Dutch Pension Fund Watch – The latest insights on Dutch pension reforms

by: Jolien van den Ende

  • Dutch pension funds will move towards defined contribution schemes with different flavours
  • There will be a new contract and the already in use “Wet Verbeterde Premieregling”
  • For the latter, new options will be added
  • Under the new contract the contribution will be age independent and will benefit your own individual reserved capital pot
  • A projected return is used to calculate the contribution and the Dutch government already committed to fix the contribution between 30% to 33% until 2036…
    …based on this we indirectly derived the projected return which would be 1.5% to 1.8%
  • An obligated solidarity fund will be used to dampen shocks…
    …and is capped at 15% of the total fund size
  • We expect that the distribution of returns will take place via a two-step approach…
    …first the protection return, which is the protection related to interest rate risks, will be distributed among plan participants and then the excess return
  • Possible sticking points are the abolition of the average system and the legal risks related to the transition to the new contract
  • We judge that existing defined contribution schemes will be exempt, which reduces the likelihood that the abolition of the average system will break the deal
  • There are two valuation methods to transfer pension entitlement to the new contract:
    –     Value based-ALM
    –     Standard method
  • In both cases, the average expected replacement ratios are higher in the median scenario for most age cohorts even though it is a zero sum game
  • The default route would be to transfer pension entitlements to the new system
  • However, we judge that DC schemes will be exempt as the most affected cohorts would lose about 10% of their expected pension due to the abolition of the average system
  • We expect that the Dutch government will approve the deal before the end of this year
  • Finally, we expect pension funds to move to the new system between the second half of 2023 and 2025
ABN-AMRO-Dutch-Pension-Fund-Watch-The-latest-insights-on-Dutch-pension-reforms.pdf (496 KB)
Download

Introduction

On 22th of June, the Ministry of Social Affairs and Employment published the preliminary new pension agreement on the shape of the Dutch pension reform. The reform affects all pension savings under the second pillar of the Dutch pension system, which are roughly EUR 1.6 trillion. We already looked at the proposed Dutch pension fund reforms in a recent note (contact Jolien van den Ende for this). This note will give an update and will discuss our latest insights on the reforms. First, we will describe what the requirements are for the new system. Secondly, we will explain the different defined contribution contracts under the new system. Thirdly, we give an estimation of what the expected pension would be under different scenarios and how plan participants, who loose part of their pension, will be compensated. We judge that the latter in combination with the transition of the current pension entitlements to the new contract could be the main sticking points for the Dutch government. Finally, we describe the transition framework including the process of the reforms. We will publish a follow up note with more details about the possible impact on the asset allocation of pension funds as well as the interest rate hedges.

New pension system: more transparent and individual with welfare-proof perspective

The new pension system should increase the likelihood that participants will receive a welfare-proof pension. This implies that pay-outs will be indexed for inflation. Furthermore, it should become easier to index pensions in good times, but also to cut pensions in bad times. In addition, the system should become more personal as well as more transparent. Finally, the new system should better match developments in the society and the job market. In addition, multiple preconditions have been set in the pension agreement last year as discussed below.

Any disadvantages for plan participants should be compensated in a fair matter

Under the new system, contributions as well as the pay-outs should be as stable as possible. On the other hand, pension providers should be able to achieve investment returns. The transition to the new system should be balanced for all generations. If there are disadvantages because of the transition to the new system, then they should be adequately compensated. This is very important for the abolition of the “Doorsneesystematiek” under the new system which we will discuss later.

The goal is a maximum of 75% of the average salary in 40 years and 80% in 42 years

The new system will keep existing elements like the collective implementation, the solidarity between generations and the fact that it remains mandatory. On top of that, the Dutch government will give tax facilitation to the extent that the current target for the retirement pay-outs remains achievable. This is a maximum of 75% of the average salary in 40 years and 80% in 42 years.

All Dutch pensions will move towards defined contribution schemes with different flavours

Under the new system, a new pension contract will be introduced with collective risk sharing in the build-up phase and the distribution phase and additional solidarity options will be added to the currently already in place “de Wet Verbeterde premieregeling” as shown in the figure below.

The shift towards defined contribution schemes means that pensions in the Netherlands will become more uncertain. Moreover, a projected return will be used and the risk free rate (UFR) and the coverage ratio will disappear. This means, that interest rates only play a role with regard to the investment return under the new system. We will discuss the new contract with collective risk-sharing below.

