Exchange Traded Fund Savings Plans as Old-Age Provision Tool in Germany

Contents

1 Introduction

2 Research question and objectives

3 Literature review
3.1 The German pension system
3.2 Exchange-traded funds
3.2.1 Suitability as a private pension alternative
3.2.2 Comparison to the commonly used pension alternatives
3.3 Current state of research
3.4 Theoretical principles of acceptance
3.4.1 Acceptance model for ETF savings plans
3.5 Summary of the literature review

4 Data and Methods
4.1 Research question and hypothesis
4.2 Research design
4.2.1 Questionnaire
4.2.2 Operationalisation of the variables
4.2.3 Data collection and sample frame
4.2.4 Data evaluation methods

5 Results
5.1 Reliability test and factor analysis
5.2 Descriptive statistics of the sample
5.3 Regression analysis of the extended TAM
5.4 Moderation and mediation analysis for the extended TAM
5.5 Influence of environmental factors on acceptance
5.6 Awareness and acceptance of ETF savings plans
5.7 Expectations related to the use of ETF savings plans
5.8 Summary of the results

6 Discussion
6.1 Original TAM variables: PU & PEOU
6.2 Added extended TAM variables: RI & FI
6.3 Impact of environmental factors
6.4 Theoretical implications
6.5 Summary, outlook, and limitations

7 Policy recommendations
7.1 Merger of financial knowledge and presented information
7.2 Germany fund as a mandatory pension scheme

8 Conclusion

9 References

10 List of figures

11 List of tables

12 List of abbreviations

A Results of the content literature review

B Results of the methodical literature review

C Translated questionnaire

Acknowledgements

Now that I have reached the summit of my academic journey, I would like to take the opportunity to thank chronologically the people who made it possible for me to get this far.

Thanks to my parents and grandparents who have supported and encouraged me in every step of my life. Thanks for your patience and trust. Thanks also to Anna, my girlfriend, for always having my back and giving me support. You have shown me how to climb, so that I could start my journey. I love you.

During my school career, especially 2 people have positively influenced my devel­opment and have therefore remained in my memory. For this reason I would like to thank my former teachers Mrs. Palmert and Mrs. Soergel, who always took me as I was and supported and encouraged me despite my sometimes special personality. You have given me the right tools for my journey. Thank you.

During my academic career I could always rely on one of my best friends, who had the power to motivate and push me even in stressful times. Without you, Lennart, this whole journey wouldn't have been as enjoyable as it was. You gave me a helping hand when I needed it. Thank you.

Finally, I would like to thank my supervisor Mr. Blizkovsky who helped me on the most difficult ascent of my academic journey with his experience and advice. You helped me cross the finish line. Thank you.

Abstract

Future German pensioners will face a pension gap in retirement age if they base their income exclusively on the statutory pension. Conventional private pension al­ternatives are not lucrative, but they are still the most widely used. Exchange-traded funds (ETFs) have features that make them particularly attractive for pension pro­vision. The aim of this dissertation is to discover which factors influence the ac­ceptance formation process and thus the usage of ETFs. Therefore, a model is devel­oped that is based conceptually on the widely known technology acceptance model (TAM) and theoretically on the results of a comprehensive literature analysis, which summarises already identified obstacles within old-age provision. This model is then tested using an online survey on future German pensioners aged between 19 and 59 years (n=615). The descriptive statistics are then calculated for the recorded data, and subsequently, the relationships within the variable model are analysed for significant dependencies. The results show that the assumed external variables have an influence on the acceptance of ETFs. In particular, the variable of financial literacy as an indicator of a person's general financial knowledge turns out to be of overrid­ing importance within the acceptance formation process. Furthermore, the results show that the use of an ETF is well recognised, but that it is still too complex to ob­tain information or acquire an ETF, which ultimately hampers acceptance for­mation, and thus the use. Finally, two concepts are developed that would have a pos­itive influence on acceptance in different ways; they focus on different variables and behavioural patterns that still act as obstacles, thereby increasing the overall use of ETFs in Germany.

Keywords: Exchange-traded funds, retirement planning, pension provision, be­havioural finance, acceptance, technology acceptance model (TAM)

1 Introduction

For some time now, the term ‘pension gap' has been circulating in German media. The main issue is not the security of pensions in Germany but rather their amount. The most recently published pension atlas, Germany 2017 by Union Investment, shows how bad the situation is regarding the old-age provision, especially for nor­mal employees of the younger generation (that are, people aged between 20 and 34 years). The average replacement rate for this group of people, provided by the first layer of the German old-age provision system, is just 38.6%. This means that they only receive 38.6% of their last gross income as a lifelong pension from the statutory pension insurance. Just how drastic the collapse of this replacement rate is can be seen from a comparison of the average replacement rates achieved for people aged between 35 and 49 years and those between 50 and 65 years, which are 43.2% and 64.1%, respectively (Raffelhüschen, Metzger & Seuffert, 2017). In addition, as a re­sult of the Retirement Income Act of 2005, the annual tax rate on retirement benefits earned through state pension provision will continue to rise until it reaches the max­imum rate of 100% in 2040 (Bundesregierung Deutschland, 2004). These facts force future pensioners to confront an ever-widening pension gap, which they must try to close through early private pension measures. This means that the old-age provision is partly privatised and is the responsibility of each individual. It is therefore clear that the standard of living of this younger generation will be exposed to major changes if these people do not make any further private provision for their old age. Therefore, the results of a 2017 survey on the interest of the German population in private old-age provision and financial security for the future are highly alarming; around 84% of Germans aged over 14 years stated they had only moderate to no interest at all in private old-age provision (IfD Allensbach, 2017). An additional problem is which pension provision people today are most likely choose. Currently, the most used alternatives are unprofitable because their profitability is linked to the guaranteed interest rate determined by the German Federal Ministry of Finance (Statista, 2019a). This rate has been set at 0.9% since 2017 (Kahlenberg, 2018). This is a huge downside since interest can massively influence the total amount of long­term investments, and can therefore help to close the pension gap. Figure 1 presents the most used pension provision alternatives among the German people at present.

However, a current study shows that more than 50% of Germans do not provide for their retirement despite knowing about the pension gap (WeltSparen, 2019a).

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: Allocation of the investment alternatives used by Germans in 2019. Source: Statista, 2019a

The chart in Figure 1 confirms that conservative pension options are still the most frequently used, even though they generate virtually no returns. The fact that funds are already above the classic Riester pension is surprising. In general, this suggests an increasing rejection of the Riester pension and an increasing interest in funds, although shares and bonds continue to be unpopular. In general, however, these sta­tistics reflect the problem already described. A large proportion of the German pop­ulation does not provide for their old age at all, and if they do it is only through low- yield instruments.

