1
Master Thesis
Defined Ambition or third generation Defined Contribution: how can
both pension plans be compared in order to come to a choice between
the two?
University of Amsterdam Faculty of Economics and Business
Master’s in Actuarial Science and Mathematical Finance Erik Veerman
October 7, 2013 5606373
University of Amsterdam
dr. J.H. Tamerus AAG
Towers Watson Netherlands B.V.
E.W.J.M. Schokker AAG dr. G.C.M. Siegelaer
2
Content
1. Introduction ... 4
2. The Dutch pension system and the development of the Dutch Defined Benefit contract ... 7
2.1 The three pillars of the Dutch pension system ... 7
2.2 First pillar ... 7
2.3 Second pillar ... 8
2.4 Third Pillar ... 9
2.5 The most common pension plan in the second pillar ... 9
2.6 Historical developments in the Dutch pension system ... 11
3. Sustainability of Defined Benefit ... 14
3.1 Exogenous developments ... 14
3.2 Endogenous developments ... 17
3.3 Lessons from retrospection of pension history ... 17
3.4 Missed opportunities ... 18
3.5 What needs to change... 19
4. Defined Contribution ... 20
4.1 What is Defined Contribution... 20
4.2 Developments in the Netherlands and Europe ... 21
4.3 First-, second- and third generation DC ... 23
4.4 Premium Pension Institution ... 24
4.5 Behavioral economics in the Defined Contribution plans ... 25
5. Defined Benefit versus Defined Contribution ... 27
6. Defined Ambition and Dutch pension reforms ... 31
6.1 The concept of Defined Ambition ... 31
6.2 Defined Ambition in practice... 31
6.3 Dutch pension reforms ... 32
7. Participating in collective after retirement versus life annuity at retirement ... 36
7.1 Defined Ambition versus Defined Contribution ... 36
7.2 Risk exposure after retirement ... 37
8. Intergenerational risk sharing versus individuality ... 38
8.1 Intergenerational risk sharing in Defined Benefit ... 38
3
8.3 Quantifying welfare effects of intergenerational risk sharing ... 41
9. Uniform policy versus Lifecycle strategy ... 43
10. Price or value ... 46
10.1 Value-based ALM ... 46
10.2 Welfare-based ALM ... 48
10.3 Preferences of participants ... 49
10.4 How to choose a pension contract ... 50
11. Research approach and model description ... 52
11.1 Research approach ... 52 11.2 Model description ... 53 11.3 Demographic assumptions ... 59 11.4 Economic assumptions ... 61 11.5 Pension scheme ... 65 12. Results ... 66
12.1 Nominal Soft (fixed parameters) ... 66
12.2 Nominal Soft ... 69
12.3 Nominal Soft (risk loading) ... 71
12.4 Real Soft (DA) ... 76
12.5 Welfare gain of risk sharing ... 79
12.6 Lifecycle DC ... 84
12.7 Summary results quantitative analysis ... 86
13. Qualitative comparison of DA and DC along different subjects ... 88
13.1 Solidarity ... 88
13.2 Differentiation and flexibility ... 90
13.3 Affordable real pension results ... 90
13.4 Risk management and transparency ... 91
14. Conclusion ... 93 15. Bibliography ... 96 16. APPENDIX ... 100 16.1 Time series ... 100 16.2 Demographic assumptions ... 102 16.3 Yield curve ... 103
4
1. Introduction and research question
When ‘The Dutch pension system’ is entered in the Google search bar, the first suggestion is ‘Dutch pension system best in the world’. The second suggestion is ‘Dutch pension system reform’. This seems like an interesting combination to me. The Dutch pension system is seen to be one of the best pension systems in the world. But apparently something has to change and a reform of the system is necessary. What elements of the Dutch pension system have led to its success? And why is a reform of the system required? The traditional Defined Benefit (DB) contract, a nominal average pay pension scheme with conditional indexation is criticized and found to be unsustainable and not future-proof, by both science and politics. This has led to the design of a new Financial Assessment Framework and a new real pension contract, based on the concept of Defined Ambition (DA). At the same time, Defined Contribution (DC) contracts are gaining popularity and are constantly developing and improving.
In a DA contract, accrued pension entitlements are not guaranteed such as in DB. Accrued pension entitlements are soft and adjustable to financial shocks. Participants bear all investment risk and interest rate risk. But this is similar to DC. In third generation DC contracts, there is a target in terms of real income after retirement. There is also some kind of solidarity and collectivity in third generation DC. And this is similar to DA. So if choosing between DB and DC is seen as a choice between black and white, the choice between DA and DC is more like a comparison of different shades of grey. This paper analyzes the main trade-offs between DA and third generation DC. The central question of this paper is how a DA plan and a third generation DC plan can best be compared with each other in order to choose the appropriate plan.
In order to be able to make the appropriate choice, a comparison is made using criteria and two methods. On the one hand, DA and third generation DC are evaluated along the following criteria:
- Solidarity
- Differentiation and flexibility - Affordable real pension results - Risk management and transparency
5 On the other hand, two methods are used to compare the two plans. These methods are a quantitative analysis and a qualitative comparison. In the quantitative analysis, a welfare-based ALM model is used to analyze pension results of both pension plans. A welfare-based ALM model is based on participant’s preferences. The (realistic) assumption that is made in the quantitative analysis is risk aversion. The idea behind this is that an individual wants to avoid uncertainty and is willing to accept a lower outcome than the expected outcome in exchange for certainty. The amount a participant is willing to accept instead of a range of possible amounts is called the certainty equivalent. Hence, for risk averse individuals, the certainty equivalent is lower than the expected value. Risk averse individuals have a concave utility function, meaning that downside risk hurts more than that upward potential of the same magnitude is loved. The set-up of the quantitative analysis is a population of active participants that enters a DA contract and a DC contract simultaneously. The population enters the contracts with a clean sheet and the premium contributions are equal in both contracts. The populations operate in the same financial and demographic world, meaning that demographic assumptions and economic scenarios are identical. The qualitative method analyzes the following differences between DA and third generation DC. In a DA plan, participation in the collective pension fund remains after retirement, while in DC, a life annuity is bought at an insurer. This has consequences for the level of risk exposure and risk reward. Furthermore, an evaluation is made based on intergenerational risk sharing in DA versus individuality in DC. Spreading out financial shocks over time leads to a better time diversification of risk and intergenerational risk sharing is found to be ex-ante welfare enhancing. Finally, the uniform policy in DA is compared with the lifecycle strategy in third generation DC.
This paper can be divided in three parts, a descriptive part, a quantitative analysis and a qualitative analysis. The descriptive part is covered by chapters 2 until 6. These chapters describe the Dutch pension system, the DB contract and its questionable sustainability, the DC contract and the DA contract. The qualitative part is covered by chapters 7 until 10 and chapter 13. The quantitative part of this paper, chapters 11 and 12, describes the welfare-based ALM model and the results. The goal of this quantitative research is to analyze welfare effects of intergenerational risk sharing in DA and the lifecycle investment strategy in DC.