New pension contract with collective risk sharing and an individual reserved pension capital

In the new solidarity pension contract, pension claims and liabilities for the pension funds disappear. There is one collective for former plan participants, retired plan participants and active plan participants. Moreover, only contributions will be certain, which will be age independent and benefit your own pension accrual. Consequently, the principle of ‘nominal certainty’ will be scrapped and therefore returns can be used to build up pensions instead of funding the buffer. The buffer will be replaced by a solidarity fund, which is capped at 15% of the total fund and will be used to dampen shocks (for more see below). Furthermore, risks will be shared collectively in both the contribution and the distribution phase and this prevents the poverty cap and big shocks in consumption. In addition, the new scheme is based on the judgement that extensive risk sharing has the advantage that returns are usually higher. Finally, the contributions will be invested based on ‘best effort’ and the ‘prudent person principle’, which means an implicit life-cycle investing approach.

The build-up and distribution phase: Contribution fixed between 30% to 33% until 2036

Plan participants have an individual share in the collectively-invested funds by the pension fund. Plan participants build-up pensions based on the contributions, the returns and the contributions from the solidarity fund, also known as the reserved pension capital. Moreover, the Dutch cabinet already committed to fix the contribution between 30% to 33% to give certainty to employers and plan participants and this will be applicable until 2036. However, if a shock of at least 5%-points occurs, then the contribution would be adjusted before 2036. Hence, without a shock the first time that the contribution would be adjusted is after the compensation period (see transition framework below).

The contribution will be used to generate returns, which will be accrued to the plan participants’ “individual” reserved pension capital on a yearly basis. Based on the reserved pension capital and a projection method, an expected pension can be estimated for each plan participant. When a plan participant retires, then the life-long pay-out will be determined based on a projected return and the “individual” reserved capital. In case the projected return is higher, then the pay-out will also be higher, but there is also a greater chance of setbacks. Each period a retired plan participant extracts part of their own reserved capital. Furthermore, plan participants share the longevity risk and therefore it is possible to safeguard a life-long pay-out. Hence, plan participants don’t buy their pension pay-outs for the whole period at retirement. Moreover, in the distribution phase, financial windfalls and setbacks are allowed to be spread over a maximum period of 10 years.

The projected return serves four different functions

The projected return will be used for multiple calculations. Firstly, the projected return will be used to determine the contribution which is needed to reach a pension with a maximum of 75% of the average salary in 40 years. Secondly, to determine the fiscal contribution limit. Thirdly, to determine the pay-out in the distribution phase and lastly, to communicate about the expected pension in good, median and bad circumstances. The idea is that, for all these different functions, an uniform calculation method, also known as “URM”, will be used. This will prevent arbitrariness. In addition, the projected return will be capped through legislation in accordance with the method that is used for the “Verbeterde Premieregeling” (Dutch pension law: article 63a, 3rd paragraph). Finally, given the fact that the contribution will be fixed between 30% and 33% for the coming 15 years, we can indirectly derive the projected return which is used for that. Therefore, the projected return is expected to be between 1.5% and 1.8% as shown in the graph below.

Obligated solidarity fund to cushion shocks

The collective solidarity fund would be able to cushion shocks. The purpose of the fund is that risks will be shared between generations as well as future generations based on clear and balanced rules over different ages. Moreover, the solidarity fund is non-distributed capital, whereas the remainder is distributed among plan participants. On average, the solidarity would result in better pension results for all participants in a collective and therefore it is also obligated under the new contract. Clear and balanced rules should be made on: how the solidarity fund will be funded, the distribution of capital from the solidarity fund to plan participants, the desired size of the fund and how the fund significantly contributes to risk sharing and stability between generations. The solidarity fund cannot be negative and will be capped at 15% of the total pension fund’s assets. The fund is expected to be funded using a maximum of 10% of the contributions. On top of this, a maximum of 10% of the positive excess returns can be used to fund the solidarity fund. In case excess returns will be used to fund the solidarity fund, then also retired participants and inactive plan participants contribute to the funding. Another option is to use existing capital to fund the solidarity fund on the transition date.

