Sustainability of the Dutch pension system: A case study of the Netherlands
University of Groningen
Faculty of Economics and Business
Department of Finance
MSc Business Administration, Finance
July, 2012
Author: Jannes Overweg
E-mail: j.overweg@student.rug.nl
Student number: 2044463
2 | P a g e
Sustainability of the Dutch pension system: A case study of the Netherlands
Abstract
This thesis examines the pension system of the Netherlands. It examines whether the current pension system is sustainable in the future. Criteria with respect to future generosity, affordability and solvability of the pension system are established to measure future sustainability of the Dutch pension system. Several scenarios are used to examine the impact of each scenario on the established criteria. These scenarios are developed with respect to future pension age, investment returns, inflation, and mortality rates. The two pillars of the Dutch pension system, AOW pension and supplementary pension are examined separately. I conclude that the AOW pension cannot sustain in its current form, because the future distribution of the labour force relative to retirees has a great impact on future generosity to retirees and/or affordability by the labour force. On the other hand, the current form of the supplementary pension system is sustainable because the current amount premiums paid by the labour force are sufficient to cover future liabilities. I conclude that there can be shift of dependency on AOW pension towards more dependency on supplementary pension and individual pension insurance.
JEL classification: G23, H55, H62, J11
Keywords: Pension system, sustainability, the Netherlands, AOW, pension funds
3 | P a g e
Table of Contents
Introduction ... 4
I. Literature overview ... 6
A. The Dutch pension system ... 7
B. Foreign pension systems ... 8
B.1. Pension system of the US... 8
B.2. Pension system of the UK ... 9
B.3. Pension system of Australia ... 10
B.4. Differences between the Dutch and foreign pension systems ... 11
C. Asset Liability Management (ALM) ... 12
II. Methodology ... 13
A. Criteria ... 13
B. Scenarios ... 14
C. Evaluation criteria supplementary pension ... 16
C.1. Calculation supplementary pension premium payments until 2010 (1st component) ... 17
C.2. Calculation amount of supplementary pension benefits (2nd component) ... 18
C.3. Calculation supplementary pension premiums between 2010 and year of retirement (3rd component) .. 19
C.4. Matching assets with future pension benefits (liabilities) ... 20
D. Evaluation criteria AOW pension ... 21
III. Data description ... 22
A. Collection of data ... 23
B. Historic and forecasted population structure ... 24
C. Historic macro figures of Dutch pension funds ... 25
IV. Results ... 26
A. Results first criterion supplementary pension ... 27
B. Results second criterion supplementary pension ... 30
C. Results third criterion supplementary pension ... 31
D. Results first criterion AOW pension ... 33
E. Results second criterion AOW pension ... 34
V. Conclusion ... 35
VI. Literature list ... 40
4 | P a g e
Introduction
The pension system of the Netherlands is often quoted as one of the best pension system in the world. An example is a study by Mercer (2011), which quoted the pension system of the Netherlands and Australia as best pension system in the world. Both systems have a sound structure, with many good features, but have some areas for improvement. Though, on what ground is this quote based? Surely, if we look at present pension benefits in the Netherlands, the average “elderly” cannot complain about the level of their income. A legitimate question is, whether the present pension system will be sustainable in the future.
Currently, the pension system in the Netherlands has three pillars. The first pillar is a public facility called Algemene Ouderdomswet (AOW). Everyone who reaches the age of 65 (in the future 671) is entitled to AOW pension. The second pillar is a supplementary pension through an employer sponsored scheme. An employee builds a supplementary pension during his/her working life, for, which the employee and the employer pay a premium. Most of the time, the employer takes account for more than half of the paid premium. The third pillar is not as common compared to the first two pillars. The third pillar is an individual pension insurance that everyone can close down voluntarily, but will not be further
elaborated in this thesis because it is not a general instrument of the Dutch pension system. The pension system in the Netherlands is a combination of a public facility (government), private facility (employer) and individual facility. The Netherlands has one of the largest pension reserves per capita in the world2. The first pillar is a Pay-As-You-Go (PAYG) system, where benefits are paid from currently paid premiums. Current employees and other beneficiaries pay the costs of the AOW benefits that are paid to current retirees. The second pillar is financed both by employers and employees. The premiums are invested by pension funds, to cover future liabilities of the employee. These future liabilities towards the employees are paid by the premiums they paid in the past and the returns on the premiums. The supplementary pension benefits used to be considered as a guaranteed amount to be paid by the pension funds. The new pension agreement of July 2011 clearly states that the amount of supplementary pension benefits depends on the returns of investments made by the pension funds and the supplementary pension
1
Governmental policies are aimed to gradually increase the legal pension age to 66 in the year 2019 and 67 in the year 2023.
5 | P a g e
premiums paid. This pension system has a lot of positive sides, though, one major negative side. Current employees and employers have to pay (relatively) a lot of taxes to provide sufficient AOW pension benefits to current retirees and sufficient futures supplementary pension benefits for themselves. Since the Dutch population is ageing, the tax burden on employees has to grow to provide sufficient AOW pension benefits to current retirees. This problem is known as “the (global) aging problem”. Jackson et al. (2010) developed the Global Aging Preparedness Index (GAP index) to measure the preparedness of 20 countries to overcome the global aging problem. The countries that are included are the most major developed countries and a selection of economically important emerging markets. The Netherlands is one of the 20 countries examined. The GAP index consists of two sub-indices, namely the Fiscal Sustainability (FS) index and the Income Adequacy (IA) index. The FS index looks at projections of spending on pension and health benefits. It also takes into account the differing fiscal room that countries have to accommodate their growing old-age dependency burdens by raising taxes, cutting other spending or borrowing costs. The IA index tracks trends in the living standard of the elderly (retirees), relative to the nonelderly (labour force), based on projections that factor in the impact of changes in public benefit programs, private pension provisions and labour-force participation rates. Not surprisingly, the Netherlands has the best score on the IA index compared to the other 19 countries. With its high living standards and a labour-force participation rate of 96.8%3 it is expected that the Netherlands provides an adequate income to their elderly. Unlike the high score on the IA index, the Netherlands scores poorly on the FS index. With a 19th place, it only scores better than Spain on fiscal sustainability. This huge difference between the two dimensions of aging preparedness is also the case for other countries. Most countries that have a high score on the FS index have a low score on the IA index, and vice versa. This indicates that today’s retirement policies entail a trade-off between fiscal sustainability and income adequacy. Therefore, the research question of this thesis: Is the pension system of the Netherlands sustainable in its present form?