A large part of this thesis is inspired by the lectures of dr. Jan Tamerus at the University of Amsterdam. Those were long days, but they made me realize the great social importance of the design of a pension system, and that triggered my motivation to expand my knowledge. I was also inspired by a presentation of drs. Dick Boeijen at an actuarial congress. The theme of the congress and the presentation was
6 solidarity. After asking an audience of actuaries if collectively sharing investment risks within and
between generations should be preserved, or if is better to switch to an individual system, the audience was literally split in two. It was interesting to see that expert’s opinions were strongly divided. At that time, I was among the larger group that was in favor of collectively sharing risks. After that, interesting conversations with Edwin Schokker and interesting articles of dr. Gaston Siegelaer, service line leaders DC consulting and DC investment at Towers Watson Netherlands B.V., have led me to also see the advantages of the sophisticated third generation DC contracts.
I would like to thank Jan Tamerus for the interesting lectures and guidance in writing this thesis. I want to thank Hans Staring of PGGM for his criticism and tips. I want to thank my colleagues at Towers Watson for their time and advice. Finally, I want to express great gratitude to my parents and girlfriend for their patience.
7
2. The Dutch pension system and the development of the Dutch Defined
Benefit contract
This chapter provides a brief overview of the Dutch pension system. Furthermore, the recently prevailing pension plan is described. That is, the commonly used pension plan before the pension reforms in 2013. Finally, the historical development of this pension plan is summarized, which serves as an introduction to the next chapter about criticism on, and sustainability of the system.
2.1 The three pillars of the Dutch pension system
Pension systems of different countries can be very different. Even countries within Europe have varying pension systems. Some countries only have a pay-as-you-go system. This means that the working population is financing the benefits of the retired population. Other countries have a system where everyone individually takes care of their own pension. The Dutch system is a combination of both. In countries such as France, Germany and the United Kingdom, occupational pensions are voluntary. In France, Italy and the United Kingdom, participants have the freedom to choose the investment strategy. In other countries, the pension fund may decide how to invest the collected premiums.
The pension provisions in the Netherlands are regulated by the government, employers and insurance companies. The goal of the system is to make sure that every retired person has a decent income for the remaining lifetime. The system consists of three pillars: the state pension (AOW), capital-funded
occupational pension schemes, and individual pension schemes. The second pillar is the largest of the Dutch pension system, and consists of the pension schemes that employers provide. This paper focuses on different pension plans in the second pillar. Nevertheless, the following paragraphs provide a brief description of all three pillars.
2.2 First pillar
The first pillar consists of the state pension. The General Old Age Pensions Act (AOW) is a Dutch law from 1956 and is the basis for old-age pension benefits. Everyone that lives or works in the Netherlands between the age of 15 and 65 accrues state pension benefits annually. Up to 2013, the benefits were paid out from the age of 65. From 2013 onwards, the retirement age is increasing in steps to the age of 67. From then on, the retirement age of the state pension will increase with life expectancy.
The state pension benefits are independent of individual salaries. Pensioners living alone receive 70% of the minimum wage, and couples each receive 50% of the minimum wage. The first pillar is a pay-as-you-go system. The pension benefits are financed by the working population. When the contributions of the
8 working population are not sufficient to cover the benefits, the additional contributions come from the government public funds.
2.3 Second pillar
The second pillar consists of collective occupational schemes provided by employers, and administrated by pension funds or insurers. The second pillar is capital funded, and is an important supplement to the first pillar. Employers and employees pay a premium annually, which is invested. The total contributions and investment returns should cover the pension benefits.
The funding can take place through a pension fund or an insurance company. The company and the pension fund are strictly separated by law. The pension fund therefore does not depend on the financial wellbeing of the company itself. Pension funds can be company-specific or industry-wide. The
government can make it mandatory for employees in a certain branch to participate in an industry-wide pension fund. The government wants to create solidarity with this mandatory participation. If
employees switch job within the same branch, they remain in the same pension fund with the same pension scheme.
There is much variety in the characteristics of pension agreements and pension schemes in the second pillar. Pension schemes can provide an old-age pension, a disability pension, partner’s pension, orphan’s pension, temporary pension benefits and more. Pension schemes differ from each other in many aspects and one can think of many forms of flexibility of the pension benefits. This paper does not focus on the differences in pension schemes characteristics, but rather on the nature of the pension arrangements that employer and employees agreed upon.
The most common pension scheme in the Netherlands is a Defined Benefit (DB) scheme. In a DB scheme, participants annually accrue pension entitlements according to a wage dependent formula. A DB scheme can be a final pay scheme or an average wage scheme. In a final pay scheme, the total accrued pension entitlements are increased according to the last earned salary. A salary increase will also be reflected in the accrued pension entitlements in the past. In an average wage scheme, the accrual consists of a percentage of the pension base (salary minus offset) each year. A salary increase does not influence the accrued pension entitlements in the past.
The annual accrual of pension entitlements is financed by employer- and employee contributions. Each year, the premium that is needed to cover the accrual is calculated, based on financial and demographic assumptions. The pension fund (and thus all the participants) face the risk that people live longer than
9 anticipated when calculating the premium, so that the contributions during the working lifetime can end up to be insufficient to cover the benefits during the retired lifetime. This risk is referred to as longevity risk.
Due to increases in the overall price level (inflation), the accrued pension entitlements can end up to be much less worth in real terms. Pension funds therefore often have the ambition to provide indexation of the pension entitlements and benefits, to protect the purchasing power of the participants.
Instead of a DB scheme, an employer can also provide a Defined Contribution (DC) scheme to its employees. As the name suggests, only the contribution is defined. The employer and employee pay a fixed – or age dependent – contribution each year, which will be invested in the financial market. With the total sum of contributions and investment returns, the participant can buy an annuity at an insurance company. In the Netherlands, it is mandatory to use the accumulated DC capital to buy an annuity at the retirement age. Phased withdrawal in two terms is also possible, but the second term cannot be at an older age than 70 years. Other forms of phased withdrawal of the DC capital are prohibited.
Further details about DB en DC schemes are provided elsewhere in this paper. The prevailing pension scheme before the major pension reforms in 2013 is an average wage DB scheme. The next paragraph will describe this scheme more detailed.
2.4 Third Pillar
The second pillar consists of pension schemes provided by employers, but it is obvious that not the entire working population works for an employer. Self-employed individuals have to arrange their own additional pension. The third pillar of the Dutch pension system consists of the individual pension provisions that are purchased at an insurance company. Individuals that also accrue pension benefits in the second pillar, can save for extra pension provisions in the third pillar, and take advantage of the tax benefits that come together with saving.
2.5 The most common pension plan in the second pillar
The prevailing pension scheme in the Netherlands is a DB scheme. Up to the year 2000, the majority of the DB schemes was a final pay scheme. Since the beginning of the twenty-first century, the majority became an average salary scheme. In 2008 only 1% of pension fund participants actively participated in a final pay scheme. In 2009, 91% actively participated in an average salary DB scheme (Goudswaard et al.,
10 2010). Up to the year 2013, the retirement age was 65 years. However, participants can usually choose to retire earlier or later.