Distribution of returns based on a two-step approach: protection return & excess return

The collective investment policy and the distribution rules should be in accordance with the risk appetite of the different age cohorts and the “prudent person” rule. The latter, implies an implicit  “life-cycle” investing approach. This is the key change in the reform with a potentially large impact on both the asset allocation and hedging strategies. Investment risks and interest rate risks will be distributed among different age cohorts and might be used to fund the solidarity fund (capped at 10% of excess returns). There will be a two-step approach for the distribution of returns. First, the protection return, which is the protection related to changes in interest rate risk, will be distributed among plan participants and second the excess return, also known as the investment risk, will be distributed. The protection return will be based on the interest rate structure, also known as the ‘rentetermijnstructuur’ (RTM), which is published by the Dutch central bank. The excess return is basically the pension funds’ return minus the protection return.

Young participants have a higher share than older participants with regard to investment returns and the opposite holds for protection return. As a result, the expected pension value is more volatile for young plan participants and more stable for older plan participants compared to the current system.

Furthermore, the new system will allow for more flexibility to change pension pay-outs. We judge that a pension fund is able to index pay-outs if the excess return is positive, which is the total return of a pension fund minus the protection return or if the total return on a pension funds’ assets is higher than the projection return. This all means that pensions will move more in line with financial markets and less with interest rates as is the case under the current system. Furthermore, windfalls and setbacks are spread over several years. This would imply that bad years would be compensated by good years.

One of the possible sticking points is the abolition of the average system

The Dutch government wants to abolish the average system, “Doorsneesystematiek”, also known as the “double” transition, as it leads to redistribution between generations. Under the new contract, the contribution is flat and age independent and therefore plan participants build-up more pension accrual when they are younger, as the contribution can be invested for a longer horizon. This means a degressive pension build-up as pension accruals decline as members age. This is not the case in the current system, where young plan participants subsidize older plan participants. This means that employees who are 40 years and older lose around a cumulative EUR 60bn based on calculation by the CPB, as they have subsidized older participants when they were younger. In most cases, the new contract gives enough benefits to compensate the group of plan participants who have subsidized other plan participants, despite the fact that it is a zero-sum game. Below we will show the impact on the expected pension under the new contract including the abolition of the average system, but first we will explain the legal risks related to the transition.

Another possible sticking point is the legal risks related to the transition

Just like the current pension system, the new system should be legally tenable, which is the case according to the steering committee. They investigated whether it is legally tenable to keep the pension build-up under the second pillar mandatory, as this is very important for the solidarity between generations. According to an analysis done by Prof. dr. Erik Lutjens, who is a pension law attorney, this is the case (see here). Furthermore, under the Dutch pension law (Article 83) individuals can use the objection right. We judge that the Dutch government will change this law in such a way so that pension funds are able to object instead of individuals in case they can justify why their fund should be exempt. Moreover, under EU law (Article 157) equal treatment should take place, which is guaranteed by age independent contributions and a gender neutral rate. Equal treatment is also applicable for the transition to the new system, which means that possible lower expected pensions for specific age cohorts should be objectively justified. In this matter, a targeted compensation for plan participants who loose part of their pension will be helpful to justify the transition. Finally, the transition should of course be in accordance with property rights under EU law. This implies that the infringement must be proportional, justified, based on a legal basis and should pursue a legitimate goal in the public interest. Case law concerning property rights in pensions show that Member States have a wide discretion in the design of national pension systems and also related to whether a measure is in the public interest. As a result, potential risks related to property rights under EU law are expected to be limited.

The expected pension will only be lower for young plan participants in the bad scenario

The expected pensions for different salaries at 13 pension funds are calculated under the current and the new contract (see here). We will first set out the assumptions that are used to calculate the expected pension for three different scenarios, which are bad (5th percentile), median (50th percentile) and good (95th percentile).

Assumptions made to calculate the expected pensions under the current and the new contract

The current contract follows the rules of the FTK. The expected pension includes the gross AOW (first pillar) of a married person, which is EUR 10625. For the second pillar, a salary of EUR 25000, EUR 35000 and EUR 55000 is used to calculate the expected pension. The coverage ratio is set at 100% and the risk free rate is used to transfer current pension entitlements into reserved capital under the new contract. Furthermore, young plan participants have a higher share in the excess return (investment risk) and are less protected against interest rate risk (protection return), the opposite holds for older plan participants. With regard to the solidarity fund it is assumed that it will be funded with 10% of the contribution and that 1/15 of the fund will be contributed in proportion to the reserved capital. Below we show the expected pension for different age cohorts under the current and the new system.