According to Jackson et al. (2010), the Dutch pension system is not sustainable in its present form. Two important strategies to mitigate the trade-off between fiscal sustainability and income adequacy are extending work lives and increasing funded pension savings. There are already plans to gradually extend the legal pension age of 65 years to 67 years in the Netherlands. AOW pension benefits are still the
3
6 | P a g e
same, but it is not certain that this will persist in the future. Also, some pension funds with funding ratios below 105%, announced that there is a chance that current retirees will be shorted on their supplementary pension benefits. It is not questioned if the government of the Netherlands takes
measures to overcome future problems of their pension system, but the question is; are taken measures enough to overcome the problems? What strategies are available to keep a sustainable and adequate income in the future? It is possible, because Australia and Chile score toward the top of the FS and IA index. Jackson et al. (2010) propose seven strategies to overcome the global ageing problem. This thesis only focuses on the sustainability of the pension system of the Netherlands. Therefore, strategies about, for example, health-care costs and immigration will not be further elaborated. The core is about the pension system. Therefore, in testing the sustainability of the Dutch pension system, I will take the following strategies into account:
1. Reduce pension benefits 2. Extend work live
3. Increase pension premiums
I have established criteria to measure the sustainability of the Dutch pension system and will test the system, based on these criteria, for scenarios. The established criteria measure the generosity, affordability and solvability of the Dutch pension system. The scenarios, which are tested, differ with respect to the retirement age, investment returns, inflation, and mortality rates.
The next section presents an overview of the literature on Dutch and foreign pension systems. Section two is the methodology section followed by the data description section where the process of data collection is explained (section three). Section four presents the results derived from my model followed by the conclusion in section five. Appendices can be found after the literature list.
I.
Literature overview
7 | P a g e
A. The Dutch pension system
Kasemir et al. (2001) argue that most developed economies have a multi-pillar pension system. The first pillar is a system that is run by the government, which is usually a PAYG system. Government-run systems are coming under pressure because they may be not be sustainable under the current demographics in many Organization for Economic Co-operation and Development (OECD) countries, where the populations are aging. However, a decrease in generosity of the first pillar can have negative consequences. Groezen et al. (2007) measure the effect of consumer possibilities of retirees when cutting PAYG schemes. They advise to keep the current percentage of GDP on PAYG schemes in the Netherlands to guarantee a balanced consumption profile between the employees and retirees. They argue that the economy is more likely to become inefficient if the PAYG scheme is significantly reduced, caused by elderly individuals who need more labour-intensive services than young people. Labour-intensive services become more expensive when there is a shift to a one pillar system, which corrodes the purchasing power of retirees. However, the Dutch dependency ratio will rise sharply in upcoming decades (see also Van Ewijk et al., 2006), which still imply a significant decrease in AOW pension benefits. This is not strange because in comparative perspective, the Dutch pension system is rather generous (Riel et al., 2002). Upcoming demographics show the weakness of the AOW pension and increase the importance of supplementary pension.
The second pillar consists of occupational payments to pension funds, which pay supplementary pension benefits when retiring. Bikker and Vlaar (2007) examined the outcomes of unfavorable scenarios on expected development of supplementary pension premiums, supplementary pension benefit cuts and funding ratios in the Netherlands. They looked especially at the possibility if requirements, set by the Financieel Toetsingskader (FTK), are not too restrictive for pension funds, which offer conditional or unconditional indexation. They conclude that the FTK implies no over-restrictive requirements to pension funds that offer conditional indexation. Pension funds that offer unconditional indexing have unfavorable simulation outcomes. These unfavorable simulation outcomes concern, for example, a higher probability of funding ratios that not meet the standard of 105%. The FTK requires large implicit capital buffers, which cannot be met by pension funds that offer unconditional indexation. Other
requirements set by the FTK are the level of funding ratios of pension funds, which expresses the level of their assets relative to expected future liabilities. The legal minimum funding ratio is 105% in the
8 | P a g e
year time horizon4. Pension funds with funding ratios below 125% have to issue a recovery plan with a time horizon of maximum 15 years if they fall below 125%. Pension funds are only allowed to make up for past recovery measures if they have funding ratios above 145%5. Nowadays, pension funds like ABP and Zorg & Welzijn face difficulties, which results in funding ratios below 105%. ABP and Zorg & Welzijn provide the supplementary pension of public servants. The difficulties that these two pension funds face are partly caused by actions from the Dutch government in the nineties. The Dutch government used public pension money to balance their budgets in the nineties. Kalloe and Kastelein (2011) argue that the Netherlands must not see pension savings as public property but rather as property of individual participants. Apparently, public finances took precedence over the level of the pensions in the Netherlands, which is not desired.
B. Foreign pension systems
As was already mentioned in the introduction, Australia, for example, scores well on both fiscal sustainability and income adequacy in the study of Jackson et al. (2010). Therefore, I will examine the pension systems of three large countries, including Australia. The other two countries that will be examined are the United States (US) and the United Kingdom (UK). All data of the three foreign pension systems are derived from the OECD (2011).
B.1. Pension system of the US
The pension age in the US is 66 years in 2008. This will be increased to 67 in the year 2022. The eligibility off receiving retirement benefits for the public scheme (social security) depends on the number of years contributing to the system. There is a minimum requirement of 10 years contribution to be eligible for retirement benefits. The benefits are based on a total of 35 years. If someone worked more than 35 years, the highest wage years will account to the calculation of the retirement benefits. If an employee worked less than 35 years, than years with zero earnings will be added to total the 35 years. The
calculations of the pension benefits are calculated as follow. The eligibility for receiving social security is based on a progressive benefit formula. The first USD 711 a month of relevant earnings attracts a 90% replacement rate. The income between USD 711 and USD 4,288 is replaced at 32%. All income above USD 4,288 has a replacement rate of 15%, up to a ceiling of USD 102,000 a year. For example, if someone earns on average USD 3000 a month, he builds up pension benefits equal to 90% of USD 711 and 32% of the resulting USD 2289. This results retirement benefits of USD 1,372 per month. This
4
This is temporarily extended to five years by DNB (OECD, 2010). 5
9 | P a g e
accounts for single wage earners. For married couples, there is a 50% dependents’ addition where secondary earners have built up a smaller entitlement.
In the US, there are possibilities for early and later retirement. Early retirement is possible at the age of 62, late retirement is possible up to the age of 70. For early retirees, each year before normal
retirement, pension benefits are reduced by yearly 6.75%, which accounts for 3 years. After three years, reduction in pension benefits falls to 5%, which applies to retirees with a normal retirement age of 66. Receipt of pension benefits can also be deferred after normal retirement age. Credit is given for deferment up to the age 70. Retirees receive an additional 8% a year, for each year deferring their pension age. Another possibility is a combination of pension and work. The pension benefits of beneficiaries, who are receiving benefits in years before the normal retirement age, are reduced with 50% of earnings that are in excess of USD 13,560. This is not the case if they reached the normal retirement age.
Furthermore, the US has also individual retirement accounts such as the Individual Retirement Arrangement (IRA) and 401(k) plans. The IRA provides tax advantages for retirement savings and can only be funded with cash or cash equivalents. An IRA comprises of an individual retirement account and the purchase of an annuity contract from a life insurance company. 401(k) plans are defined
contribution plans. 401(k) takes its name from a subsection (401k) of the Internal Revenue Code. Contributions to a 401(k) plan are tax deductible, but are set to a limit. Employees pay taxes if they withdraw money from the 401(k) plan. The major benefits of 401(k) plans are that interest is paid on contributions and tax payments are delayed.