Social partners (employers and employees) negotiate about several aspects of a pension scheme, such as accrual rate, social security offset, maximum pensionable salary, indexation ladders, premium policy, investment policy et cetera. The indexation policy, premium policy and investment policy can be used as instruments by the pension fund to intervene if necessary. The use of policy instruments depends on the financial position of the fund.
Due to inflation, pension funds have the ambition to increase the pension entitlements with price inflation or wage inflation. This indexation is conditional on the financial position of the fund, and future indexation depends on the investment returns. If the funding ratio of the fund is below the minimum required funding ratio for several years, indexation can be negative. In recent years, the Dutch Authority Financial Markets pays extra attention to supervising that pension funds clearly communicate that indexation is conditional.
Each participant pays the same uniform premium in percentages of pension base or wage. This
percentage is based on the total premium needed to cover the total annual accrual. The premium that needs to be paid can be calculated by discounting the annual accrual of benefits with interest rates and mortality rates. Young active participants have a longer horizon until the retirement age, so the premium needed to cover the accrued pension benefits is lower than the premium needed for an older individual. Since everyone pays the same average premium, younger participants are paying more than actuarially needed. When they are older, they profit from this average system, because they pay less than needed. A pension fund board can adjust the premium according to the financial position. When the funding ratio is very high, premium rates could be decreased. When the funding ratio is high, or when demographic and financial assumptions change, a higher premium could be required.
The average premium creates some sort of solidarity. This solidarity is not a form of intergenerational risk sharing, but rather predictable subsidizing solidarity from young participants to older participants. The contributions are all together are invested in the financial markets. This means that every individual is participating in the same uniform collective asset mix. The assets are invested in different asset classes with different risk and return characteristics. Extra investment risk requires extra expected return, as a reward for taking the risk. Pension funds invest in risky assets in order to generate investments profits to finance indexation of pension entitlements. Because of the collective property of the funded DB
11 schemes, every individual faces the same risk profile, even though individuals may have different risk preferences.
With the introduction of the Financial Assessment Framework (FTK) in 2007, pension funds had to value their liabilities market consistently (fair valuation). Also, the probability of a shortfall may not be greater than 2.5% in any year. Therefore, pension funds are required to hold an extra solvency buffer on top of the solvency ratio of 100%. The buffer depends on the amount of risky assets in the investment portfolio. If the funding ratio equals the required solvency ratio, there is a probability of 97.5% that a shortfall will not occur within the year.
Pension funds have to value their liabilities with the yield curve as published by the Dutch Central Bank (DNB). This curve fluctuates from day to day, resulting in highly volatile funding ratios. Therefore, pension funds can frequently conflict with the FTK regulations. When interest rates are low, the corresponding value of liabilities is high. Since the FTK is designed to protect the nominal certainty of pension entitlements, pension funds need to hedge the interest rate risk. One way of doing this is to invest in fixed income assets, which replicate the liabilities. The nominal guarantees in the FTK conflicts with the ambition of indexation, which requires taking investment risk. This ambiguous target will be discussed further in the next chapter. The next paragraph will summarize the historical developments in the Dutch pension world.
2.6 Historical developments in the Dutch pension system
This paragraph will first briefly summarize important developments in the history of the Dutch pension system. In the next chapter, exogenous developments that put tension on the sustainability of the pension contract are described. Also, the next chapter describes criticism on the pension contract and necessary measures to make the contract sustainable are described. This paragraph and the next chapter are mainly inspired by the lectures of dr. J.H. Tamerus at the University of Amsterdam in 2013.
Information provided in this paragraph and the next chapter is mainly derived from Tamerus (2011). Van Marken is seen as the pioneer of the current Dutch pension contract. His vision was that employees, rather than employers, have the responsibility to save for a pension. However, according to Van Marken, the employee does not realize the importance of saving for future income, and prefers direct
consumption. Furthermore, it is too complex and expensive for an individual to arrange insurance properly. Therefore, Van Marken’s vision was that the employee needs to be protected and assisted by the employer to make sure that there is sufficient income after retirement. His definition of sufficient
12 income was 75% of the last earned salary. In the year 1880, the company of Van Marken offered a pension scheme to its employees, but this final pay scheme soon turned out to be unaffordable.
However, prior to the twenty-first century, most pension plans were still based on a final pay DB scheme. In 1949, participation in an industry-wide pension fund was made mandatory. Participants paid an average uniform premium, independent of age, combined with an average uniform accrual rate. This should make it easier for older employees or employers with relatively old employees to save for a sufficient pension income. In 1952, the ‘Pensioen- en Spaarfondsenwet’ (PSW) was introduced. This law was introduced to provide assurance and guarantees to pension fund participants. Employees are protected and they need to have certainty about their pension results. Pension funds must have a sound investment policy, and risks have to be reinsured. The PSW focuses mainly on a nominal guaranteed pension entitlements, but disregards the necessity of indexation (Tamerus, 2011). The 50’s and 60’s were characterized by substantial economic growth. In the beginning of the 70’s, pension funds were dealing with deficits, mainly caused by high wages and inflation, high interest rates combined with low
investment returns. In the second half of the 70’s, interest rates started declining, leading to increasing returns on fixed income assets. Also, inflation decreased, which led to decreased back service costs. Between 1980 and 2000, inflation was very low and pension funds invested relatively more in risky assets, which provided substantial high returns in this period. In the late 1980’s and 90’s, pension funds operated in times of prosperity. However, the prosperous times were not used to pay off outstanding deficits or building-up buffers. Instead, money was spent by lowering premiums. Pension funds
introduced Asset Liability Management (ALM) and tried to stabilize the premium level and invested more in risky assets. The goal was to guarantee the promised pension entitlement, but also to realize a
sufficient pension in real terms. The ambition to realize an inflation-protected pension at an affordable price requires high investment returns. Therefore, taking investment risk was necessary, and this risk had to managed properly. Indexation of pension benefits became more important to pension fund boards. As a reaction to the developments described above, the Dutch Central Bank (DNB) introduced the ‘Actuariële Principes Pensioenfondsen (APP) in 1997, which provided policy guidelines for pension funds. The APP were based on the philosophy in the PSW from 1952, which expressed the importance of nominal certainty. The focus on nominal guarantees resulted in increased costs for the indexation ambition. Pension funds needed to hold a solvency buffer on top of a minimum required funding ratio. The magnitude of the required buffer depends on the asset mix of the fund.
13 In the second half of the 1990’s, the government was the initiator of a discussion about the sustainability of the pension contract, mainly due to demographic changes. Social partners wanted to control the costs of pensions and to make the contract increasingly flexible. Indexations were made conditional. Other changes to the pension contract were needed due to the aging populations. Cost control and
sustainability of the pension became an important item on the agenda. In 1999, the legislating authority made pension accrual over past service years prohibited. Also, recovery periods were tightened. These measures increased nominal certainty. However, this nominal guarantee comes with a high price, especially for young participants.