2 Research question and objectives

The question this thesis seeks to answer is not whether younger generations should invest in private old-age provision, but rather how they should do so. The steadily falling guaranteed interest rate is causing an increasing loss of profitability for the conventional old-age provision alternatives. For this reason, increasing attention has been paid in recent years to investing in exchange-traded funds (ETFs). The in­creasing range of savings plans in this asset class confirms this growing interest. In 2017, the industry even reported a new boom year (ETF Magazin, 2018). However, funds and other financial products are still not very popular as old-age provision among the German population despite some attractive features for retirement pro­vision. Therefore, the main objective of this thesis is to identify reasons why ETF savings plans are not often used as old-age provision among the German people. To do this, how well known this pension alternative is must be clarified. If awareness is eliminated as the main reason against the use of ETF savings plans, the acceptance of this asset class needs to be measured in the retirement planning framework of Germans. Therefore, an integrated and comprehensive acceptance measurement model must be developed, which can also be used to measure the acceptance of other financial products. To answer the main research question, three research subquestions have been formulated:

Question 1 : Which factors have an influence on the acceptance formation process and thus on the usage of ETF savings plans products in the context of private old-age pro­vision?

Question 2 : What is the current level of awareness and usage of ETF savings plans and how high is their acceptance within the framework of private old-age provision in Ger­many?

Question 3 : What would consumers want regarding ETF savings plans as an alterna­tive to old-age provision to consider them as part of their old-age provision?

In the following paragraphs, the procedure of this dissertation is described to an­swer the defined research questions. The following figure illustrates the whole pro­cess from identifying the research gap and then to answering the research question to provide a scientific contribution. This is done to provide an overview of the con­tent this dissertation covers and which methods are used to archive the aforemen­tioned objectives.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2: Structural approach of this thesis. Source: The author.

With regard to the objectives, this dissertation answers the research question in a four-step approach, which combines two different methodological research ap­proaches. The introduction of this work is intended to make a significant contribu­tion to providing a comprehensive overview of the contents dealt with. The starting position is the derivation of the problem and the question that the entire work re­volves around. These are intended to provide an indication of the topicality and rel­evance of the subject matter dealt with. In addition, this part describes which re­search question is derived from the problem and consequently investigated.

Subsequently, it is necessary to create a common understanding of the research sub­ject and the relevant terminology; therefore, the theoretical and conceptual founda­tions of the German pension system and ETFs are explained. Chapter 3 initially takes a critical look at the existing literature on the German pension system to explain how this system works and to theoretically derive and explain the aforementioned pension gap. Subsequently, the products of the private old-age provision are ana­lysed according to their exact function to determine whether they are able to reduce or even close this provision gap theoretically. The selection of the alternatives pre­sented is limited to capital-based old-age provision products of the old classic. This is mainly justified by the fact that these products are currently among the most fre­quently used for private old-age provision. In addition, their characteristics and functions make them easier to compare with an investment in ETF savings plans because some of them are highly similar. For example, it would be difficult to com­pare an ETF savings plan with a property as a means of old-age provision in the cat­egories of return and risk, because these key figures are difficult to generalise for a property. The advantages and disadvantages of these products must be identified in the course of the presentation of the pension alternatives to compare them later with the advantages and disadvantages of an ETF savings plan product. However, before this comparison can be made, the ETF savings plan products must also be analysed and examined. In the final comparison, the selected pension products are then compared in terms of critical factors such as return on investment, capital risk, capital flexibility, and investment transparency. Due to the complexity, scope, and largely nonexistent generalizability of the comparison, tax considerations are not included in the comparison.

After this comparison is performed, the relevant research from the areas of behav­ioural finance, old-age provision, and ETFs is structured and analysed to reveal in­terrelations between these topic areas. This step is intended to help identify the state of the art regarding Germans' current perception of financial products as old- age provision, which is one of the most critical objectives of this study. Therefore, a literature review based on peer-reviewed journals is conducted. Both steps help to develop an acceptance measurement model, which is then used in the actual empir­ical investigation; however, before that, some of the basics of acceptance research must be stated. This is a necessary step to derive the framework model that is used to measure the acceptance of ETF savings plans among respondents and to answer the main research question of this paper. Therefore, the focus is on Davis' technol­ogy acceptance model (TAM).

Chapter 4 marks the start of the empirical part of this dissertation, which intends to use the findings from the theoretical part to derive hypotheses to help answer the research question. These hypotheses mainly refer to the impact of the variables, which are more precisely defined in the framework model of this research.

Subsequently, Chapter 5 describes and justifies the chosen methodology and re­search design in detail. Within this scope, the structure of the questionnaire used and the implementation of the quantitative survey are also examined. Finally, the main methods used to evaluate the collected data are presented. The aim of this chapter is to provide complete transparency regarding the approach of this re­search.

Chapter 5 also deals with the analysis of the collected data using descriptive and analytical statistics. The aim is to test the hypotheses previously made and, if neces­sary, confirm or falsify them.

The most crucial part of this thesis is dedicated to the interpretation of the results and the final response to the research question, which is done in chapter 6. In addi­tion, the theoretical implications are discussed. This chapter concludes with a criti­cal appraisal of the research and an outlook for potential future investigations. As a result, the findings of this dissertation should provide guidelines for researchers and practitioners in future research and the implementation of ETF savings plans as a common old-age provision alternative. Based on the findings of this chapter, con­crete policy recommendations for action are made in chapter 7.

Finally, Chapter 8 concludes by summarising the most important core statements of the entire dissertation.

3 Literature review

At the very beginning of a scientific research project, it is essential to first examine the necessary theoretical background regarding the chosen topic and then analyse and reflect on the current state of research to identify the research gap in depth. This chapter is aimed precisely at these two tasks. For this reason, all principles of the German pension system are explained and the pension gap is derived theoretically. In a second step, the functions of ETFs are analysed in more detail and applied to the field of pension provision to highlight their benefits. Subsequently, a systematic lit­erature analysis is performed to assess the current state of research regarding pos­sible factors influencing the decision making of future pensioners. The final aim of this chapter is to develop and derive of an acceptance measurement model for ETF savings plans, in order to be able to answer the research questions posed in the fol­lowing empirical part.

3.1 The German pension system

As previously mentioned, a partial goal of this study is to theoretically prove the emergence of the pension gap in old age to first fundamentally grasp the problem with regard to the research question. For this reason, the functioning of the current pension system in the Federal Republic of Germany is examined and analysed. This is not intended to deal with every facet of the sometimes highly complicated and extensive overall system, but rather to focus on the essential components that are indispensable for deriving the problem. In addition, only selected private old-age provision products that can be compared with the investment form of the ETF sav­ings plans in terms of structure and other criteria are analysed in more detail. The company pension scheme is not considered further in the following section because the contributor ultimately has no influence on the final use of contributions. Since 2002, an employee has had the right to claim a company pension plan in the form of deferred compensation without the employer's consent, but the employer has sole decision-making power over the use of contributions (Clemens & Förstemann, 2015). This would become an indefinable variable in the course of the later compar­ison of the various alternatives, and thus make objective consideration impossible. The so-called three-layer model of old-age provision forms the basis of the current old-age provision system in the Federal Republic of Germany. The model is essentially based on the older three-pillar model of old-age provision, and thus as­sumes that the income of a retired person comes from three different sources (Mierzejewksi, 2015). Therefore, this is not necessarily a substitution for the mod­els, but rather an extension, which is largely based on the tax law-oriented Retire­ment Income Act introduced in 2005 (Ilg, 2010). Since its enactment, this law has regulated the tax treatment of all pension expenses and the later retirement pay­ments that follow (Sachverständigenkommission zur Neuordnung der steuerrecht­lichen Behandlung von Altersvorsorgeaufwendungen und Altersbezügen, 2003). In this model, all investment alternatives for old-age provision are differentiated ex­clusively according to their degree of use for old-age provision, and are thus allo­cated to the individual layers. The principle applies that the more a product contrib­utes exclusively to maintaining lifelong income security, the more likely it is that it will receive tax incentives. For this reason, the tax incentive for investment alterna­tives decreases from the first layer to the third layer, in which no state subsidy is granted (Myßen, 2006).