B.2. Pension system of the UK
10 | P a g e
potential working lives. Working live in the UK is from 21 to 60 for women in the “old” situation and 21 to 65 for men and women. In the Pension Act 2007, the number of years contributing to (potential) working live is decreased to 30 years. Therefore, someone that lived the last 30 years of his/her live in the UK before reaching the pension age has entitlement to a full basic state pension. If a person reached the pension age in 2008/09 in the UK, that person received an amount of GBP 90.70 per week, which is close to GBP 363 a month. In comparison, this amount is 14% of average earnings in 2008/09. From 2010, the UK has two income bands to specify a replacement rate. There is a replacement rate of 40% between an income of GBP 4,680 and GBP 13,500 per year and a replacement rate of 10% between an income of GBP 13,500 and GBP 40,040. For example, the average income of a person is GBP 25,000 per year. That person receives, after reaching the pension age, an amount of 52 times GBP 90.70 from the basic state pension, resulting in a yearly amount of GBP 4,716. Note that this is close to the lower earnings limit of GBP 4,680. The retiree receives an additional amount of his earnings related pension. This amounts to 40% of the difference between GBP 4,680 and GBP 13,500 plus 10% of the difference between GBP 13,500 and GBP 25,000, which in total is a yearly amount of GBP 9,394 of pension from the public scheme. Employees in the UK complement their pension with an additional private pension plan. Moreover, most employees contract out of the earnings related public pension scheme and solely complement their pension by a large voluntary private pension sector.
B.3. Pension system of Australia
Australia also has a distinction in pension age for men and women, respectively, 65 and 63.5 year. The pension age for women will gradually increase to 65 by 2014. After 2014, pension age will be further increased for both men and women to 67 by 2023. The pension system of Australia has three
components to supplement income after reaching the pension age. The first component is a public age pension, which is funded by general taxation revenues. The second component is a compulsory
employer contribution to a private superannuation guarantee. This is a pension arrangement
11 | P a g e
In most countries, defined-benefit pension schemes dominate the pension systems. In Australia, most employees are members of defined-contribution schemes. Pension benefits from the superannuation guarantee are converted to a price-indexed annuity. Mortality data of Australia is used to calculate the average age of Australian people, which is needed to calculate the annuity. The public age pension is designed to help the people who need it the most. It is paid once in two weeks’ time. Furthermore, every 6 months, the public age pension is increased with the Consumer Price Index (CPI) of Australia. In 2008, a full annual payment of public age pension was AUD 14,313. However, this amount could get reduced if someone did not “pass” the assets and fixed income tests. There was a free area of an average income of AUD 3,510. If someone had a fixed income above this threshold level, they would get a part-rate public age pension. Another test is the assets test in Australia. In 2008, 56% of all Australian retirees received a maximum rate public age pension. Therefore, 44% of retirees were reduced in their public age pension. 82% of these retirees were reduced in public age pension because they did not passed the fixed income test, 18% did not pass the assets test.
B.4. Differences between the Dutch and foreign pension systems
The first difference between the Dutch pension system and the foreign pension systems are the
differences in legal pension ages. Australia and the UK have different pension ages for men and women, but this will be equalized in the future. The UK and Australia have similar pension age for men compared to the Netherlands and the US has currently a pension age of 66. The pension age will not deviate much in the future between the four countries. The structure of the pension systems also deviate from each other. The US has multiple pillars, which are social security6 and individual retirement accounts such as IRA and 401(k) plans. Social security is based on historic income of the retiree, which differs from the Dutch pension system. The UK divided the public pension system into two components; a flat rate basic pension and an income related pension. These two components are complemented by a large voluntary private pension sector. This system is similar to the Dutch pension system, except that the average flat rate basic pension is much lower compared to the Netherlands. In the UK, retirees have to rely more on earning-related additional pension benefits compared to the Netherlands. Australia has multiple components in its pension system. The first component is a public age pension, which is funded by general taxation revenues. The Netherlands has a PAYG system for its public age pension. This causes Australia to score considerably higher on the FS Index because general taxation revenues do not depend on the distribution of the population. The second component is a compulsory employer contribution to
6
12 | P a g e
private superannuation guarantee named as the superannuation guarantee in Australia. This component is almost the same in the Netherlands and is most of the time also compulsory in the Netherlands. The third and last component is a voluntary superannuation contribution of the employee and other private savings. This can be compared to the third pillar in the Netherlands, which is an individual pension insurance that everyone can close down voluntarily.
Ezra (2011) examined the differences between the Dutch and US pension system and noticed that roughly 60% of the employees in the US belong to a pension plan, in the Netherlands roughly 95% do. The total Dutch occupational pension assets are 134% of GDP and this is 104% of GDP in the US. In the Netherlands, 94% of the total pension plans are Defined Benefit (DB) plans. This is currently between 43% and 57% in the US. Ezra (2011) argues that there are two main reasons why the Dutch can manage these DB benefits: by a realistic regulatory regime and a culture of cooperation.
C. Asset Liability Management (ALM)
13 | P a g e
of pension plans are DB plans (Ezra, 2011), which indicates that pension funds ensure a certain amount of supplementary pension benefits to their clients. This means that liabilities should be at the center of the decision-making process and performance and risk should be measured relative to liabilities (Berkelaar and Rouwenberg, 2010). Pension funds should not be obsessed by short-term asset-only performance but should move to ALM instruments that meet the retirement needs of their current clients, who retire in the future.
II.
Methodology
The core of my analysis is based on a model that takes into account the two major sources that Dutch people have with respect to their pension benefits, namely, AOW and supplementary benefits.
A. Criteria
The sustainability of the Dutch pension system is evaluated with respect to several criteria. In order to evaluate the sustainability of the Dutch pension system, I use the following 3 criteria.
1. Pension premium as percentage of the average income.
The first criterion is the premium as a percentage of the average income in the Netherlands. This is a criterion of sustainability since it measures the premium relative to average income, which means that future factors, such as growth of average income, are taken into account. A future labour force can pay more pension premiums in absolute terms because of the growth of their average income. This criterion is important to examine, because it shows insight into how much premium is paid nowadays and how much there must be paid in different scenarios in the future. If, for example, total pension premium equals 20% of the average income in 2010 and certain scenarios show that total pension premiums increase to 40% of the average income in the future, the Dutch pension system is not sustainable in its present form, since this is a large part of income, which leaves little room for paying other taxes and basic consumption needs.
2. AOW and supplementary pension benefits relative to average income.
14 | P a g e
3. Funding ratio
This third criterion is only applicable to supplementary pension. The funding ratio compares the value of assets with (future) liabilities. The funding ratio is the ratio of the terminal value of pension premiums relative to the present values of pension benefits at one future point in time. This is the same as comparing the total sum of assets built up before retiring with the present value of the total sum of supplementary pension benefits at the start of the retirement period.