Tamerus (2011) describes that the twenty-first century started euphorically. Times were prosperous, investment returns were booming, and welfare had enhanced significantly in prior years. The prosperous times of high welfare made risks invisible, and the costs of pension was lost from sight. However, after the year 2000, times changed radically. Events like disasters, terrorist attacks, the bankruptcy of giant companies and banks, led to new legislation and a new philosophy. Safety first became the new philosophy. Risks were not acceptable anymore. Meanwhile, the aging population and increasing life expectancy continued. Financial shocks occurred more often and more severe and we lived from one financial crisis to the other. The prosperous times of the 90’s made way for times of underfunded pension funds and suspicion amongst the public.
In 2007, the legislating authority introduced the new financial assessment framework (FTK). The protection of nominal pension entitlements was given priority. Market consistent fair valuation of liabilities became mandatory, leading to increased volatility of the funding ratio. Pension funds were required to hold a buffer, dependent of the investment policy, and perform a solvency test every year. The probability of underfunding in one year could not be greater than 2.5%. When the funding ratio drops below the minimum required funding ratio, pension funds were required to recover within three years. Pension funds had to communicate clearly that indexation was conditional. In a DB plan, nominal guaranteed pension entitlements and nominal funding ratio certainty became sacred. To comply with the new legislation, pension funds used hedging strategies and fixed income asset classes in their investment policy. On the other hand, the target of the pension field was to realize an affordable real indexation ambition. This requires sufficiently high investment returns, thus investing in risky assets. The FTK improved risk awareness, but did not resolve the conflicting targets of nominal certainty and
14
3. Sustainability of Defined Benefit
The previous chapter described the prevailing Dutch Defined Benefit contract, before the pension reforms in 2013. Also, a short overview of historical development in the Dutch pension world was provided, based on a much more extensive analysis of Tamerus (2011). This chapter describes endogenous and exogenous developments that threaten the sustainability of the current pension contract. Furthermore, this chapter discusses some of the lessons that can be learned from the retrospection according to Tamerus (2011). Next, some of the missed opportunities in the past are mentioned. Finally, the necessary changes towards a sustainable pension contract are described, based on Tamerus (2011).
3.1 Exogenous developments
Tamerus (2011) describes that the Dutch pension industry has missed some opportunities in the past. Looking back in time with present-day knowledge, we can conclude that the Dutch pension system has not been perfect and there are and always have been shortcomings. Nevertheless, the Dutch pension system is seen as one of the best pension systems in the world (Mercer, 2011). However, the world is changing and, as described before, the pension system is under a lot of pressure the last years. The Dutch government appointed a committee to investigate the future sustainability of the second pillar pension contract. The main conclusion was that the current contract was not future-proof and would become unsustainable (Goudswaard et al., 2010). This paragraph summarizes the main exogenous developments that make fundamental changes to the pension contract a necessity.
The first exogenous development that threatens the sustainability of the pension contract is the radical and structural demographic changes. The following graphs show the composition of the Dutch
population per age, in the years 1950, 2013 and 2060. This information is derived from the Dutch central bureau for statistics (CBS, 2013).
15 1950 2013 2060
It can be seen from the graphs above that the population composition was pyramid-shaped in 1950. Around the year 2013, the post-war baby-boom generation just retired. The composition of the population is now more shaped like a clock. In the year 2060, the population will almost be equally divided. One cause of this development is that life expectancy keeps increasing. People live longer on average, and thus need to receive pension benefits for a longer period. In addition, less people are born. As a consequence, the average age of the population keeps increasing, and the retired population relative to the working population keeps increasing. This development is very important to pension funds, since for every retired participant, there will be less active participants. If a pension fund has a relatively large active population with respect to the retired population, financial shocks can easily be absorbed with the premium policy instrument. Increasing the premium results in a relatively large profit relative to the liabilities. In contrast, a premium increase will not be very effective when the population is composed as expected in the year 2060. Financial shocks then need to be absorbed in some other way. Another problem of an aging population is that the possibility to take investment risk reduces. If negative investment returns result in incomplete or negative indexation, retired participants are immediately affected by this. The investment risk will generally be reduced when the average age of the population increases. The danger of this is that the pension fund will not earn enough investment returns, and the cost of pension increases.
Since the premium policy instrument is not very effective in a mature population, the investment policy gains importance. Investment returns are essential to afford the costs of pension. This brings us to the second exogenous development that threatens the sustainability of the DB contract, which is a
16 development in the financial market. The expected return on stocks shows a decreasing trend. At the same time, the volatility of stock return increases. As described by Dimson et al. (2009), the financial markets operate increasingly efficient, which leads to a decline of the risk premium on stocks. If the premium policy and investment policy loose effectiveness in absorbing financial setbacks, the indexation policy will have to be used. If the guaranteed pension benefits are unaffordable with premium
contributions and investment returns, the ambition needs to be adjusted or accrued benefits need to be cut.
Next to demographic and economic developments that require rethinking of the pension contract, there is another development which makes the appropriateness of the traditional DB contract questionable. This is the changing society. The DB contract fitted well in the post-war society, when risk aversion, collectivity and risk sharing were prevailing principles (Ponds et al., 1999). In past years, individualization, globalization, mobility and flexibility have become a trend. Competition is an important element in a globalizing world. Increasing pension costs in the form of increasing premiums have a negative influence on the competitiveness of a company. Furthermore, due to increasing mobilization in the labor market, the system of uniform premium rates and uniform accrual rates needs to be reconsidered. Elements in DB that are conditioned on the continuity of future service years are in contrast with increasing mobility in labor market (Tamerus, 2011).
The traditional DB plan becomes unsustainable due to these exogenous developments. The increasing life expectancy, the aging population and the lower expected investment returns will lead to increasing premium rates in a DB contract. In addition, the retirement age remains unchanged, while life
expectancy keeps increasing. This increases the costs of pension accrual even more. The increasing premium due to financial en demographic developments leads to labor market distortions. Also, the premium policy has a pro-cyclical effect on the economy. In times of economic stress and low returns, premium rates increase, so costs of labor increase and consumption decreases. This will make a
recession even more severe. In contrast, in times of economic growth, premium reductions will stimulate economic growth even more.
As mentioned before, the traditional DB under FTK suffers from a mismatch between the pension funds target and the investment policy (Frijns et al., 2010). The target is to provide an inflation-protected wage related income after retirement. To realize this ambition at an affordable price, this requires investment returns. However, the investment policy is designed to comply with the nominal certainty requirements.
17
3.2 Endogenous developments
Next to the exogenous developments that threaten the sustainability of the DB contract, such as described in the previous paragraph, Tamerus also describes endogenous elements that have to be eliminated or reduced to create a future proof pension contract.