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: Three-layer model of the German pension system.

Source: Vermittlungsgesellschaft für Verbraucherverträge AG, 2018.

The first layer comprises all forms of old-age provision whose entitlement is neither eligible as collateral nor saleable or capitalisable. Furthermore, they are not inher­itable or transferable. The pension is paid to the beneficiaries on a monthly basis at the age of 60 years at the earliest. Unscheduled capital withdrawal is categorically excluded. These restrictions ensure that the assets saved can only be used for old- age provision and simultaneously exclude any economic use. In addition, forms of old-age provision of this stratum are tax-privileged in their savings phase. This layer includes, among other things, the classic statutory pension insurance as well as the basic pension or Rürup pension.

The second layer, in addition to the company pension scheme, includes the certified and state-subsidised old-age pension scheme, also known as the Riester pension. The conditions for these investment products to belong to the second tier of the sys­tem are not as restrictive as in the first tier. For example, although the pension in­vestment must still grant a lifelong pension payment, the possibility exists of receiv­ing a certain capital amount in the form of a one-off payment.

In principle, the third layer comprises all investment products that do not meet the requirements of the first two layers, but which nevertheless pursue the objective of accumulating assets for old-age provision at the beginning of their use. In reality, the final use of the saved assets can deviate from the original provision because they are not earmarked. This corresponds to a large difference from products of the preced­ing strata because they can be used exclusively for old-age provision. For this rea­son, third-tier investment products have the characteristics of a freely available cap­ital investment. Typically, this class includes endowment life insurance, private pen­sion insurance, real estate, bank savings plans, and securities (Sachverständigen­kommission zur Neuordnung der steuerrechtlichen Behandlung von Al­tersvorsorgeaufwendungen und Altersbezügen, 2003). Consequently, the ETF sav­ings plans presented later also belong to this stratum.

The statutory pension insurance, as part of the first layer, forms the basis of the Ger­man pension system and is therefore explained separately. It is a compulsory insur­ance by law that initially applies to all persons who are either gainfully employed for a wage that goes beyond the remuneration of minor employment, or who are employed for their own vocational training. Notably, compulsory insurance is linked to the employment relationship. This relationship essentially reflects the em­ployee's individual dependence on the employer, which results from integration into the employer's work organisation; thus, it is directly linked to the employer's authority to issue instructions. This also includes self-employed persons who are similar to employees. All other persons who are not covered by the statutory direc­tives can nevertheless voluntarily pay into the statutory pension insurance scheme.

The object of this insurance is the lifelong payment of a pension to the contributor, which usually begins when the statutory retirement age is reached. The financing of this system is largely based on contributions paid by insured persons, employers, and state subsidies. In the course of the wage deduction procedure, the employees retain an employee share of 50% of the contribution and then pay the full amount to the statutory pension insurance. In fact, the financing of this pension system is ensured by the current national income. This financing method is therefore called an allocation procedure. In principle, the retirement pension of a retiring contribu­tor is realised by the preceding generations of contributors (Mierzejewksi, 2015). For this reason, an obligation for this form of insurance is indispensable for gener­ating a continuous flow of contributors and guaranteeing security for pension pay­ments (Bäcker, 2016). Theoretically this is a waterproof system, but the reality shows that this is not the truth. The result is the already described pension gap that future German pensioners now face.

The origin of this pension gap can essentially be explained by a combination of two factors, namely the type of financing of the statutory pension insurance and the de­mographic development of the German population, which have a negative influence on each other. As previously described, the pension amounts are financed within the framework of the allocation procedure by the people's ongoing income. Thus, a di­rect quantitative relationship exists between contributors who pay into the statu­tory pension insurance and pensioners who receive individual contributions. The size of the working age population is therefore ultimately decisive for financing be­cause it reflects the number of potential contributors (Fenge & Peglow, 2018). The latest forecast of the Federal Statistical Office estimates that the number of these persons will fall by approximately 8.5% by 2030 and by as much as 30% by 2060; therefore, the aforementioned ratio will deteriorate. In addition, the ratio will be worsened by the rising life expectancy of Germans because the retirement age will be extended simultaneously (Statistisches Bundesamt, 2015). Increasingly fewer contributors are now facing an increasing number of beneficiaries, which represents an enormous challenge for the statutory pension insurance. To continue financing this insurance through the allocation procedure, it was necessary to continually lower the statutory pension level per person to distribute the capital to larger parts of the population. This step was implemented within the framework of the 2001 pension reform. At the same time, a voluntary private funded pension insurance scheme was introduced with the Riester pension. This was intended to compensate for the falling statutory pension level (Börsch-Supan & Gasche, 2010). In general, the pension gap is understood to mean a financial gap that arises as a result of the change in income caused by retirement, which thus leads to a negative change in living standards (Mierzejewksi, 2015). Terminology also used in this context is the replacement rate of the last gross income; this merely reflects what proportion of the last gross income can be compensated by the individual retirement provision of all three layers (Keller, 2012). A rate of 60% is enough to maintain one's standard of living in old age, since cost structures are also changing significantly. If one com­pares this value with the replacement rates, which can only be achieved with the help of statutory pension insurance, the current groups of people between the ages of 20 and 49 years only have an average replacement rate of 39.9% (Union Invest­ment, 2017).

