures have already been taken by the board of directors to prepare the company on the new capital requirements.
For the following questions, which can be seen as ex ante preparations on the new requirements, I investigate if the company answers them affirmatively in the annual report of the financial year 2009. If the phrase in the annual report lies in the given range for each question, it counts as an affirmative answer. The questions and range can be found on the next page.
Research questions and corresponding range:
1. Does the company actively participate in the development of the framework?
2. Is there an internal team specialized in Solvency II issues, composed for preliminary operations? 3. Is the insurance company already designing or revising the Own Risk and Solvency Assessment
(ORSA)?
4. Is Solvency II a point of particular interest to risk management? 5. Is Solvency II a point of particular interest to asset management? Range concerning question
no.
Phrases / answers counted as an affirmative answer
1 - Participation in the Quantitative Impact Studies (QIS) conducted by the former CEIOPS in association with De Nederlandsche Bank;
- Participation in the project “RiSK”4 conducted by De Nederlandsche
Bank during 2009;
- Participation in the European discussion regarding Solvency II as part of a branch group.
2 - Presence of a project team regarding Solvency II;
- Performance of a gap analysis and determination of a roadmap until implementation in 2012;
- Education, training or guidance on the subject of Solvency II.
3 - The insurance company already uses an internal model to calculate the capital requirements, already work with ORSA and states in the annual report this model will be revised to attain compliance with Solvency II; - The procedures and components of ORSA are already formulated, but
preliminary operations are not completed yet – also regarding the calculation of SCR.
- The insurance company has to set up an ORSA, compliant with Solvency II and states that the designing of this model has already started.
4 and 5 Only an affirmative answer is possible when the company states that the risk management and / or asset management function give(s) significant attention on the issues raised by Solvency II.
When stated that Solvency II pillar 2 and 3 receive significant attention, this is counted as an affirmative answer for the risk management function since compliance with both pillars is primarily a responsibility of the insurer’s risk management function.
4 Since there has not been a QIS on a European level during 2009, De Nederlandsche Bank launched the RiSK project. This
project had the same goal as the European studies, but was also meant to keep Solvency II as an area of special interest to Insurance companies.
This part of the research is included in the next chapter.
4.2.2 Asset allocation and the shift from risky to safer asset categories
The asset allocation of the companies included in the sample is extracted from the statistics regarding financial year 2009 published by De Nederlandsche Bank. The assets can be categorized in different ways and on several risk levels. In accordance with the original design of this research I would have to use the rating of Standard & Poor’s or Moody’s. Since this data does not classify the investments on that level but only in terms of ‘stocks’ and ‘bonds’, I rearranged the classification of risk for my research in more simple terms since I only use publicly available information. According to the theory described, stocks clearly have a higher risk even though the rating is the highest possible. Bonds, especially government bonds, are a less risky investment since they guarantee a return.
The quantitative analysis, question 2 and 3, starts by calculating the balance of acquisitions and disposals (bonds and stocks) in a standardized portfolio. This approach is necessary since the size of the investment portfolio varies per company in the sample. I measure the balance of acquisition minus disposal in a portfolio of € 1.000.000 to exclude the influence of size. Then I calculate the descriptive statistics of the sample and test the assertion if insurance companies favor bonds in stead of shares using the balance of acquisitions and disposals per category. The next question is if correlation exists between the redemption of shares and acquisition of bonds. The answer is deduced from the results of the statistical test on correlation (on the basis of the Pearson correlation coefficient). The formulas used for the statistical analysis are listed below:
• Calculation of sample mean (̅):
̅ = 1
• Calculation of sample variance (s2):
= 1
− 1
( − ̅)
• Calculation of confidence interval:
̅ ± /; ∙
√
• Execution of t-test:
=̅ − /√
• Comparison of two means:
=̅ − /√
• Regression analysis:
β=∑(X − X ) ∙ (Y − Y ) ∑(X − X )
• Pearson correlation coefficient
= ∑(X − X ) ∙ (Y − Y ) ∑(X − X )∙ ∑(Y − Y )
• Test on correlation
=r√n − 2 √1 − r
The last part of this study explains if there is any significant influence of the funding ratio on the disposal of stocks. The reason that the funding ratio is also considered as a factor of influence, lies in the conclusion of the research Rauh (2009) has conducted. He concludes that when firms have poorly funded pension plans, the assets are mostly allocated to safer securities such as government debt and cash, while firms with well-funded plans select more volatile asset classes like equity. Although his research focuses on corporate pension plans, I think his conclusion applies to the funding ratio of insurance companies as well.