One of the elements why critics criticize the current DB contract are the labor market and economy distorting effects of the DB contract. This is caused by the increasing premium in case of increasing life expectancy or financial stress. Also, the fixed retirement age leads to increasing labor costs. Another endogenous element why the DB contract is criticized are the conflicting objectives of nominal
guarantees and real ambition. Nominal certainty comes with a high price. It requires investing with low risk, while the ambition of indexation requires investment returns. High expected investment returns require taking investment risk. Other endogenous elements are the incomplete contract, subsidizing solidarity and the uniform policy. An in-depth analysis of these endogenous element is out of the scope of this paper.
3.3 Lessons from retrospection of pension history
This paragraph describes some conclusions and lessons drawn by Tamerus, based on the analysis of historical developments in the pension world, with present-day knowledge and insights. The following conclusions are a selection from Tamerus (2011). The sentences are no citations, but rather a free translation from Dutch.
Only define an ambition or benefit if it is affordable and if it is based on a decent actuarial fundament.
Prosperity and booming welfare lead to non-transparency of risks and costs of pension. Stay alert in prosperous times and overfunding.
Accrued pension benefits or benefits in payment should not stay behind much with respect to the ambition or purchasing power.
It is in the nature of the Dutch society to show solidarity with minorities.
Eventually, the employers bear the risk in DB. This risk exposure will become too severe to employers, due to the aging population, fair value valuation of liabilities, tight recovery periods and accounting standards.
18
No pension contract can beat economic developments or shocks. When the economy is booming, premium reduction or pension scheme improvements are almost unavoidable. In times of economic stress, increasing premium contributions or indexation cuts are necessary.
Legislation and authority can’t beat the economy either.
Final pay DB schemes are unsustainable. Eventually, intervention is unavoidable. You can only realize as much as you can afford.
There should be a decent actuarial fundament that is based on a wage related pension target and indexation ambition, in order to keep a pension contract sustainable.
The pension scheme should be complete and it should ex-ante be clear who bears what risks.
In a situation of overfunding, it is recommended to reduce investment risk and secure the financial position.
Negative and positive are asymmetric. A deficit hurts more than a surplus is enjoyed.
When the pension field and the supervising authority are in a dispute, the supervising authority should not use its power to force its rules. Instead, both should turn to the legislating authority.
A pension contract should not have two conflicting targets, such as nominal guarantees and a real ambition.
3.4 Missed opportunities
This paragraph describes some missed opportunities in the past as described by Tamerus (2011), based on the analysis of historical developments in the pension industry. This paragraph is a free translation of the findings of Tamerus (2011).
According to Tamerus (2011), a switch to a Defined Contribution (DC) contract is a missed opportunity, or rather a bad decision, since a DC plan lacks a wage relate pension target that is monitored constantly. This statement will be questioned further on this paper, where third generation DC is described. No clear rules about how to deal with overfunding and surpluses have also been a missed opportunity. Looking back, the prosperous times during the late 1980’s and the 1990’s could have been a great opportunity to use surpluses to improve the system. Instead of reducing premium rate levels, which dropped far below the actuarial fair premium, financial deficits could have been paid off. The surpluses could also be used to change the system of uniform premium and uniform accrual rate, which lead to subsidizing solidarity of young participants with old participants, and therefore threaten the continuity of the system. Finally, pension funds did not grasp the opportunity to reduce investment risk and secure the financial position.
19 The ongoing dispute about nominal guarantees versus real indexation ambition has been a major missed opportunity. The pension industry should have considered letting go of unconditional guarantees a long time ago, and accept the consequences of conditional indexation and the ambition of an inflation-protected pension. The nominal certainty requires safe investments and hedging strategies, while the ambition to realize an inflation-protected pension at an affordable price requires investing in risky assets and earn the risk reward. Trying to realize both goals is not an option. A pension funds has to choose one of these targets and follow a matching policy.
3.5 What needs to change
Based on the historic analysis of the Dutch pension system, Tamerus (2011) describes what changes have to be made to make the pension contract sustainable. Tamerus suggests letting go of the unconditional defined benefits. Accrued benefits should be conditional, and adjustable to exogenous developments. Also, the fixed retirement age should be replaced with a variable retirement age that increases with life expectancy. The possibilities of age differentiation and freedom of choice within the pension contract should be investigated. Although solidarity is an advantage, it should be more balanced. The system of average premium and average accrual causes too much subsidizing solidarity of young- and future generations with older generations. An alternative system may be necessary in the future.
Intergenerational risk sharing is proven to enhance welfare (ex-ante), but when certain generations contribute too much to risk sharing, the discontinuity risk increases. Tamerus also describes that nominal guarantees should be dropped, so that the policy of the pension fund can be designed to match the ambition of an inflation-protected pension benefit.
20
4. Defined Contribution
The first two chapters provided a descriptive analysis of the Dutch pension system and the DB contract. This chapter provides a descriptive analysis of the DC contract. It is first described what a DC contract is. Secondly, the increasing popularity of DC contract in the Netherlands and Europe is discussed. Some characteristics and obstacles with respect to first and second generation DC contracts are described. Furthermore, the so called third generation DC (or next generation DC) is described. Finally, some information about premium pension institutions (PPI) is provided.
4.1 What is Defined Contribution
As mentioned in the first chapter, in a DC contract, only the contribution is defined. The premium rate is age dependent. Premium rate ladders are subjected to fiscal maximum percentages. Similar to DB contracts, DC contracts can provide insurance against death and disability during the accrual period. The accumulated capital depends on the total contributions (after subtracted costs and risk premiums) and investment results. At the retirement age, the accumulated capital is used to buy a life annuity. This annuitization is mandatory in the Netherlands, phased withdrawal in more than two terms is prohibited. After retirement, an insurer pays an annual benefit to the participant. The insurer therefore has taken over longevity risk and investment risk. The height of the annuity depends on the actuarial assumptions underlying the insurer’s tariffs at the retirement date. The tariffs are subjected to interest rates and mortality rates (and costs). While the total accumulated capital is subjected to investment risk, the height of the annuity also depends mainly on interest rate risk and life expectancy.
It is also possible that a pension fund provides a DC scheme and benefits after retirement are paid by the pension fund. In the Netherlands, the majority of the pension funds provide a hybrid pension scheme, which is a combination of DB and DC (Pensioenfederatie, 2010). For example, up to a certain maximum salary, a DB contract is applicable. For the salary above the maximum, participants save for their pension in a DC scheme. The DC contract that is analyzed in this paper is based on a pure DC contract, in which the capital is used to buy an annuity at an insurer, rather than in a pension fund. This means that after retirement, there is no volatility of the annual benefits. The default risk is neglected in this paper. Also, in the quantitative comparison of DA and DC, costs are assumed to be zero, and risk premiums are
neglected. However, in reality there can be substantial differences between pension funds and insurers regarding costs.