Now, a subobjective of this work is to determine which retirement provision prod­uct is best suited to achieving these approximately 20.1 missing percentage points of the replacement rate required to maintain the standard of living as constant even at retirement age. For this reason, some selected capital-based private pension al­ternatives are analysed below with regard to their functioning and advantages or disadvantages to make them comparable. Moreover, the alternatives are examined with regard to the following criteria: capital flexibility, capital yield, capital safety, investment costs, and investment transparency. It should be mentioned that not all of the selected retirement provision products belong to the private retirement pro­vision according to the definition of the three-layer model. This section only catego­rises whether retirement provision is legally mandatory. In addition, all the follow­ing alternatives have four characteristics in common. First, the occupancy of these pension assets is not mandatory, but rather voluntary. Second, they are financed by the capital cover procedure. In this procedure, the contribution payments of the in­sured person are invested on the capital market (Siebert, 1997). At the time of re­tirement, the insured person is then entitled to the accumulated capital and the in­terest earned with it. These interest rates essentially reflect the return on the pen­sion products, which is made up in detail of the annual interest rate and the final profit participation. The law currently guarantees a nominal return of at least 0.9% p.a. on the capital saved over the entire term of the contract (contracts starting from 2017). However, the maximum current rate of interest ultimately achieved depends to a large extent on the net interest income participation of the providers, which they determine annually (Kahlenberg, 2018). The term ‘surplus participation' obliges the provider to allow the policyholder to participate in the income generated by the premiums. Cost savings on the side of the insurance company also contribute to an increase in profit participation (§ 153 Gesetz über den Versicherungsvertrag, 2013). The steady decline in the guaranteed interest rate is mainly caused by the current low interest rate environment, which is a logical consequence of the political decisions of the European Central Bank. For this reason, insurance companies are also being forced to adjust the interest rates of their products (Brühl & Walz, 2015). For 2018, the average current interest rate is 2.61% for all forms of old-age provi­sion dealt with in this paper (Assekurata, 2018). The final profit participation will not be included in the yield comparison of this work because it cannot be general­ised. The third characteristic is the fact that all later disbursements of almost all var­iants are taxed later on. Only the Rürup pension still benefits from a limited tax lia­bility. However, this advantage expires in 2040 because of the equal tax treatment of pension payments (Sachverständigenkommission zur Neuordnung der steuer­rechtlichen Behandlung von Altersvorsorgeaufwendungen und Altersbezügen, 2003). Last but not least, since 2008 all of the following forms of provision have been required to publish a product information sheet, which provides the insured with information on the type of insurance, risks, premiums, duration, benefits to be ex­pected, and costs (§ 4 Abs. 2 Gesetz über den Versicherungsvertrag, 2013). This en­sures a certain degree of transparency for these products, which makes them more predictable and comparable to a certain extent.

The Riester pension is the official alternative of the state to close the pension gap created by the pension reform in 2001. Currently, four different types of Riester pen­sion exist, namely classic life insurance, the fund savings plan, the bank savings plan, and the residential Riester. To make the acquisition of a Riester policy particularly attractive, this old-age provision alternative is supported by state allowances and special tax deductions. To receive the full allowance subsidy, the insured person must continuously save at least 4% of his or her annual gross income (maximum of 2100 euros) minus individual allowances for example, basic and child allowance and career starter bonus (Rüb & Lamping, 2010). In the context of the tax special ex­penditure deduction, the paid Riester contributions can be marked in the annual tax return as special expenditure and thus obtain a tax advantage. If this tax advantage then exceeds the state paid allowances, the insured person receives a tax refund of the difference. If this is not the case and the tax advantage is lower than the amount of the allowances, then this has no effect on the insured person. Irrespective of the type of Riester contract, which is usually entered into with insurance companies, the contractual arrangement must essentially correspond to the already described re­quirements of a first-tier old-age provision to be considered a certified subsidy con­tract (§ 5 Gesetz über die Zertifizierung von Altersvorsorge- und Basisrenten­verträge, 2017). Only a withdrawal of capital amounting to 30% of the accumulated assets at the end of the term constitutes a difference from the abovementioned cri­teria. In addition, since June 2013, providers have been required to disclose full transparency about their products, which also shows what costs the contract entails and what return can be expected (§ 7 (§ 5 Gesetz über die Zertifizierung von Al­tersvorsorge- und Basisrentenverträge, 2017, 2017). The Riester pension is classi­fied as a safe form of investment because the payment of the saved premiums and allowances must be guaranteed by the insurer. Furthermore, the saved assets re­main untouchable even in the case of private insolvency or joblessness, but only if the monthly contributions have not exceeded the maximum contribution. Overpay­ments are therefore not covered. In addition to the abovementioned advantages, some disadvantages must be highlighted. First, the hereditability is usually strongly limited; only the spouse would have a claim on the complete Riester fortune in the case of death of the insured. All other surviving dependants would have to repay the paid allowances and saved taxes to the state. Furthermore, the flexibility of this age pension form is strongly limited because capital withdrawals are not possible before the 60th year of age without cancelling the contract, which would lead again to the repayment obligation of the received promotions (Benölken, Bröhl & Blütchen, 2011). In addition, although a change of provider is possible within the framework of capital flexibility, this in turn entails costs, which ultimately have a negative effect on the net return (Reichert, 2013).

The basic or classic Rürup pension offers a further possibility to build up a private and state-subsidised old-age pension with its affiliation to the first old-age pension layer of the three-layer-model (Sachverständigenkommission zur Neuordnung der steuerrechtlichen Behandlung von Altersvorsorgeaufwendungen und Alters­bezügen, 2003). Compared with the Riester pension, the subsidy for this provision is limited to the tax advantage. However, with a deductible annual maximum amount of 23,712 €, this subsidy is many times higher. There is no mandatory pay­ment amount, as prescribed by the Riester pension, but a maximum of 23,712 € can be paid in, of which 86% (20,393 €) can be claimed for tax purposes as other ex­penses (status 2018). In 2025, the maximum contributions will be 100% deductible, which represents a clear advantage of this form of old-age provision in the future. In the disbursement phase, this form of provision behaves exactly like the statutory pension, which means that a one-time capital pay-out is not possible. In case of the death of the insured person, no survivor compensation is guaranteed in principle. However, depending on the provider, it is possible to contractually carry out this compensation elsewhere but, as a consequence, the costs are higher and the net yield is lower. Basic pension contracts cannot be cancelled, and thus there is basi­cally no right of access to the saved capital, which practically eliminates the capital flexibility of this form of provision. Even a change of provider in this pension form is tolerated by only a few providers, which also has a negative effect on the capital flexibility. Only the contribution payments can be flexible in their amount, and an exemption from contributions is also possible. The security level of this pension al­ternative corresponds exactly to that of the Riester pension because both products have the same characteristics in this category (Ilg, 2010).

Life insurance policies are normally designed to provide financial protection for sur­viving dependants only in the event of the insured's death. The capital life insurance is the only variant which, in addition to the surviving dependants' insurance, can also pay the saved capital to the insured person at the end of the contract, and is therefore suitable for old-age provision. At this time, the insured person is entitled either to a one-off or a continuous payment of the accumulated capital, which is in­creased during the term with a guaranteed interest rate, plus a profit participation. This profit is generated by the insurance company with the premiums paid. To sign a capital life insurance policy, information about the applicant 's state of health must be provided in advance. The capital life insurance contract can be cancelled at any time, but the insured only receives the repurchase value of his or her paid-in capital, which is equivalent to a capital loss (Bundesanstalt für Finanzdienstleistung­saufsicht, 2016a). It should also be noted that only the savings element of the paid- in contribution is used to build up assets, and thus bears interest. The remaining portion of the premium is allocated to the risk share to build up the agreed sum in the event of the death of the insured and to the cost share. Conversely, this means an even lower expected return on this investment product. The amount of the pre­mium paid is determined individually by age, the desired term, the sum insured, and the state of health. Therefore, the insured person only has a partial influence on the premium to be paid. Once the contract has been signed, it is no longer possible to adjust the premiums without further delay. This is usually associated with costs that have a negative impact on the return and reduce the flexibility of this asset class. German life insurers are subject to state supervision by the Federal Financial Super­visory Authority (BaFin) and are therefore confronted with strict investment rules. The investments that an insurance company carries out with the premiums are also subject to strict rules with regard to their risk liability. However, should an insur­ance company experience payment difficulties, a state institution will take over the contracts and claims of the policyholders and continue them under the same condi­tions. The capital created is therefore safe in this form of investment, assuming that the insured person experiences the end of the contract. This fact has an immense effect on the flexibility of the capital, because it is only possible for the insured per­son to withdraw capital prematurely at a loss (Bund der Versicherten, 2018).