4.3
Data selection / collection
In the first place I would like to indicate that only insurance companies competing in the Dutch market are included in the research. The conclusion thus only refers to the Dutch insurance industry. On the basis of the statistics published by the DNB I made a selection of insurance companies with bonds as well as equity in their asset portfolio. For these companies I checked if the company’s financial statements could be obtained. Most insurance companies are a business unit of a bigger group and the consolidated financial statements do not contain the information regarding the purchases and redemptions of asset categories for the simple company. The sample is adapted to the availability of the annual reports of the simple firm.
The databank of the DNB regarding the insurance sector (year 2009) counted 62 life insurance companies and 219 non-life insurance companies. Among the category of non-life insurance are different classes to distinguish, like health insurance and the property casualty insurance companies. These classes are further divided by branches (fire and other damages, accidents, motor vehicles et cetera). I take all the categories into account for the calculation of the total population, so for life insurance N=62 and for non-life N=219. For the sample of the life insurance industry I selected 15 companies, which represent 64,5% of the gross premium earned in this industry. From the population of non-life insurance companies, I also selected 15 companies which represent 12,4% of the gross premium (including health insurance). When health insurance is left out (only property casualty insurance companies), 41% of the gross premium is covered by my sample. The exact sample is included in the appendices, section 7.1.
Unfortunately during the research process I found out that, for one life insurance company, the required table with the exact movements of financial investments was missing. Therefore I had to remove this company from the sample.
5
Other indications of early anticipation of insurers
In this chapter I perform a small case study on the annual reports (financial year 2009) of the life insurance and property casualty insurance companies included in the sample. The aim of this qualitative research is to prove that insurance companies, even though it might not be visible in the investment allocation, anticipate the new regulations by mentioning their preliminary activities regarding Solvency II (for example in the ‘recent developments’ section or ‘outlook 2010’ in the annual report). The relation of this small study to the hypothesis, is that voluntary disclosures may indicate anticipation on the Solvency II framework even though the asset allocation has not been adjusted. The research questions to be answered:
1. Does the company participate in the development of the framework?
2. Is there an internal team specialized in Solvency II issues, composed for preliminary operations? 3. Is the insurance company already designing or revising the Own Risk and Solvency Assessment
(ORSA)?
4. Is Solvency II a point of particular interest to risk management? 5. Is Solvency II a point of particular interest to asset management?
The possible answers for each question are: yes, no or not mentioned in the annual report. I use the sample as described in chapter 4; this sample contains 14 life insurance companies and 15 property casualty insurance companies. There are a few remarks noted in section 5.6, in addition to the given answers in the sections 5.1 to 5.5.
5.1
Participation in the development of the framework
The research question in this section runs as follows: “does the company participate in the development of the framework?” The scope in which answers count as “yes” are for example the mentioning of the company’s participation in the Quantitative Impact Studies (QIS) or in the project “RiSK”. Also is participation in the European discussion regarding Solvency II counted as an affirmative answer. The results deducted from the annual reports:
Life insurance companies
Yes No Not mentioned
Number 10 0 4
Property casualty insurance companies
Yes No Not mentioned
Number 7 0 8
5.2
The presence of an internal team for Solvency II issues
As participation in the QIS study can be the responsibility of only one or two employees, the next research question is: “is there an internal team specialized in Solvency II issues, composed for preliminary operations?” An example of a clear affirmative answer is the presence of a project team regarding Solvency II. What also gives evidence of the presence of an team operating on the subject is the performance of a gap analysis and determination of a roadmap until the regulation comes into effect in 2012. The training or other guidance of employees on the subject of Solvency II is also counted as an affirmative answer. The results can be found on this and the next page:
Life insurance companies
Yes No Not mentioned
Number 8 0 6
Property casualty insurance companies
Yes No Not mentioned
Number 6 0 9
5.3
Designing and revising of the ORSA
The question I research in this section: “is the insurance company already designing or revising the Own Risk and Solvency Assessment (ORSA)?” The insurance industry has to control risk in a manner compliant with Solvency II and the ORSA is a significant part of this. A possible answer which indicates ex ante preparation is that the insurance company is already using an internal model to calculate capital requirements, is already working with ORSA and only has to revise it to attain compliance with Solvency II. Also a complete formulation of procedures and components of the ORSA or the start of designing the ORSA counts as a sign of ex ante preparation. The results from the annual reports are the following:
Life insurance companies
Yes No Not mentioned
Number 8 0 6
Property casualty insurance companies
Yes No Not mentioned
Number 6 0 9
5.4
Solvency II and the risk management function