In a DB contract, the employer faces risk with regard to pension liabilities. Exogenous developments can cause increasing premium rates. If a pension fund that provides a DB contract is in underfunding, the
21 employer has to pay additional premiums. Due to international accounting standards, an employer has to take the volatile pension liabilities into account on the financial balance sheet. In a DC contract, the employer only has the obligation to pay a predictable premium contribution. In a pure DC contract, the employer is not obligated to pay additional contributions when the pension target is out of sight. The advantage for the employer is that there is no volatile liability on the balance sheet and the future costs are very transparent. Experts believe that cost calculability is one of the biggest advantages of DC plans (Allianz, 2009). If employees (participants or retirees) bear all risks, employers do not have to take the risk with regard to pensions into account.
There are different variants of DC contracts. The DC contract as defined in this paper assumed that participants bear the investment risks and interest rate risk. It would also be possible that an insurer guarantees a certain capital. The annuity is still subjected to interest rate risk and life expectancy. Another variant is a DC contract, in which the contribution is converted each year into a defined benefit. The DB entitlement is then adjusted annually with indexation.
4.2 Developments in the Netherlands and Europe
In recent years, there is a global shift from DB towards DC (Allianz, 2009). Also in Western Europe where DB contract prevailed, the DC contract is gaining popularity. Experts, except the Dutch, expect that the DC contract will be the dominating contract in the future (Allianz, 2009). In the Netherlands, the DB contract is still the dominant contract. However, according to the survey of Allianz (2009), experts in the Netherlands also expect an increasing popularity of DC plans. DC contracts in different European
countries have different designs and different characteristics.
There are several developments that drive the shift from DB towards DC plans. One of the main reasons for employers to prefer a DC plan over a DB plan is the predictability of costs. Employers that strongly prefer steady and predictable pension costs could prefer a DC plan (Siegelaer, 2012). The sponsoring company has no further obligations next to the predictable contributions. Another driver for employers to shift from a DB plan to a DC plan has to do with HR strategies. For example, multinational companies might value labor mobility of their employers. A DC plan is better suitable in this context (Siegelaer, 2012). Furthermore, employers might want to offer freedom of choice to employees.
In some countries, cost reduction is the main driver. Furthermore, the reduction of investment risk and longevity risk of the employer is an important driver for shifting to DC (Allianz, 2009). Finally, one of the drivers mentioned in the survey of Allianz is the adoption of international accounting standards. The
22 importance of arguments differs among European countries. However, it seems that a shift from DB to DC is mostly preferred by employers rather than employees.
Siegelaer (2012) states that a DC plans also have certain benefits to the employee. A DC plan could lead to more involved participants than a DB plan. If a participant will have to make a choice about the amount of voluntary additional contributions or the risk profile of investments, the participant will become involved with his pension savings, given that there is proper education and communication (Siegelaer, 2012).
The financial crisis also has a major impact on the developments in the pension industry. Pension funds, authorities and employers will rethink the pension contract due to the financial crisis (Allianz, 2009). According to the survey of Allianz, it is not expected that the financial crisis will stop the shift towards DC plans. In contrary, the financial crisis will even cause an even greater shift to DC plans. Experts do expect that the financial crisis will cause an increasing importance of protection mechanisms in DC plans. The survey of Allianz also describes what experts see as obstacles towards better DC plans. The following obstacles (amongst others) are mentioned in the survey of Allianz (2009):
Poor financial education or inadequate advice;
Insufficient employee participation or lack of automatic enrollment;
Insufficient tax incentives;
Amount of costs (only in the Netherlands);
Lack of investment choices;
Too many investment choices.
It seems that all of the obstacles above are not unavoidable. The obstacles could be solved by the legislating authority or in the plan design.
Concluding this paragraph, the survey of Allianz (2009) also concluded that pan-European pension experts are mostly positive about a future pan-European pension market. They expected changes in legislation that will favor a pan-European market. The experts think that a possible European-wide pension plan will be a DC plan rather than a DB plan. This seems to fit well in the changing society as mentioned before, which is characterized by globalization, individualization and mobilization.
23
4.3 First-, second- and third generation DC
This paragraph describes the development of DC plan design in the Netherlands. The information is mainly derived from Siegelaer (2012). DC plans during the 1980’s and 1990’s are referred to as first generation DC. In first generation DC plans, participants can choose how their contributions are invested. The costs of asset management and administration were very high and have a substantial negative impact on the final DC capital. The main disadvantage of first generation DC is the mismatch between the investment policy and the pension ambition. When contributions are invested in a bank account, participants face substantial interest rate risk. A decrease in interest rates results in a higher price for an annuity, while the amount on a bank account does not increase (Siegelaer, 2012). On the other hand, when contributions are mainly invested in risky assets such as stocks, participants face substantial investment risk prior to retirement. Suddenly negative investment returns just before the retirement age affect the pension income for the remaining lifetime.
The second generation DC plans, at the beginning of the 21st century, were characterized by dynamic life cycle investment strategies. The interest rate risk of the annuity is hedged by investing in assets with matching interest rate sensitivity. As the retirement age approaches, the investment risk is decreased. The investment policy is improved compared to first generation DC plans. However, costs of asset management remain high and reduce the net returns. Also, the strategy is designed to reduce the risks with respect to a nominal annuity. However, participants also face inflation risk. In a world with inflation, a nominal annuity quickly loses value in real terms. Second generation DC plans do not provide sufficient protection against inflation risk after retirement (Siegelaer, 2012).
Siegelaer describes the concept of third generation DC. The key element of third generation DC is that design, management and communication are focused on individual preferences and planning.
Participants have to make decisions about desired real income after retirement and allowed uncertainty. The appropriate investment strategy is derived from the pension ambition by a technique called
reversed engineering. The result is an optimal life cycle strategy that maximizes the probability of achieving the defined ambition and risk preferences. Also, possibilities to increase this probability can be suggested, such as increasing the contribution of retirement age. The idea that people value high
outcomes less than they regret bad outcomes is taken into account. The idea of risk aversion and utility is described further in this paper.
24 In third generation DC plans, participants can choose to hedge inflation risk. The asset mix is a
combination of a matching portfolio and a return portfolio. The matching portfolio consists of assets that match the interest rate sensitivity and inflation risk sensitivity of life annuities. The return portfolio consist of diversified risky assets to generate investment returns. As the retirement age approaches, the part of the capital that is invested in the matching portfolio increases. The balance between the
matching portfolio and return portfolio depends on the risk appetite of the participant. Low costs of asset management are also an important element of third generation DC. When the pension provider is a so called premium pension institution (PPI), it can lead to substantial cost reductions. The next paragraph shortly describes the main idea behind a PPI.
Third generation DC plans thus have some improvements compared to current and older DC designs. Participants become more involved and responsible for their own pension. Also, instead of
communicating in nominal terms, the real value of income after retirement is the most important element. Participants set a clear pension target in terms of purchasing power. The important property of DC, freedom of choice, is preserved. However, participants do not have to make difficult decisions about investment strategies or an asset mix, instead they make decisions that are easier to understand. Complete freedom of choice in DC plans may not always be desired and can lead to non-optimal choices. The last paragraph of this chapter describes why freedom of choice can be non-optimal, based on the theory of behavioral economics. It is also described how third generation DC plans deal with this issue
4.4 Premium Pension Institution
A PPI is an institution that can act as pension provider for employers. A PPI can only administrate DC plans. The PPI collects the contributions, but does not pay out retirement benefits. At retirement, the accrued capital is used to buy an annuity at an insurer. A PPI may not be exposed to actuarial risks such as mortality risk, longevity risk or disability risk. If the PPI also wants to provide a spouse’s pension or disability pension in a pension scheme, this will have to be insured at an independent insurer. A PPI cannot guarantee a certain capital or annuity to participants.