B. Scenarios
In order to evaluate the sustainability of the Dutch pension system, I will examine the outcomes of the model under different scenarios. First of all, I will examine four possible retirement years, namely, 2020, 2030, 2040 and 2050. This refers to the situation that a person retires in the year 2020, 2030, 2040 or 2050. Evaluation of multiple retirement years will show insight into how the criteria can change over time. Different scenarios are the result of external factors that can deviate from forecasts of these external factors. The following scenarios are examined:
1. Legal pension age.
I will examine 3 legal pension ages, namely, 65, 66 and 67. With respect to the different time periods, there are 12 outcomes for each criterion. If someone retires at age 66 or 67, that person pays, respectively, 1 or 2 years extra supplementary pension premium to the pension fund compared to a person who retires at age 65. Therefore, someone who retires at age 66 or 67 pays, respectively, 41 and 42 years supplementary pension premium. With respect to AOW pension, a pension age of 67 results in additional AOW pension premiums from persons of 65 and 66. Furthermore, these persons will not receive AOW pension benefits if a pension age of 67 is applicable.
2. Return on investments.
Return on investments is only applicable to the supplementary pension. The supplementary pension premiums are invested by pension funds, which results in an average return. In the period 1980 to 2009, the Dutch pension funds achieved an average nominal return of 8.2% on investments7. Therefore, supplementary pension premiums paid before 2010, are invested with a yearly return of 8.2%. A return of 8.2% could be a good benchmark return to use for a normal return in the period after 2010. However,
7
15 | P a g e
according to Goudswaard et al. (2010), this is not true. They suggest a nominal return of 6%8 for at least the upcoming 15-20 years. This is mainly caused by low long-term interest rates, for example,
government bonds. Therefore, the normal return on investments after 2010 is set at 6%. Furthermore, I will examine average returns on investments of 8% and 4% after 2010. A future return on investments of 8% is considered to be optimistic, a future return on investments of 4% is considered to be pessimistic.
3. Inflation
This set of scenarios examines the influence of inflation on the criteria. The inflation scenario is applicable to both AOW and supplementary pension. This scenario shows the risks of unequal
indexation ambitions between the average income and pension benefits. According to the figures of the Centraal Bureau voor de Statistiek (CBS), the Netherlands had an average inflation rate of 1.84% over the last 10 years (period 2002-2011) and 2.31% inflation in the period 1990 to 2011. Furthermore, the European Central Bank (ECB) strives to inflation close to 2%. This suggests that 2% inflation is a good approximation as long term percentage during normal economic conditions. Furthermore, 1% inflation is taken as low inflation and 3% is taken as high inflation. These percentages are chosen because it
concerns long term inflation rates. Inflation rates for individual years can be a lot higher or lower. For example, in 1975, the Netherlands experienced an inflation rate of 10.2% and in 1987 it experienced a deflation rate of 0.5%. However, since the scenarios of long term high inflation rates of 10% and
deflation rates of 0.5% are very unlikely, inflation rates of 1% and 3% are chosen. For the calculations of the scenario inflation rate, I assume that the average income still growths with a normal inflation. On the other hand, pension premiums and benefits are indexed with low or high inflation. In practice, most of the time the income is indexed with realized inflation. Therefore, the scenarios of different levels of inflation are hypothetical but it gives insight into a situation when the average income is not indexed with realized inflation.
4. Mortality rates
The last set of the scenarios regards the uncertainty related to mortality rates. Pension funds use
mortality rates to forecast future liabilities. This set of scenarios is examined differently for the AOW and supplementary pension. For the AOW, a 10% deviation is used compared to forecasted mortality rates provided by the Actuarieel Genootschap (AG). A 10% deviation is best explained by an example. The AG
8
16 | P a g e
forecasted that someone of 96 years has a probability to die of 30% in that year. Scenarios are examined that this probability will be 27% or 33% in the future. In this example, there is a 10% deviation between the forecasted probability and the actual probability to die for a 96 year old person. The scenarios, with respect to supplementary pension, are examined on an individual basis. Therefore, mortality rates regarding supplementary pension is examined differently compared to AOW pension. For the calculation of supplementary pension benefits, the average age is taken for each time period that is examined. Therefore, if mortality rates are lower, I count one additional year to the average age a person becomes. When mortality rates are higher compared to forecasted mortality rates, I subtract one year from the average age a person becomes. See appendix A for an overview of the forecasted average age.
C. Evaluation criteria supplementary pension
The criteria of 2010 are solely based on historic figures and can be regarded as benchmark figures to compare future scenarios with. The first criterion, premium as percentage of average income, is
calculated by comparing the total amount of supplementary pension premiums with the total amount of income from employees on a yearly basis. Both amounts are divided by the number of employees. This results in the calculation of the individual supplementary pension premium and the average income of an employee in 2010. The average supplementary pension premiums paid in the period 2002-2010 are taken to calculate the average percentage of supplementary pension premium paid relative to average income in 2010. This period is taken as the benchmark period, because only after 2001, there is a catch-up of scatch-upplementary premiums. Therefore, I indicate this period as the best benchmark period. The second criterion for 2010, pension benefits as a percentage of average income, is calculated by the same method as the future scenarios. This criterion measures how much supplementary pension benefits a person receives with an age of 65 in 2010 and receives benefits until the age 859. The funding ratio is the last criterion and is the minimum acceptable funding ratio in 2010, namely 105%.
Before I explain the calculation of future scenarios of 2020, 2030, 2040 and 2050 of the first criterion, it is important to understand the implication of the figures of the first criterion, premium as percentage of average income. For example, the figures of 2020 for the first criterion, which refer to the set of
scenarios in that a person retires in the year 2020, show how much supplementary pension premium relative to average income must be paid between 2010 and 2020. The figures of 2030 show how much supplementary pension premium relative to average income must be paid between 2010 and 2030. In
9
17 | P a g e
the first case the figure is an average over 10 years, the second case gives an average figure over 20 years. Consequently, the figures of 2040 and 2050 that refer to the set of scenarios that a person retires in the year 2040 or 2050 are average figures over, respectively, 30 and 40 years. The calculation of future scenarios with respect to the first criterion consists of 3 components.