The question here is if Solvency II is a point of particular interest to risk management. This has not only to do with designing or revising ORSA (this can also be done by a third party) since there are a lot of other issues involved in complying with Solvency II concerning risk management. The risk management function is responsible for accomplishing all these tasks. The annual reports show the following distribution of insurance companies which give significant attention to the Solvency II issues within the risk management function and insurance companies which do not:
Life insurance companies
Yes No Not mentioned
Number 7 0 7
Property casualty insurance companies
Yes No Not mentioned
Number 5 0 10
5.5
Solvency II and the asset management function
The same research question as mentioned in section 5.4, but now for the asset management function: “is Solvency II a point of particular interest to asset management?” When the insurance industry knows the
capital requirements and the matching charges on capital, the asset management function is able to revise the strategic allocation of their investments to reduce the capital charges for example by diversifying the portfolio to diminish risk or by choosing other asset categories. The results of the research in the published annual reports if the asset management function gives significant attention to the issues raised by Solvency II:
Life insurance companies
Yes No Not mentioned
Number 0 0 14
Property casualty insurance companies
Yes No Not mentioned
Number 0 0 15
5.6
Remarks
It is remarkable that in the annual reports none of the insurers can be categorized in “no” for each question. One explanation for this observation is the that companies rather not disclose certain facts when those facts admit the company ‘fails’ in a certain way. In this case: not preparing. It makes sense since no company wants to have shareholders or a supervisory organ reading that the company left some important issues undone. Another explanation is that all companies reporting on the preparation on Solvency II requirements are voluntarily disclosing this information, since disclosure on this subject is not required by legislation. This voluntary disclosure goes beyond the minimum required disclosures laid down by IFRS or Dutch GAAP. Besides this reason, it is also sensitive information that the competitors don’t need to know.
Another remark has to do with the number of answers in the category “not mentioned”. AEGON has not mentioned anything in the annual report about the activities concerning the steps towards Solvency II. Yet we know that a company that size gives attention to the issues, it simply has to. Research on their website shows that this life insurer did participate in QIS and is revising the ORSA. A plausible explanation is the same as the argument in the previous section: reporting on the preparation on Solvency II happens voluntarily and not all insurance companies are disclosing information on this subject since it is not legally required.
Furthermore a few companies are business units of one group. Preparation and implementation of measures for Solvency II happens at group level in stead of business unit level. This concerns for example ING, Nationale Nederlanden en RVS in the category life insurance. Also Achmea, Avero and Interpolis are part of the same group called Eureko. ABN AMRO is a part of the Delta Lloyd Group. AEGON is as well part of the group AEGON Nederland.
6
Results
Panel 1 in the appendices section 7.2.1 includes the disposals and additions of shares in the portfolio of life insurance companies. These are absolute numbers since only disposals and acquisitions have been measured. The return on shares (unrealized gains or losses), depreciations and movement in valuation (caused by changes in exchange rates for example) have not been counted for. Panel 2 contains the disposals and additions of bonds in the portfolio of property casualty insurance companies,
The same calculations have been made for property casualty insurance companies for shares as well as bonds. The results can be found in panel 1 (shares) and panel 2 (bonds) in appendices section 7.2.2. The same conditions as applied for property casualty insurance companies are valid for the life insurance sector. This means, summarized, that returns, depreciations and changes in valuations have not been counted for. There is one difference between both of the table series. For life insurance companies were also 15 companies selected, but it appeared that the one company’s financial statements lacked a table showing the exact movements of the financial investments. Therefore this company has been removed from the selected data.
The following methodology is used to calculate the amounts recorded in the tables. The aim of this study is to examine if there is a relationship between Solvency II and the shift from relatively risky investments (shares) to safer assets (bonds and other fixed-interest securities). First, to measure the shift it is not recommended to subtract the book values at year-ending (t – t-1) mentioned in the financial statements. The book values at year-ending contain noise since the result of the deduction contains also returns and depreciation, which are not caused by buying or selling the investments. This is the reason that the disposal and acquisition as absolute numbers are used. In the second place the size of the companies and thus the size of the investment portfolio differ. Acquisition minus disposal is also dependent of the factor ‘size’. Therefore is decided to measure the balance of acquisitions minus disposal as if it had taken place in a portfolio of € 1.000.000. For example in table X.4 the no. 11 seems to dispose a huge amount of bonds, but this is a disproportionate view. On a portfolio of € 1.000.000 it is even less than the no. 8.
6.1
Descriptive statistics of the sample
Bases on the tables on the previous pages, descriptive statistics have been calculated for the sample selection. Panel A Category Investment category Sample median ∆ acquisition / disposal per €1.000.000 N n Life insurance companies Stocks € 17.232,19- 62 14 Bonds € 44.074,70 62 14 Non-life insurance companies (property and casualty insurance)
Stocks € 464,70 219 15