In a way, the PPI acts as a director of administrators, asset managers and insurers. The possibility to unbundle services gives the advantage that the PPI can choose the insurer or asset manager with the best price. A PPI can create cost efficiency through the advantage of collectivity. By buying annuities or asset management as a collective, costs can be reduced. A DC plan provided by a PPI can therefore benefit from collectivity efficiencies, and also provide a form of solidarity. The solidarity provided by a
25 PPI is the sharing of risks with respect to mortality and disability. Another form of solidarity is
intergenerational risk sharing. DC plans cannot benefit from the welfare enhancing properties of intergenerational risk sharing.
4.5 Behavioral economics in the Defined Contribution plans
In DC plans, participants have the freedom to choose a certain strategy. Individuals can make decisions about for example asset mixes, voluntary additional contributions, retirement age or hedging of inflation risk. However, according to the theory of behavioral economics, individuals often don’t make rational choices. Bodie and Prast (2011) describe that people deviate from the rational choice model, because people have nonstandard beliefs, nonstandard preferences and exhibit systematic and predictable biases in decision making. They also state that the traditional assumption that more choice is better is
challenged. When there are too much choices, people are discouraged to choose anything or simply just pick anything (Bodie and Prast, 2011).
Nijboer and Boon (2012) describe some relevant insights from behavioral economics in the pension domain. They state that governments can act paternalistic to prevent people from making bad decisions. One argument is that people don’t see the importance of saving for a pension because of the long horizon until retirement. Also, financial decisions can be difficult, since people might not have the knowledge to make a rational choice. Even if a deliberate decision is made, people still make errors in their choices, such as investment decisions. Furthermore, people tend to underestimate the effect of inflation on the value of their income in real terms. Pensions have a big impact on welfare and wrong decisions are difficult to correct later in life. When people end up with an insufficient income after retirement, they will have to be helped by social security. This way, governments, and thus the whole society, will have to pay to make up for wrong or irrational decisions.
The optimization of life cycle profiles in third generation DC plans takes into account that people are risk averse. A very bad result is given more weight than extreme high results. So in a way, the behavior of individuals is incorporated in defining the appropriate strategy. The fact that complete freedom of choice can lead to non-optimal decisions, or the reality that most people don’t want or can’t make a decision (Siegelaer, 2012), results in default choices. Siegelaer suggest that most effort should be given to the design of standard default lifecycles that are designed to match the membership characteristics, rather than spending most of the time to a relative small group of participants that clearly want to make the investment decisions their selves.
26 Despite the fact that people might make wrong or irrational decisions, freedom of choice fits in the current day society. People nowadays find it increasingly important that their pension matches their individual preferences and financial needs (Siegelaer, 2012). Pension has become part of personal financial planning. Siegelaer suggests to take insights from behavioral economics into account when participants are given the freedom to make decision about their pension. He states that people should not be provided too much information and possible choices. Instead, people should first become more emotionally involved with their pension, so that they become more prepared to think about possible choices and consequences with respect to pensions. Also, people should be forced to decide within a certain time limit, since people tend to postpone decicisions. Finally, it can be helpful to design default choices and guide people in their decisions, since the stress of making a difficult decision can lead to the wrong choice.
27
5. Defined Benefit versus Defined Contribution
Recently there is a growing interest in DC system in the Netherlands, partly because of the widening global economy and the need for cost control. Also, the number of Dutch pension funds is reducing in recent years. The Dutch society is becoming increasingly individualized. Furthermore, the pension contract is becoming increasingly flexible, pensions contracts need to meet different personal
preferences and participants become more and more aware of their financial situation with respect to pension (Kuné, 1999).
Employers have put their focus on the control of pension costs. In the past this has led to a shift from final pay contracts to average pay contracts, since the final pay contracts led to substantial increasing premium contributions (Tamerus, 2011). Indexation of pension entitlements became conditional. Now, since the unsustainable increasing costs of conditional average pay DB contracts has come to light (Goudswaard, 2010), DC contracts become interesting to employers. In a DC system with fixed
predictable (but age-dependent) premium contributions, employers face maximal cost transparency and low administration costs.
A shift from a DB to a DC system will have generational effects. In a DB system everyone pays an average uniform premium. The actuarial fair premium for young participants is much smaller than the actuarial fair premium for older participants. A shift from DB to DC will have negative effects for the older active participants, since they paid more premium than actuarially needed when they were young, but don’t benefit anymore from the value transfers of the current young participants to the older participants. In a DB system, risks and returns are shared between generations. If one generation profits from this
solidarity, others will have to contribute to solidarity. In other words, a pension fund is a zero-sum game. At the end of the twentieth century, the economy became increasingly globalized and internationalized. As a result, the effect of national legislation diminishes. European legislation increasingly replaced national laws. The internationalization has led to more and more shifts to DC systems (Kuné, 1999). In a DB system, when real investment returns are low, the costs of indexation are passed on to the active working population of the pension fund. This leads to increasing premium rates. On the other hand, when investment returns are high, premium rates can stay at a relative low level. The impact on the premium is largest in a mature fund.
28 The overall level of premium contributions is reaching a maximum (Goudswaard, 2010). Because risks are shifted to future generations, the current nominal DB pension contract is in danger of becoming unsustainable. Future generation might refuse participating in the pension fund. In a DC system, there are ex-ante no value transfers between generations. However, the realized pension results will ex-post fluctuate between generations, because results are completely depending on realized investment results. Ex-post there will be winning and losing generations.
Ponds et al. (1999) describe the main difference between DB and DC as being a trade-off between risk sharing (solidarity) and freedom of choice. The DB contract fitted well in the post-war society, when risk aversion, collectivity and risk sharing were prevailing principles (Ponds, E.H.M., Bosch, R., Breunesse, E.A., Willemsen, B-J., 1999). In past years, individualization, globalization, mobility and flexibility have become a trend. So a DC contract became an interesting alternative. In a DC contract, the participant bears all risks, so there is more uncertainty about the pension result. On the other hand, there is more room to realize personal preferences, since there is more freedom of choice. However, Ponds et. al (1999) point out that complete freedom of choice can lead to non-optimal results due to non-optimal choices of individuals. Also, pension funds often offer flexibility in DB contracts, since completely uniformly designed contracts do not match with the heterogeneity of these days.