I How much supplementary pension premium is there paid until 2010 II How much supplementary pension benefits must be paid after retiring
III How much supplementary pension premium must be paid between 2010 and year of retiring to match future liabilities
C.1. Calculation supplementary pension premium payments until 2010 (1st component) To calculate the first component, I have to calculate the age specific supplementary pension premium. The CBS published the average income between 2001 and 2010 in age classes of 5 years. For example, it shows the total income of persons between the age of 25 to 30 and 30 to 35. I calculated for each age class how much (in percentages) the total income of an age class contributes to the total income of all persons between the age of 25 and 65. This percentage is taken from the total amount of
supplementary pension premiums to calculate an age specific supplementary pension premium. The data of the CBS, with respect to income in age classes, is available from 2001. Therefore, the income distribution that is applicable for 2001 is used also for the years before 2001. The data of total supplementary pension premiums paid to pension funds are published by De Nederlandsche Bank (DNB). The data of DNB concerns total supplementary pension premiums from 1997 to 2010. To calculate the amount of total supplementary pension premiums paid to pension funds before 1997, I calculated the average growth of the total supplementary pension premiums between 1997 and 2001. Only these years are taken into account because there was a significant increase in total supplementary pension premiums in 2002. I consider this significant increase as a one-time event and, therefore, it is not taken into account to calculate past supplementary pension premiums. The amount of
supplementary pension premiums paid to pension funds in 1997 is indexed backwards to 1970 with the average percentage of growth between 1997 and 2001. If, for example, a person can retire at age 65 in 2020, that person is 25 years old in 198010. Therefore, I need to know the age specific pension premium paid by a 25 year old in 1980. This has to be taken into account because, on average, a 25 year old
10
18 | P a g e
person does not pay as much supplementary pension premium as a person of 60 years old does, because the 25 year old is at the start of his career and the 60 year old at the end of his career. This is important to notice because, with a pension age of 65, the premiums paid by a 25 year old can be invested for 40 years and the premiums of a 60 year old only for 5 years. The terminal value of supplementary pension premiums paid by a 25 year old person is, therefore, higher compared to the terminal value of premiums of a person who is 6o years old.
(1)
Formula (1) shows the calculation of the terminal value of a one year supplementary premium payment. This is the present value in 2010 of a one year premium payment before 2010. In this formula is the yearly supplementary pension premium, refers to the year when the premium is paid, refers to the last historic year of the model, namely 2010, and 0.082 is the average return on investments achieved by the pension fund until 2010.
{ ∑ [ ]
} (2) Formula (2) shows the sum of each individual terminal value of one year supplementary premium payments for a person who starts to pay premiums in 1991 and can retire at age 65 in 2030. The first year of premium payment is indicated by the lower bound of summation, , which refers to a premium payment in 1991. The upper bound of summation is the premium paid in 2010, indicated by
. Calculations with a legal pension age of 66 and 67 results in lower bounds of summations of, respectively, and for a person who can retire in the year 2030. If, for example, a person can retire in 2020 and the legal pension age is 65, 66 or 67, the lower bounds of summations are,
respectively, , and . For reasons of convenience in further calculations, the sum of individual terminal values of pension premiums until 2010 is replaced by X.
C.2. Calculation amount of supplementary pension benefits (2nd component)
19 | P a g e
future average AOW pension benefits are calculated by taking average AOW pension benefits of 2010 as starting point and the amount is yearly indexed with 2% inflation and 1.5% GDP. The average age a person becomes is used to calculate the number of years that supplementary pension benefits are paid. The Dutch pension funds take also the average age a person becomes to calculate their future liabilities (Bikker et al. 2009).
(3)
Formula (3) shows the calculation of the present value of a one year supplementary pension benefit. The present value refers to the value of one year pension benefit at the start of the retirement period where Z is the amount of pension benefits to pay in one year, refers to the year when the pension benefits must be paid, R is the return on investments and refers to the year at the start of the retirement period.
∑ [ ]
(4)
Formula (4) shows the calculation of the present value of the total sum of supplementary pension benefits that has to be paid to one person. In other words, it is the present value of the total sum of supplementary pension benefits at the start of the retirement period. The lower bound of summation is the start of the retirement period indicated as , the upper bound of summation is the last year payment of pension benefits, indicated as T. The upper bound is determined by the average life expectancy (see appendix A).
C.3. Calculation supplementary pension premiums between 2010 and year of retirement (3rd
component)
After determining the amounts of components one and two, the amount of supplementary pension premium can be calculated that must be paid between 2010 and the year of retirement.
(5)
20 | P a g e
retirement period where P is the pension premium, refers to the year when the premium is paid, R is the return on investments and refers to the year at the start of the retirement period.
∑ [ ] (6) Formula (6) shows the total sum of the terminal value of premiums. This is the total amount of money saved before the retirement period, between 2011 and the retirement period. Consequently, the lower bound of summation is , which refers to supplementary pension premium payment in 2011. The upper bound of summation is the start of the retirement period, indicated by .
C.4. Matching assets with future pension benefits (liabilities)
The total amount of assets has to be matched with future liabilities. The assets are the terminal values of all paid supplementary pension premiums at the start of the retirement period. The terminal values of all premiums indicate a present value at a future point in time. The future liabilities are the present value of the total supplementary pension benefits at the start of the retirement period.
{[ ] ∑
} ∑ [
]
(7)
Equation (7) shows that the assets built up before the retirement period must match the present value of the total amount of supplementary pension benefits. Equation (7) shows that amount X (see formula (2)) can be invested for - years, which is indicated as the period between 2011 and the start of the retirement period. This is the first part of the assets built up before retirement. The second part consists of the total sum of the terminal value of premiums paid between 2011 and the retirement period, which is indicated by formula (6). The sum of these two parts must be equal to the present value of the total amount of benefits, which is indicated by formula (4).
Many calculations of the first criterion are equal to calculations for the second criterion. The second criterion measures the generosity of the pension system by calculating the percentage of supplementary pension benefits relative to the average income. In the period between 2002 and 2010, supplementary pension premiums equaled almost 10% of average income. Therefore, in calculations of the second criterion, a person pays supplementary pension premiums equal to 10% of the average income. Since premiums until 2010 are historic values, these premiums will not change anymore. The supplementary pension premiums that were paid before the year 2001 were structural lower compared to the
21 | P a g e
2010 to equal an average paid premium of 10%. Appendix B shows the average percentage of
supplementary pension premium to pay, between 2010 and the year of retirement, to equal an average premium of 10%. Calculations of the terminal value of supplementary pension premiums are further equal to formula (2) and (6), with the exception that average premiums equal 10% of the average income. The terminal value of supplementary pension premiums at the start of the retirement period is now leading to determine how much supplementary pension benefits can be paid to retirees.
∑ [ ] (8) Equation (8) shows the calculation of how much supplementary pension benefits, relative to average income, there can be paid to retirees. The symbol Y is the total terminal value of supplementary pension premiums at the start of the retirement period, T is the upper bound of summation, which is the last year payment of supplementary pension benefits, refers to the year when the benefits must be paid,
refers to the year at the start of the retirement period, p is the percentage relative to average income and V is the average income of employees in a certain year. The maximum number of years between the year of retirement and payment of supplementary pension benefits ( - ) is determined by the average life expectancy (see appendix A).
The calculations of the last criterion, the funding ratio, are, with a few exceptions, similar to the
calculations of the first two criteria. The calculation of the total terminal value of supplementary pension premiums at the start of the retirement period is based on supplementary pension premiums of 2010 that are indexed with 2% inflation and 1.5% GDP growth. The calculation of supplementary pension benefits is based on 70% of the average income of employees minus the AOW pension benefits. The year 2010 is, also the for average income and AOW pension benefits, taken as starting point and indexed with 2% inflation and 1.5% GDP growth. All calculations after 2010, with respect to the criterion funding ratio, are calculated with a return on investments equal to 6%. The total terminal value of
supplementary pension premium relative to the total present value of supplementary pension benefits results in the funding ratio. The total terminal value and present value are, again, based on one point in time, which is the start of the retirement period.