The classic private pension insurance is highly similar in form to a capital life insur­ance, but some differences must also be mentioned. For instance, the insurance ben­efit includes a lifelong pension payment and thus covers the longevity risk. Further­more, coverage of the survivors is not the standard of this type of insurance but can be added by contract in individual cases. In general, survivors' insurance always leads to higher costs, which is why it can be generalised that a private pension in­surance in its original form exhibits a higher yield compared with a capital life in­surance. When a policy is signed, the insured person has a capital option and can determine whether the saved capital is to be paid out at once or in continuous life­long pension payments. It should be noted, however, that the private pension insur­ance cannot be terminated after retirement and therefore there is no more access to capital. Furthermore, within this investment form, the paid contribution does not go to 100% into the savings portion of the insurance because costs and commissions are subtracted beforehand (Lindmayer, 2008). The contributions can be paid in each case monthly or annually, which gives, in addition, the possibility of a one-time pay­ment. This variant is particularly interesting for people who are close to retirement age. In contrast to capital life insurance, the insured person can completely deter­mine how high these contributions should be arbitrarily. The amount of the contri­bution thus also determines the amount of the final pension (Bundesanstalt für Fi­nanzdienstleistungsaufsicht, 2016b).

3.2 Exchange-traded funds

This section deals with ETFs in detail to comprehensively review this asset class and then make a comparison with the aforementioned private pension alternatives. The cornerstone of this investment class was laid at the end of the 1980s by Nathan Most and Steven Bloom, who further developed the classic index funds. The first ETF was traded on the American Stock Exchange in 1993 and tracked the S&P 500 index. In subsequent years, other ETFs followed, such as the Dow Jones Industrial Average or the NASDAQ100. In Europe, an ETF was first traded on the stock exchange in 2000, which represented the Euro STOXX 50 (Tarassov, 2016). In 2005, the first legal reg­ulations were introduced at the European level, which influenced the trading of ETFs. The Undertakings for Collective Investment in Transferable Securities (UCITS) guidelines are still valid today and have been continuously further developed (Deutsche Börse Group, 2018). The ETF market has always been growing steadily and even recorded a new boom in 2017 (ETF Magazin, 2018). By 2016, there were over 4500 ETFs administered worldwide (Deutsche Bank, 2017a). In total, more than US$ 3.4 trillion are invested in this asset class, and the trend is rising (Deutsche Bank, 2017b). As a result, this study now analyses the functioning and characteris­tics of this type of fund. The aim of this section is to highlight the suitability of ETFs for private pension provision. In this context, two of the most important financial theories are explained. A comparison between traditional private pension alterna­tives and ETF savings plans concludes this section. This will help to identify ad­vantages or disadvantages of building old-age assets through ETFs.

An ETF is a special type of investment fund that usually attempts to replicate an in­dex as closely as possible and can be traded on the stock exchange. In an investment fund, the capital of many investors is bundled and then invested in securities by the investment company. Legally the fund assets are considered a special fortune and therefore must be strictly separated from the assets of the fund company. This re­sults in the crucial advantage of this fortune not being exposed to the emitter risk, and thus in case of insolvency of the fund company it is not sizeable (Niedermayer & Wagner, 2012). An index comprises the performance of an entire submarket of secular financial activity, which essentially represents a grouping of various securi­ties. Usually, the securities are grouped according to regions, countries, sectors, strategies, asset classes, or the most important companies of an economic area, and their joint performance development is represented by an index. Thus, the index is a key figure that determines the overall condition of a market. Furthermore, it is a benchmark for traditional mutual funds (Lettau & Madhavan, 2018).

An ETF attempts to copy the composition of an index. This can be achieved through three different methods: full replication, sampling, or synthetic replication. The first two methods can be summarised under physical replication. With full replication, the composition and weighting of the securities in the ETF are identical to those of the underlying index. This method ensures the highest transparency because the ex­act composition can be taken from the index, which is easily accessible information (Niedermayer & Wagner, 2012). As the number of different securities in an index increases, complete replication becomes increasingly complex, which in turn leads to rising transaction costs. The sampling method attempts to reduce these costs by replicating only parts of the underlying index. For reasons of cost and complexity reduction, securities are not used that either have little influence on the perfor­mance of the index because of their minimal weighting within the index or that are particularly difficult to trade. To keep the difference in value development between the index and the ETF resulting from this replication method as small as possible, the elementary characteristics of the index are retained in the ETF.

Synthetic replication is characterised by the fact that parts of the index representa­tion are realised by swaps. Swaps are exchange agreements on future cash flows between at least two parties. This swap transaction usually occurs between the ETF and a financial institution (the swap counterparty). The swap counterparty is com­mitted to maintaining the return on the index represented by the ETF. In return, the counterparty receives a swap fee and the yield of the securities in the security port­folio. This exchange transaction gives rise to the so-called counterparty risk, which physically replicated ETFs do not have because they actually acquire the securities shares. This risk describes the default of the swap counterparty due to insolvency (Kosev & Williams, 2011). Under the legislation of the UCITS, this risk may not ex­ceed 10% of the invested assets, and thus the collateral amounts to 90%. The type of synthetic replication differs depending on where these securities are located, whether in the ETF itself or with an independent trustee. For this reason, within synthetic replication, a decision is made between an unfunded swap and a funded swap (Niedermayer & Wagner, 2012). In practice, some measures are taken to re­duce counterparty risk. For example, several different counterparties are used to minimise this risk (Johnson et al., 2012). Besides this risk, there is also an advantage to this replication method. It makes it possible to invest in indices that cannot be replicated physically or can only be replicated at high cost (Kosev & Williams, 2011).

ETFs are subject to the UCITS guidelines, which are implemented in the German Capital Investment Act. These guidelines are designed to protect investors and in­crease the efficiency of investment companies. Only if an investment fund fulfils these guidelines can it be sold throughout Europe without further delay. In addition, these funds can be traded in Asia, Latin America, and the Middle East. For an ETF to comply with the guidelines, it must fulfil special requirements with regard to its li­quidity, diversification, security, transparency, and safekeeping of securities. An ETF must and can only be labelled a UCITS ETF if these conditions are met (ESMA, 2014).