Ponds et al. (1999) extensively describe the aspects of real interest rate risk in DB and DC contracts, which they call the most important risk in pension funding. They state that an intergenerational risk sharing collective can create protection against downside risk of real interest rate risk. When periods of high interest rates are followed by periods of low interest rates, the risk can be reduced ex-ante by means of a better time diversification. Spreading out risks, by sharing risk between (future) generations, will be (ex-ante) welfare improving. Ponds et. al also refer to the philosophy of Rawls, who assumes a situation in which all individuals are ex-ante equal and risk-averse, since they do not know anything about the future and will therefore be prepared to share risks and thus create some protection against economic shocks. In the starting situation, everyone is ignorant about economic developments, but when time has passed and ignorance has vanished, participants will know whether they have contributed or have benefited from intergenerational risk sharing.
In addition to protection against economic risks, DB contracts can also provide certain protection to pension funds or participants when participation is mandatory. Mandatory participation in a collective pension fund can rule out problems like adverse selection, bounded rationality and myopia (Ponds, 1999). This could also be realized in a DC scheme though.
29 Ponds et. al state that the essential question when choosing the appropriate contract is the desired degree of real interest rate risk sharing. The real interest rate risk in a DB contract can be managed by means of three policy instruments: indexation policy, investment policy, and premium policy. The policy determines the degree in which real interest rate risk is reflected in the three risk measures: solvability risk, probability of incomplete indexation, and premium volatility. A change in policy that leads to an improvement of one of the three risk measures, will lead to a decline of one of the other risk measures. The investment policy requires special attention. There is a certain trade-off in the investment strategy between risk and return. Investing relatively risky leads to relatively high returns on average. High returns lead to a lower premium contribution. However, this comes together with more risk. The investment policy should also depend on the desired degree of hedging real interest rate risk. Since liabilities need to be valued market consistently, the funding ratio of a pension fund is very sensitive to interest rate fluctuations. Investing in fixed income assets that have real interest rate hedging qualities can reduce the volatility of the funding ratio. However, the downside of hedging, is that it will lead to lower investment returns on average, which will result in a higher required premium contribution. Investing more in stocks means higher expected returns, but the interest rate risk increases since stocks have worse real interest rate hedging qualities compared to bonds or credits (Hoevenaars et al., 2008). In an individual DC contract, participants have more opportunities to choose an investment strategy that fits their specific preferences. Participants can profit from upward potential of investment returns, but also face all downside risk of investments. The real interest rate risk is also present in the pension results of DC participants. In the DC contract that is discussed in this paper, the DC capital is used to buy a life annuity at the retirement age. Phased withdrawal is assumed to be not allowed. This is the current situation in the Netherlands.
The insurer’s tariffs, that determine the annuity that can be bought with the accumulated DC
capital,mainly depend on interest rates. Suppose that a participant is 67 years old and has accumulated a DC capital which he or she uses to buy a life annuity when interest rates are relatively low. If interest rates were high, the annuity would provide a higher yearly benefit. The dependence of the pension results on interest rates creates a great deal of uncertainty prior to retirement. If interest rates suddenly drop one year before retirement, the accumulated capital provides a significantly less annuity. The participant could create protection against this interest rate risk. When the retirement date approaches, the participant will invest less in risky assets such as stocks, and more in fixed income assets with interest
30 rate hedging qualities, such as government bonds. If interest rates drop prior to retirement, the value of the fixed income assets increases, so the pension results are less volatile.
The volatility of stock returns will also cause participants to reduce investing in risky assets when approaching the retirement date. If stock returns are extremely negative just before retirement, the DC capital could vaporize quickly. One year, the participant may believe that his pension is secure, and the next year he may be left with a very humble pension. On the other hand, at younger ages participants can invest relatively more in risky assets, since they have a long horizon until retirement and financial setbacks can be compensated with future profits. The theory behind this is described more detailed in the chapter about lifecycle strategies.
Risks in a DC contract can be reflected in pension results, such as described above, or in the premium contribution during the working lifetime. If premium rates are variable and based on a target benefit, the uncertainty will be translated in the premium contribution. If the ambition is not expected to be met, premiums will have to increase to achieve the desired goal. If premiums are fixed, the uncertainty will be translated in a probability distribution of pension results. The latter is the assumption in the quantitative research in this paper.
In Ponds et al. (1999), a DB and DC contract are evaluated by means of an ALM analysis. They conclude that a DC participant faces significantly more uncertainty, while results are only slightly better on average. The ALM analysis in this paper evaluates a DA plan and a DC plan. In a DA plan, participants bear more risk compared to participants in a DB plan. In a DB contract, the employer eventually bears the risk. In a DA contract, the accrued pension entitlements are soft, meaning that they are adjustable and depend on financial shocks. A DA contract has more similarities with a DC plan than a DB plan does. However, Tamerus (2011) describes that a DA contract still is more similar to a DB contract than to a DC contract. Ponds et al. (1999) conclude that an individual in a DC contract can only realize a higher return than a DB participant, when the share of stocks in the asset mix is higher on average.
31
6. Defined Ambition and Dutch pension reforms
In chapter 2, it was described what changes are necessary in the Dutch DB contract, according to Tamerus (2011), in order to make the contract sustainable. Tamerus also provides a concept for a new pension contract, which is called DA. This chapter briefly describes the concept of DA. Furthermore, the recent developments with respect to the new Financial Assessment Framework (nFTK) are described.
6.1 The concept of Defined Ambition
Tamerus (2011) evaluated sustainability along different dimensions, such as affordability, solidarity and labor market. One of the conclusions of Tamerus is that it is necessary to let go of the unconditionality of pension entitlements, the nominal certainty, and the fixed retirement age. Also, labor market mobility should be improved and solidarity between generations should be more balanced. A wage related target and inflation protection is the core of Defined Ambition. Pension entitlements and retirement age are soft, and are dependent on exogenous developments. In DA, there is no nominal guarantee, but there is security assurance. The defined ambition should be monitored and protected by the social partners. Since pension entitlement are soft in DA, the differences with DC gets smaller. However, DA is still more similar to DB, but without the unconditionality. Similar to DB, a pension fund with a DA still collectively accrues pension entitlements, with a wage related and inflation protected pension target. Pensions are still financed with a system of capital funding. Also, employers are involved, and can act as a last resort if the target gets out of sight. Collectivity, solidarity and mandatory participation are preserved.
6.2 Defined Ambition in practice
Tamerus describes how a DA contract would work in practice. The first step is to determine a target, in other words to define an ambition. The social partners should agree a accrual rate, offset, retirement age et cetera. Furthermore, the ambition should be an inflation protected accrued pension. For active participants this could be wage inflation, and for retired participants this could be price inflation. Social partners should also agree on the level of risk exposure, or risk appetite. The ambition is an expected target, but the maximum downside risk should also be clear. The investment risk should be matched with the risk appetite of the different stakeholders.
Similar to DB, pension entitlements are accrued for each service year. There are accrued pension entitlements, a reserve and a real funding ratio. The difference with DB is that the retirement age, accrual rate, offset and accrued pension entitlements are adjustable to exogenous developments. The adjustments can also affect accrued pensions in the past.