D. Evaluation criteria AOW pension
22 | P a g e
the AOW pension system have to be group-based because the generosity of AOW pension benefits depends on the amount of AOW pension premiums and vice versa. Calculations of the two criteria for AOW pension are straight forward, because I assume that the total amount of AOW pension benefits and the total amount of AOW pension premiums are paid in the same year. For 2010, the first criterion is calculated by taking the total amount of AOW pension premiums relative to the total income of employees and people that are currently unemployed. For the calculation of this ratio, the incomes of people that are currently unemployed are also taken into account, because these people also pay AOW pension premium. The second criterion for 2010, AOW pension benefits as percentage of average income, is the average amount of AOW pension benefits of 2010 relative to the average income of 2010. The average amount of AOW pension benefits in 2010 is the total amount of AOW pension benefits divided by the total number of people that are 65 years and older. The data of total collected AOW pension premiums, and payment of AOW pension benefits in 2010, are provided by the annual report 2010 of the Sociale Verzekeringsbank (SVB).
The calculation of the first and second criterion, for the years 2020, 2030, 2040 and 2050, are based on different assumptions. The calculation of the first criterion is based on the fact that future retirees receive (relative) equal AOW pension benefits compared to retirees in 2010. This means that people, who retire in 2040, receive an equal percentage of average income of AOW pension benefits compared to retirees in 2010. Consequently, employees have to pay AOW pension premiums in 2040 that equal the total amount of AOW pension benefits to pay in 2040. The scenarios of the second criterion are based on the fact that the employees pay an equal percentage of AOW pension premiums, relative to the average income in 2040, compared to employees in 2010. Consequently, retirees of 2040 receive average AOW pension benefits equal to the total amount of AOW pension premium paid in 2040 divided by the number of retirees in 2040. Therefore, the first criterion takes as starting point relative equal future AOW pension benefits compared to 2010 and the second criterion takes as starting point relative equal future AOW pension premiums compared to 2010.
III. Data description
23 | P a g e
A. Collection of data
First, the data with respect to expected mortality rates of the Netherlands were collected. This data is published by the AG. The AG published in August 2010 their newest expected mortality rates of the Netherlands in the report: Prognosetafel 2010-2060. This report takes into account the latest available data provided by the CBS. The AG published in 2007 the report ‘Prognosetafel 2005-2050’. After 2007, the CBS published that real mortality rates were lower caused by medical innovations and behavioral changes. Following this information, the AG developed a new forecasting model to predict mortality rates in the Netherlands. The report Prognosetafel 2010-2060 takes into account this latest available data. In the appendix of the report, the forecasted mortality rates for men and women in the
Netherlands are published. Mortality rates are published separately for men and women because, overall, women have lower mortality rates. Furthermore, mortality rates are published yearly and in an age range of 0 to 120. All mortality rates of men and women are used in the datasheet. I take the average of both sexes to calculate the mortality rate of an average person in the Netherlands. One minus the mortality rate is called the chance of surviving a particular year. The chance of surviving is used to calculate how the population pyramid of the Netherlands will look like in the future.
After collecting the data of mortality rates, I collected the data of how the real population structure of the Netherlands looks alike. The latest available data is from 2011, provided by the CBS. The CBS
provides data of the number of men and women of every age between 1 and 100 years. The year 2011 is taken as a starting point to forecast the population structure of the Netherlands. To determine the population structure after 2011, I take the chance of surviving for each age to calculate how many persons there are left next year. One problem was that, for example, I could not calculate how many 8-year olds there are in 2025. This is solved by taken the 8-yearly number of every future new born also provided by the CBS. They provide how many babies there are expected to be born yearly, until the year 2060. With all this data I can structure the present and future population structure of the Netherlands. The future population structures are solely calculated by taken into account the future mortality rates of the Netherlands. This means that other sources that affect the population structure are not taken into account, for example, emigration and immigration of people.
24 | P a g e
classes of future retirees, namely, above 65, 66 and 67 year. I assume that persons of 25 years old begin to pay pension premiums. Consequently, there are three groups of employees, namely, 25 to 65, 25 to 66 and 25 to 67.
B. Historic and forecasted population structure
An examination of the historic and future population structure is important with respect to the AOW pension. The Dutch AOW is a PAYG system, which means that the current employees pay the AOW pension benefits of the current retirees. Therefore, the future distribution of the number of employees relative to the number of retirees (dependency ratio), gives important implications of the sustainability of the AOW pension system.
Table I: Historic figures population of the Netherlands
Year 1997 2000 2004 2007 2010 Number of people between 25 - 65 years 8,482,890 8,658,132 8,891,740 8,961,831 9,068,947
Number of people 65 years and older 2,083,839 2,152,442 2,251,154 2,368,352 2,538,328
Dependency ratio 24.57% 24.86% 25.32% 26.43% 27.99%
Source: www.cbs.nl
Table I shows the historic figures of the number of employees and retirees from 1997 until 2010. It shows that the group of retirees grew with more than 450,000 persons between 1997 and 2010. The labour force grew with approximately 600,000 people between 1997 and 2010. This grow in number of employees ensured that the dependency ratio only grew from 24.57% in 1997 to 27.99% in 2010. Table II: Overview of future expected number of employees and retirees and the average income.
Number of employees (25-67) Number of retirees (67>) Average income employees Year of retirement 2010 9,452,000 2,239,000 € 34,707 2020 9,285,804 3,083,056 € 48,958 2030 8,868,173 4,003,955 € 69,060 2040 8,155,923 4,713,252 € 97,415 2050 7,985,009 4,579,542 € 137,414
25 | P a g e
Chart I: Expected population in the Netherlands from 2011 to 2060.
Chart I shows the development of the number of employees and retirees from 2011 to 2060. This chart shows directly two threats that the AOW pension system faces. The first threat is the growth in the number of retirees until the year 2040. The growth of retirees is mainly caused by the birth wave after World War II. After 2040, this group will (for the most part) not live anymore and the group retirees will steadily decline. The second threat is the decline in number of employees in the upcoming decades. The number of employees declines sharply until the year 2040. After the year 2040, the number of
employees also decline, but not as sharp as before 2040. This chart shows that a worst case scenario of the AOW pension can be expected in the year 2040 (see also Krakes and Broeders, 2006).
C. Historic macro figures of Dutch pension funds
For the supplementary pension system it is important to analyze the historic development of the amount of supplementary pension premiums and benefits paid.
Table III: Historic macro figures Dutch pension funds (amount in million euros).