Furthermore, ETFs can only be traded on the secondary market; that is, on the stock exchange. The process of issuing and redeeming ETF shares on the exchange is called the creation/redemption process and occurs on the primary market. In this process, the ETF company and the authorised market participants (market makers) exchange cash or baskets of securities for ETF shares. In detail, during the creation process, the cash or baskets of securities flow from the market maker to the ETF provider; that is, new ETF shares are placed on the market. The other way around, the flow of cash or baskets of securities behaves in the redemption process, as the provider buys back his ETF shares from the market maker. The market maker is therefore the authority that enables ETFs to be traded without restriction on the stock exchange. Moreover, the market maker is responsible for setting binding buy­ing and selling prices and ensuring sufficient liquidity for trading (Dieckmann, 2008). In summary, the high liquidity of ETFs is a major advantage of this asset class. On the one hand, this can be justified by the tradability on the secondary market and, on the other hand, by the aforementioned creation/redemption process (Osterhoff & Kaserer, 2016). The whole market process is presented in figure 4.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Market structures of exchange-traded funds (ETFs).

Source: Ramaswamy, 2011.

The price of an ETF unit is determined by the net asset value (NAV), which is calcu­lated once a day by the depositary bank. Therefore, all assets of the ETF are added and the liabilities are subtracted before being divided by the number of units in cir­culation (Angel, Broms & Gastineu, 2016). A second key figure, the indicative NAV (iNAV), enables continuous trading of an ETF unit at the NAV. This key figure is cal­culated every 15 seconds by the exchange and market makers, which is done based on the current market prices of the individual ETF components. If the price of an ETF share moves too far away from its NAV, then arbitrage traders make a correc­tion. These traders aim to take profits without risk by exploiting this price differ­ence. For this reason, the value of the iNAV is always very close to the value of the NAV (Kreis & Licht, 2018).

Like all other securities investments, ETFs are subject to a certain degree of risk, which must generally be distinguished into market and product risk. Since the ETF is directly linked to an index, its performance is correspondingly similar, and there­fore volatile. Which factors influence the market risk essentially depends on the composition of the respective index that an ETF tracks. For example, the perfor­mance of a DAX-ETF is largely determined by the share prices of the DAX30 compa­nies. Or if an ETF tracks a commodity index it is heavily dependent on the supply and demand of this commodity. An ETF based on bonds is significantly influenced by the development of key interest rates. However, other environmental factors such as a change in government strategy or legal changes, which are clearly assign­able to the respective market represented by the ETF, can also have a massive influ­ence on its development. This is due to the fact that the securities within the ETF are often strongly correlated with one another. Thus, this risk is called correlation risk (Chen, Liu & Hsu, 2016). Another risk, the currency risk, is one of the product-spe­cific risks of an ETF. If the home currency of the private investor differs from the currency of the companies represented by the ETF, then the exchange rate between the currencies has a direct influence on the value of the ETF. Some ETFs invest in indices that depict different currency regions, which directly increases the currency risk. To minimise this risk, various ETF providers offer exchange rate hedging, but this results in rising costs (Williams, 2016). Two further risks are caused by the structure of an ETF. ETFs usually have no maturity date and therefore run endlessly. However, it may occur that an ETF is not profitable for the issuing company and is therefore liquidated. Generally, the issuer takes back the issued shares at the price of the NAV. It is also possible that two similar ETFs that represent the same index merge. In this case, the assets will be transferred between the ETF providers. Usu­ally, the liquidation of an ETF does not result in further costs for the investor, but it may lead to legal structural changes, which may affect the investor (Picard & Braun, 2010). The second risk caused by the structure of an ETF is called security borrow­ing. Securities that are part of the ETF's total assets are provided to a lending coun­terparty for a fee. The ETF provider does this to generate additional income and re­duce costs. While borrowers must deposit collateral at all times to compensate for a default in the return of the borrowed security, this type of add-on business poses a risk that must ultimately be shouldered by the investor (Bioy & Rose, 2012). Two other product-specific risks, the counterparty risk and total default risk, were ex­plained previously.

ETFs are considered a particularly cost-effective investment alternative because they represent a passive investment strategy. As previously mentioned, ETFs at­tempt to track an underlying index as accurately as possible, and therefore do not require a fund manager to actively determine the selection of the fund's securities. This leads to extreme cost savings compared with an active managed fund (Com- Stage, 2016). The following paragraph explains the costs that must be considered in the context of an ETF investment.

The costs of an ETF essentially consist of two main cost blocks. The first block is described by the total expense ratio (TER), which is calculated in accordance with the regulations of the Federal Association for Investment and Asset Management (BVI). This percentage cost figure always refers to all costs incurred in the previous year on the average fund assets of the ETF. In addition to all fees arising, such as those for administration, licences, or distribution costs, this cost block also includes costs for annual financial statements or marketing activities (BVI, 2015). The second block includes the trading costs of an ETF, which are not included in the total cost ratio. For a physically replicated ETF this includes the costs of an index adjustment, whereas for a synthetically replicated ETF this includes the swap fees. In this context it should be mentioned that not all swaps lead to an increase in costs. However, this may happen when certain securities are extremely difficult to access (BlackRock Inc., 2011). Furthermore, the second cost block includes the bid/ask spread. This cost component increases equivalently with the intensity of the investor's trading activities. The difference between the money and the ask price benefits the market maker, who uses it to cover his or her costs. These costs include transaction costs, hedging costs, management costs, and costs generated during the creation/redemp- tion process. How high these costs are depends significantly on two factors. The first is the liquidity of the index which is replicated through an ETF. Therefore, a higher liquidity results in decreasing costs. The second factor is the market volatility; a lower volatility level especially reduces the costs caused by the hedging process done by the market maker. Ultimately, this cost block also includes all stock ex­change fees (Angel, Broms & Gastineu, 2016). The so-called tracking difference is used to determine all the costs of an ETF. It compares the return achieved by the ETF within one year with the return achieved by the underlying index. The resulting negative deviation must be justified by the costs that effectively reduce the return of the ETF (Johnson et al., 2013).

The good transparency of an ETF investment is caused by the structure of this asset class because it attempts to track an index as accurately as possible. Therefore, the composition of an index is completely accessible to the public, which already con­tributes to the transparency of the ETFs. The disclosure obligation regulated in the UCITS Directives also contributes to an increase in transparency, which is mainly through the sales prospectus, the Key Investor Information document, and the an­nual and half-year reports. The sales prospectus must specify which index the ETF tracks and which underlying assets form its basis. Furthermore, which replication method the ETF uses to model the index must be defined. All influencing factors that affect the tracking difference must be named. Finally, there must be specific warn­ings about securities trading and its risks. The sales prospectus is the most detailed version in terms of information provided (Verordnung über Vermögensanlagen­Verkaufsprospekt, 2004). The Key Investor Information document only summarises the essential information. Thus, only the objectives and the investment policy, risk and return profile, costs, and past performance of the fund are often provided (Ow­ens & Senior, 2011). The annual and semi-annual reports clarify the following key figures: the ETF asset position and development, the income and expense statement, the tracking difference, and the tracking error (ESMA, 2014). The tracking error shows the standard deviation of the index return and the ETF return (Johnson et al., 2013). At the beginning of 2018, the Investment Tax Act MiFID II was introduced, which was also intended to contribute to transparency. Brokers must now inform the investor of all costs incurred before buying a security (Deutsche Börse Group, 2018).