Year 1997 2000 2004 2007 2010 Total sum of supplementary pension premiums paid to pension funds € 7,001 € 10,405 € 22,399 € 24,883 € 28,660
Total sum of supplementary pension benefits paid by pension funds € 9,070 € 11,692 € 16,438 € 20,539 € 24,232 Source: www.DNB.nl 0 1000000 2000000 3000000 4000000 5000000 6000000 7000000 8000000 9000000 10000000 20 11 20 14 20 17 20 20 20 23 20 26 20 29 20 32 20 35 20 38 20 41 20 44 20 47 20 50 20 53 20 56 20 59 Nu m b er of p eo p le Year
26 | P a g e
Table III shows the development of total supplementary pension premiums paid to Dutch pension funds and the amount of supplementary pension benefits paid out over years. The table shows a big increase of premiums, namely, a quadrupling of premiums paid in 2010, compared to 1997. The amount of premiums increased with an average of 12.08% between 1997 and 2010. The biggest increase was in the year 2002. In 2002, total supplementary pension premiums increased from 12.2 billion in 2001 to almost 18 billion in 2002. This is an increase of 47% in total premiums. This increase was (partly) due to the internet bubble of 2001. High return on investments resulted in relatively low supplementary pension premiums to pay before 2001. When the bubble bursted in 2001, Dutch pension funds felt they needed to raise premiums largely in 2002. The total sum of supplementary pension benefits increased also spectacular between 1997 and 2010. The benefits increased with a yearly average of 7.9% between 1997 and 2010. These results are important to take into account when different scenarios are examined. For example, a person who will retire in the year 2020. It is important to examine the development of average supplementary pension premium because that person started to pay premiums already in 198011.
IV. Results
In this section, the results will be discussed with respect to sustainability of the Dutch pension system. The following three criteria were established to measure future sustainability of the Dutch pension system: premium as percentage of the average income, pension benefits relative to average income and funding ratio. The criterion funding ratio is only applicable to the supplementary pension. Table IV and V show the figures of criteria that are applicable to 2010. These figures are used as benchmark figures and can be compared to the figures in future scenarios.
Table IV: Figures supplementary pension 2010.
Premium as percentage of average income (2002-2010) Pension benefits as percentage of average income Minimum acceptable funding ratio Supplementary pension 9.71% 20.65% 105.00% 11
27 | P a g e
Table V: Figures AOW pension 2010.
Premium as percentage of average income in 2010 Pension benefits as percentage of average income AOW pension 10.24% 36.43%
The first criterion, applicable to supplementary pension in 2010, shows that a person had to pay a supplementary pension premium equal to 9.71% of his average income. Retirees received on average supplementary pension benefits equal to 20.65% of the average income in 2010, which is the second criterion. The third criterion is funding ratio, which is 105% in 2010. The figures of table V show that employees had to pay, on average, AOW pension premiums equal to 10.24% of the average income in 2010. Retirees received, on average, AOW pension benefits equal to 36.43% of the average income in 2010.
A. Results first criterion supplementary pension
The first criterion is the percentage of supplementary pension premium to pay relative to average income. I assume that all retirees receive total pension benefits equal to 70% of average income. Consequently, the pension fund has to pay supplementary pension benefits to retirees equal to 70% of average income minus the average AOW pension benefits paid out. When I examine the first set of scenarios, regarding pension age, the premiums have to be more than doubled in the retirement year 2020, compared to 2010, if the pension age is held at the age of 65. If the pension age is 67 in
Table VI Total overview of impact on criteria supplementary pension in different scenarios
Criteria supplementary pension
Premium as a percentage of average income
(1st criterion) Pension benefits as a percentage of average income (2nd criterion) Funding ratio (3rd criterion)
Year 2020 Year 2030 Year 2040 Year 2050 Year 2020 Year 2030 Year 2040 Year 2050 Year 2020 Year 2030 Year 2040 Year 2050 Pension age 65 21.00% 12.50% 9.45% 9.10% 35.30% 28.88% 39.02% 36.96% 81.75% 99.32% 145.26% 120.80% 66 18.50% 11.30% 8.65% 8.48% 37.22% 30.30% 41.52% 39.67% 86.76% 105.33% 156.58% 130.09% 67 16.00% 10.10% 7.85% 7.85% 39.42% 31.85% 44.25% 42.30% 92.03% 111.64% 168.54% 139.69% Return on investments Low 4% 31.70% 21.30% 16.50% 14.68% 28.50% 21.80% 25.75% 23.08% 65.63% 70.20% 119.38% 76.79% Normal 6% 16.00% 10.10% 7.85% 7.85% 39.42% 31.85% 44.25% 42.30% 92.03% 111.64% 168.54% 139.69% High 8% 3.70% 1.85% 2.01% 4.08% 54.00% 46.20% 77.50% 79.80% 127.81% 177.29% 243.45% 261.23% Inflation Low 1% 9.30% 5.55% 4.57% 5.62% 47.50% 40.50% 60.00% 58.45% 109.10% 141.65% 229.48% 192.10% Normal 2% 16.00% 10.10% 7.85% 7.85% 39.42% 31.85% 44.25% 42.30% 92.03% 111.64% 168.54% 139.69% High 3% 23.40% 15.30% 11.78% 10.84% 32.40% 25.00% 32.85% 31.00% 77.61% 88.27% 124.82% 102.83% Mortality rates Minus 1 year 14.20% 9.30% 7.40% 7.57% 41.00% 33.10% 46.00% 43.82% 95.93% 116.06% 175.21% 144.86% As forecasted 16.00% 10.10% 7.85% 7.85% 39.42% 31.85% 44.25% 42.30% 92.03% 111.64% 168.54% 139.69% Plus 1 year 17.70% 10.80% 8.30% 8.12% 37.90% 30.80% 42.70% 40.95% 88.53% 107.66% 162.53% 135.00%
compared to the persons that retire in the year 2030.The second reason is that the persons who retire in 2020 (only) have 10 years to compensate the lower supplementary pension premiums paid in the past. The persons who retire in 2030 have 20 years to compensate the lower premiums paid in the past. Retirement in the years 2040 and 2050, results in premiums that are below 10% of the average income. For persons retiring at an age of 67 in the retirement years 2040 and 2050, the premium is 7.85%. This is considerably lower compared to the 9.71% in 2010. Figures of 2040 and 2050, with respect to a change in pension age for the first scenario, show that a one year increase of legal pension age can decrease the average supplementary pension premium to pay, between 0.8% and 0.6% of the average income. The second scenario examined is the influence of return on investments on the first criterion. The results of the second scenario are based on the assumption that the pension age is 67. Therefore, the results with a normal return on investments (6%) are similar to the results for the first criterion
regarding a legal pension age of 67. A scenario of return on investments equal to 4% in the period 2010 to 2020, results in average supplementary pension premiums of 31.7%. This is a tripling compared to the premiums paid in 2010. On the other hand, in an optimistic scenario with a return on investments of 8% in the period 2010 to 2020, 3.7% of the average income has to be paid as premium. These results show that the return on investments has a huge impact on the first criterion. The huge impact is also shown in further retirement years, but to a lesser extent. For example, retirement in the year 2050 results in supplementary pension premiums of 14.68% and 4.08% relative to the average income. These results are applicable with returns on investments of, respectively, 4% and 8%.
30 | P a g e
The last scenario shows the impact of mortality rates on the first criterion. A person, who is 65 years old in the year 2020, lives on average 21 years after he turns 65. Therefore, that person becomes on
average 86 years old (see appendix A). If the average age is 85 years, then as a result, supplementary pension premiums can be lowered with 1.8% to 14.2% of average income in 2020. An average age of 87 years increases premiums with 1.7%, to 17.7% of the average income in 2020. In other future retirement years, the differences between forecasted and 1-year deviations from forecasted figures become even smaller. For example, in the retirement year 2050, the premiums can decline with 0.28% relative to the average income if the average age becomes 1 year shorter. The supplementary pension premiums must rise with 0.27% if the average age is extended with 1 year in 2050. These results show that the set of scenarios of 1-year deviations from the forecasted average age does not have a big impact on the first criterion.
B. Results second criterion supplementary pension
31 | P a g e
The figures of the first criterion showed that the set of scenarios, which refer to return on investments, has a huge impact on the first criterion. The huge impact of return on investments is also shown in the generosity of supplementary pension benefits. The percentage of benefits, relative to average income, varies between 28.5% and 54% with, respectively, a return on investments of 4% and 8% in the period between 2010 and 2020. Other future retirement years also show a huge deviation in generosity of benefits. For example, the retirement years 2040 and 2050 show that an 8% return on investments results in benefits of, respectively, 77.50% and 79.80%. Supplementary pension benefits are,
respectively, 25.75% and 23.08% of the average income if the average return on investments yields 4% in 2040 and 2050.
The set of scenarios, which refer to inflation, show also a big impact on the second criterion. The retirement year 2020 shows that an inflation of 1% results in supplementary pension benefits equal to 47.50% of the average income. On the other hand, an inflation of 3% results in benefits equal to 32.40%. The inflation set of scenarios shows also the highest generosity in the retirement year 2040.
Supplementary pension benefits, equal to 60.00% of the average income, are paid out if the average inflation is 1% between 2010 and 2040. An average inflation of 3% results in benefits equal to 32.85% between 2010 and 2040.
The last set of scenarios examined for the second criterion is mortality rates. A subtraction of 1 year in average age leads to small increases in generosity of supplementary pension benefits. The retirement years 2020, 2030, 2040 and 2050 show increases in benefits of, respectively, 1.56%, 1.25%, 1.75% and 1.52% relative to average income, if the average age is reduced with one year. An increase of 1 year in average age shows similar results. For example, in the retirement years 2040 and 2050, benefits decrease with, respectively, 1.55% and 1.35% if 1 year is added to the average age.
C. Results third criterion supplementary pension
32 | P a g e
2020. The funding ratio exceeds the minimum funding ratio only, if the pension age is 66 and 67, in the retirement year 2030. A pension age of 65 results in a funding ratio of 99.32% in the retirement year 2030. The retirement year 2040 shows the highest funding ratios, which is a result of the premiums, paid after 2001, that are invested for 9 years, with a return on investments of 8.2%. A pension age of 65 still results in a funding ratio of 145.26% in 2040. The funding ratio increases to 168.54% if the pension age is 67 in the retirement year 2040.The funding ratios decline to 120.80% and 139.69% in 2050, with pension ages of, respectively, 65 and 67. This can be explained by the fact that all supplementary pension premiums are invested with 6% return on investments, compared to 8.2% return on investments before 2010.
The second set of scenarios shows that an average return on investments, equal to 4%, results in
funding ratios below 105% in the retirement years 2020, 2030 and 2050. Retirement year 2040 results in a funding ratio of 119.38% with an average return on investments of 4%. Apparently, the (high)
premiums paid before 2010 and the return on investments of 8.2%, compensate the low return on investments, to result in a funding ratio of 119.38%. The results with a return on investments of 8% show excessively high funding ratios. For example, the retirement years 2040 and 2050 show funding ratios of, respectively, 243.45% and 261.23%. These results show that the current supplementary pension premium level is more than sufficient to guarantee total pension benefits equal to 70% of average income, if the future average return on investments equals 8%12.
The set of scenarios regarding inflation show also a big impact on the funding ratio. The retirement year 2020 shows an increase in funding ratio of 17.07% in a scenario of 1% inflation compared to the scenario of 2% inflation. On the other hand, the funding ratio will decrease with 14.42% if the average inflation equals 3% in 2020. All funding ratios, with respect to the retirement year 2040, are above 105%. An average inflation of 3% results in a funding ratio equal to 124.82% in 2040. An average inflation of 1% results in a funding ratio equal to 229.48% in the retirement year 2040.
The last set of scenarios examined for the third criterion is mortality rates. The results of the first two criteria showed that a change of 1 year in average age have a small impact on these criteria. These results are also observable for the third criterion. Funding ratios decline between 3.5% and 6% if the
12
33 | P a g e
average age increases with 1 year and funding ratios rise between 3.9% and 6.7% if the average age is reduced with 1 year.
D. Results first criterion AOW pension
The first criterion examined for AOW pension is the AOW premium as percentage of the average income of employees. In 2010, retirees received, on average, AOW pension benefits equal to 36.43% of the average income. Calculations of the first criterion are, therefore, based on the assumption that all future retirees receive AOW pension benefits equal to 36.43% of average income. In 2010, AOW premium as percentage of average income is 10.24%. All scenarios show that AOW pension premiums will have to rise in the future, compared to the premiums paid in 2010. The results show a direct link with the future population structure of the Netherlands shown in chart 1. Chart 1 shows a sharp increase in the number of retirees and a decrease in number of employees, after 2010.
Table VII: Total overview of impact on criteria AOW pension in different scenarios
Criteria AOW pension
Premium as percentage of average income (1st
criterion)
Pension benefits as percentage of average income (2nd criterion) Year 2020 Year 2030 Year 2040 Year 2050 Year 2020 Year 2030 Year 2040 Year 2050 Pension age 65 14.37% 19.56% 24.07% 23.74% 25.97% 19.08% 15.50% 15.71% 66 13.19% 17.94% 22.55% 22.27% 28.28% 20.79% 16.54% 16.75% 67 12.10% 16.45% 21.05% 20.89% 30.84% 22.68% 17.72% 17.85% Inflation Low 1% 10.98% 13.55% 15.73% 14.17% 33.98% 27.54% 23.71% 26.33% Normal 2% 12.10% 16.45% 21.05% 20.89% 30.84% 22.68% 17.72% 17.85% High 3% 13.32% 19.94% 28.09% 30.70% 28.01% 18.71% 13.28% 12.15% Mortality rates Lower -10% 12.36% 16.93% 21.75% 21.67% 30.18% 22.04% 17.15% 17.22% As forecasted 12.10% 16.45% 21.05% 20.89% 30.84% 22.68% 17.72% 17.85% higher +10% 11.85% 16.01% 20.42% 20.20% 31.49% 23.30% 18.27% 18.47%