Some ETFs accept a deposit via a savings plan, which allows the investor to regularly invest a constant amount of money over a long period of time. With the chosen pre­mium, the custodian then purchases ETF shares on the market. Only for this reason is it possible to purchase fractions of an ETF, which is normally not possible through a direct buy on the stock exchange (Götte, 2010). Therefore, the investor can benefit from the cost-average effect due to the constant contributions. Because of the con­stant saving amount, the number of ETF shares purchased varies depending on their price. This means that fewer units can be purchased at a high price level and accord­ingly more units at a low price level. In the long term, this effect leads to a lower average purchase price of the acquired shares. Furthermore, any timing effects are eliminated because the investment is always made automatically at the same time, usually on the first or last day of the month. This can prevent a positive or negative timing effect depending on the price level of the ETF unit (Thiery, 2010). The savings rates of an ETF savings plan can be suspended or changed at any time. A minimum contribution is set by the custodian bank but is usually at least 25 € per month. This also enables people with a low income to participate in the exchange market. More­over, the saved capital can be paid out quickly or flexibly reinvested in other securi­ties investments. Some providers charge an additional fee for an ETF savings plan, which has a negative impact on its returns. Since these costs differ between the de­posit banks, the choice of a low-cost custodian is essential for a long-term invest­ment of this kind, because the costs have an enormous impact on the return.

3.2.1 Suitability as a private pension alternative

Since the asset class of ETFs has now been extensively analysed, it is now possible to verify whether this special form of capital investment is suitable for private re­tirement provision before making a comparison with the other pension alternatives. To make this assessment, two elementary financial market theories must be consid­ered, and are therefore explained first to help to answer this question. This will in­clude the efficiency market theory of Eugen Fama and the portfolio theory of Harry M. Markowitz. Both theories have been awarded the Nobel Prize and therefore should help to legitimate an investment in an ETF savings plan as a private old-age provision as well as highlight its advantages.

Within the framework of his theory, Fama dedicates himself to the question of whether it is even possible at all to gain a higher return than the full efficient total market. The term ‘efficiency' plays a key role in this theory, which Fama (1970) de­scribes as follows: ‘A market in which prices always “fully reflect” available infor­mation is called “efficient”' (p. 383). Today, information spreads all over the world within seconds, which means that it can be analysed and used almost in real time.

This, in turn, means that all security prices must already take this information into account and thus be fairly valued in relation to their risk. Fama divides this infor­mation efficiency into three strengths, with the strongest form always including the weaker forms.

The weakest efficiency level assumes that security prices only consider all historical information and are not related to future price changes. The medium efficiency level goes a little further and assumes that all publicly accessible information such as an­nual financial statements or press reports is already reflected in the share price. Fi­nally, the strongest information efficiency also includes nonpublic information such as insider knowledge or managerial skills. Thus, all different efficiency levels have a different influence on the intention to generate an overperformance. Over the course of time, the hypotheses have been tested repeatedly and partly refuted both empirically and theoretically. In summary, the overall markets are often - but not always - efficient, an thus it is possible for active funds to generate excess returns (Mehwish & Yasir bin, 2015). However, these excess returns must be at least as high as the fees charged by the fund manager for the management of the capital invest­ment, because otherwise the investor would receive a worse net return on his or her capital investment. In the context of the old-age provision, this is a further risk factor that can substantially affect the value of the pension. Statistically speaking, the ma­jority of actively managed funds do not beat the benchmark (Lyxor, 2017). This fact implies that ETFs are even more attractive and suitable for private retirement pro­vision.

Markowitz's portfolio theory examines the effects of risk, return, and correlations of different asset classes on the entire asset portfolio. It thus forms the basis for the question of the correct asset allocation. The correct diversification of the individual investments is intended to improve the efficiency of the overall portfolio. The aim is therefore either to generate more return with the same risk or to generate the same return with less risk. Markowitz divides risk into two categories: the first, system­atic risk, cannot be influenced by an organisation, such as natural disasters, political changes, or economic crises. This also means that this risk always exists and cannot be diversified. However, the second category, unsystematic risks, can be reduced by diversification because they are directly influenced by an organisation's decisions and strategies. When a portfolio is classified as efficient depends on the risk toler­ance of the investor. In this context, Markowitz also mentions the fact that investors perceive volatility in yield development as a risk. An efficient portfolio is thus ulti­mately constructed in such a way that the maximum return opportunities can be taken advantage of, while the return development is constant, to reduce the volatil­ity. The volatility of returns is minimised by a negative or neutral correlation be­tween the securities included in the portfolio. Figure 5 presents the efficiency curve.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5: The efficient frontier of Markowitz Source: Zhang, 2017.

The more shares a portfolio has, the more the risk increases but also the return, and it moves northeast on the efficiency line. Only portfolios that are on this curve can be classified as efficient because there is no other constellation of the portfolio that promises more return for the same risk. To guarantee this, the portfolio must in­clude at least 20% shares and 80% bonds to be considered efficient (Markowitz, 1952). Above the curve there is an unattainable risk-return profile, and thus no port­folio can be there (Grujic, 2016). ETFs are already diversified by their structures and thus minimise the unsystematic risks, which contributes to the suitability of this as­set class for retirement provision. Diversification is usually much more expensive and requires advanced financial knowledge.

To assess the suitability of ETFs for private pension provision, it is first necessary to identify the objectives that people are pursuing as part of their pension strategy. The objectives most often identified in the literature are security, return, flexibility, and administrative burden. In addition, comprehensibility, transparency, and state sub­sidies play a role in decision-making on the desired pension alternative (Buhl, Frid­gen & Kaiser, 2003). If one considers state support as the sole selection criteria in advance, an ETF savings plan is not suitable for private old-age provision because there are no support options; therefore, this objective is no longer relevant.

The security of the gained savings is of overriding relevance in the context of old- age provision, as the sources of income significantly decrease at retirement. The re­maining sources must be able to secure the standard of living of the retiring person. In general, ETFs are a safe investment because they belong to the category of special fund and are therefore unseizable. However, ETF savings plans are subject to certain risks, as already explained. These risks cannot be completely eliminated, but they can be reduced to a minimum, which is also proven by portfolio theory. ETFs offer the optimal conditions because they are already diversified and thus reduce unsys­tematic risks. However, additional attention must be paid to the correlations be­tween the securities used within the index. Therefore, in terms of risk reduction, it is advisable to diversify the ETFs contained in a portfolio as well. For this reason, it is not recommended to acquire several ETFs at the same time that partially or com­pletely cover the same market. If the investor does not want to rely on different ETFs, he or she should choose a more broadly diversified ETF to ensure greater se­curity for the investment. The MSCI World ETF is a good example of a broadly diver­sified ETF. This index tracks the performance of more than 1600 different securities from 23 industrialised countries (MSCI Inc., 2018). In addition, the replication method of the ETF influences the risk to a certain extent. To avoid counterparty risk, a fully physically replicating ETF should always be chosen if possible. The reason why investors should accept any risk at all within the framework of private old-age provision is the expected return. Based on the abovementioned example of the MSCI World, an investor has been able to achieve an average annual return of 6.8% since 1970 (Zinnecker & Neumann, 2018). According to the efficiency market theory, it is possible - but difficult - for actively managed funds to beat the return of the bench­mark. In any case, an actively managed fund is more expensive than a passive in­vestment fund such as an ETF. A rising investment horizon also increases the prob­ability that the actively managed fund will not outperform an ETF. For reasons of long-term cost savings, an ETF is therefore better suited for private retirement provision (Morningstar, 2018). It may be the case that an MSCI ETF does not exactly achieve the return of its benchmark and a tracking difference is created. However, the development of this index is sufficient to show that historically a good return could be achieved in the long term. At this point it should be mentioned that histor­ical values do not justify future forecasts, but they do show how the index reacted in times of crisis. In this context, the exact development of the annual yield of this index is visualised in Figure 6.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6: Annual return of the MSCI World.

Source: Warnecke, 2017.

This chart indicates a high risk that arises with an investment in ETFs as a part of a private pension plan. Economic crises have had an enormous impact on the devel­opment of the MSCI World and led to large losses at this time. If the global economy is in a crisis, while an investor is going into retirement, enormous losses of capital should be expected if this capital must be accessed, which is usual at the start of retirement. This risk can only be reduced if the investor has the time to wait for a recovery of the stock prices. This fact reduces the capital flexibility of an ETF invest­ment if the investor does not want to take capital losses into account. However, di­versification is also worthwhile here because not all indices are affected by every crisis.

Examining the flexibility, in addition to the aforementioned restriction, ETFs are ei­ther particularly good or particularly bad for private retirement provision, depending on how disciplined the investor is about his or her saving behaviour. The capital can be transferred to other investments or paid out at any time. For people with little self-discipline, there is a risk that they will not use this capital for retire­ment provision but spend it elsewhere. For these people, an ETF savings plan is therefore not suitable for private retirement provision. One point that speaks in fa­vour of an ETF savings plan in terms of flexibility is the flexible contribution pay­ments. These can be adjusted or suspended at any time. In addition, the minimum contribution is low. To use an ETF savings plan, a depot must be opened first. The administrative effort is very manageable, especially due to the online portal, which normally has every custody account.

Considering the transparency of an investment as an investor objective, ETFs are excellent for private pension provision because their construction, structure, costs, and much other information can be seen at any time. However, transparency is only an advantage if the information given can be processed, which in turn defines the comprehensibility of an ETF. To select an ETF that meets the needs, the investor must deal partly with highly specific financial information. In the context of private pension provision, this requires a certain amount of initiative and interest from the investor to select the optimal ETF. Therefore, it is very helpful to deal intensively with the financial markets in general and ETFs in particular, which requires a certain basic financial understanding. Robo-advisers are a new upcoming trend and can as­sist an investor in putting together a suitable portfolio. These electronic, algorithm­based advisory programmes allow investors to invest in ETFs even if they only have the most superficial financial knowledge. Since this advice also generates costs, the use of a robo-adviser reduces the return of an ETF (Nowak, 2017).

3.2.2 Comparison to the commonly used pension alternatives

This subsection is devoted to making a final comparison between the pension alter­natives presented and the investment in an ETF savings plan. In this process, the products are evaluated with regard to the pension objectives that have already been identified. In this way, the advantages and disadvantages of investing in ETF savings plans are highlighted. The results of this comparison provide the theoretical content for the following empirical study. Within the framework of the acceptance analysis for the use of ETF savings plans, the perceived usefulness for the investor is determined by the advantages, while the disadvantages illustrate the variable of the perceived disadvantages.

Comprehensibility

In general, an investment in securities requires a higher level of personal initiative because the investor must acquire a certain degree of general financial education. To understand insurances, it is often enough to know the terms cost, interest, and contribution. In addition, the insurer helps to understand the products, which is not the case with an ETF savings plan. Furthermore, since the insurance products are made for retirement provision, the descriptions and formulations of the products are often very simple. By contrast, the relevant vocabulary of a securities invest­ment, which is required to understand the functioning of the products, goes far be­yond the standard terms and is often also written in English. In summary, the gen­eral comprehensibility of an ETF savings plan is lower than that of other pension alternatives due to its complexity. Nevertheless, it is possible for people who want to prepare a good old-age provision, and thus take care of it themselves, to under­stand the functioning of an ETF savings plan with the help of various sources of in­formation.

Return

Investors who prioritise this objective higher than capital security cannot avoid us­ing an ETF savings plan in the context of the compared products. The current low guaranteed interest rate on insurance products and the fact that only the savings component of the contribution is interest-bearing can result in no real increase in assets at the end of the term. The average return of an ETF savings plan depends on which index the underlying ETF tracks and how it develops. This means that no re­turn is guaranteed but you can draw certain conclusions from the historical perfor­mance of the indices, such as how an index reacts to a financial crisis and whether it has recovered from it. There are also indices which are riskier than others but there­fore promise a better return. The costs of all products also have an impact on the return that is ultimately achieved. Even if a quantified cost comparison is difficult to accomplish due to the product diversity and partly nontransparent cost structures of the conventional alternatives, it can still be shown that the costs of an ETF savings plan are lower than those of the alternatives. Because an ETF savings plan is a pas­sive investment, neither a fund manager nor an insurer will earn from it. The elimination of these two positions has an extremely positive effect on the cost struc­ture, which is why it is cheaper. The cost-average effect also helps to lower the av­erage costs of the ETF investment, and thereby increase the yield with rising prices. This effect is a unique feature of the ETF savings plan. Due to the very long term of a retirement provision investment, costs have a very large influence on the total re­turn at the end of the term. For this reason, it is important to pay attention to keeping the costs of a pension product low. In summary, on the basis of the information pro­vided, it can be said that an ETF savings plan can be expected to generate a much higher return than the other alternatives. However, in any case, they provide a bet­ter chance to generate return on the saved assets.

Capital safety

The counterpart to the return is the capital security or risk. The risk can be further divided into value loss risk and total loss risk. The total loss risk that would result from an insolvency of the insurance company or fund company is practically nonex­istent with all alternatives, because in the event of insolvency the saved assets are always safe. To completely eliminate this risk within an ETF investment, it is crucial to ensure that the ETF physically replicates the index to avoid the counterparty risk. The value loss risk does not exist within an insurance-based private pension be­cause the saved-up capital is always guaranteed. This means that the insured person is at least entitled to his or her paid-in assets at the time of retirement. An ETF sav­ings plan cannot provide this guarantee because it does not have an insurance cover. This guarantee is also a reason for the high cost of insurances. The ETF savings plan is exposed to volatility because it is based on securities. In this context, a long in­vestment horizon and broad diversification are essential to minimise the unsystem­atic risks of the market. The historical development of the abovementioned MSCI World also makes an investment in an ETF savings plan appear less risky, because the value of the index has recovered even after several crises. However, this does not necessarily apply to the future.

[...]