The average expected pension pay-out is higher under the new contract than under the FTK

Overall, the average pension pay-out is higher under the new contract in the median and good scenario, except for the age cohort 0 as shown in the graph below. The former is mainly driven by the fact that under the FTK more will be distributed to the buffer. This is about 30% under the FTK and only 3% under the new contract in the median scenario and the difference becomes even large under the good scenario (5% versus 100%). Moreover, young and future plan participants are especially better off in the bad scenario under the FTK instead of the new contract, due to the fact that the accumulated reserve generates returns. For simplicity, we only showed the results of one pension fund instead of for the whole sample, which is pension fund B. There are small differences between funds. Having said that, the pension fund which we used is in line with the general picture.

Indeed, pensions are easier to index and cut under the new contract compared to the FTK

The table below shows that expected pensions will be more often indexed under the new contract and the chance of a decrease of the expected pension is also lower. However, the average cut and indexation are also slightly lower based on calculations by the CPB.

The above results all depend on the final details of the new contract and the scenarios that are used. However, they give an indication of the difference between the current system and the new contract. Below we will discuss how current pension entitlements can be transferred into the reserved capital pot for each plan participant.

Two valuation methods to calculate reserved capital under the new contract

The transition consists of three steps. First of all, the contribution should cover all costs, second the abolition of the average system and third the transition to the new contract. This is also known as the ‘Double Transition’. For the transition to the new contract, the current value of the pension entitlement should be calculated for each plan participant. The transition should be balanced including an adequate compensation for the abolition of the average system. For the transition of the pension entitlements under the FTK to the new contract, a pension fund has two options, namely the value based-ALM method and the standard method.

Under the standard method, a pension fund’s capital will be divided by calculating the current value of the defined benefits via the risk free rate. If there is a shortage or a surplus, then this will be divided over 10 years. After 10 years, the coverage ratio would be 100%. This is also the case for the value based-ALM method, however the transition will be designed in such a way that the effects of the transition remain as limited as possible. In addition, market rates on a specific date will be used for the valuation. As a result, under the value based-ALM method it is about the difference of the net benefit between the old and the new contract, which implies differences in the average present values of the pay-outs minus the contributions for both contracts. To keep the difference as small as possible, a pension fund has more degrees of freedom under the value based-ALM method compared to the standard method. Below we show the impact on the expected pension for both methods.

Average replacement ratios are higher despite the fact that it is a zero-sum game

The net benefit for each age cohort is shown in the left graph below for the new contract based on the standard method (VKV a standaardmethode) as well as under the value-based ALM method (VKV a VB ALM). In addition, the results for the current system the FTK (Huidig FTK) and the FTK without the average system (FTK zonder DSS) are shown. These calculations are done by the Dutch CPB and show that the double transition leads to a negative impact of roughly -3% for the affected cohorts in case the standard method is applied. Under the new contract, pensions will be more often indexed and this is especially beneficial for retirees under the new contract, as they have a positive net benefit under the standard method. In case the value-based ALM method will be used, then the net benefit of retirees will disappear as current plan participants will be compensated.  Moreover, the median replacement ratio shows that active plan participants get a higher expected pension as % of their pensionable wage compared to older plan participants and also future plan participants as is indicated by the light pink line (VKV a VB-ALM).

The difference between the net benefit and the replacement ratios are driven by the fact that the additional return on the investments is taken into account under the average replacement ratio, but not under the net benefit measure, as the net benefit method corrects the returns for the associated risks. Therefore, the additional return on equities disappears under the net benefit, for example. To conclude, under the value based-ALM method plan participants will directly be compensated for the double transition, which is not the case under the standard method. Furthermore, the average replacement ratios are expected to meet the target, which is  a maximum of 75% of the average salary in 40 years.

The effects of the transition are dependent on the economic situation at the time of the transition. For example, if the coverage ratio at the time of the transition is roughly 90%, then the most effective cohort which is 1980, will have a lower negative net benefit of about -1% instead of -3%. Due to a lower coverage ratio, it is expected that the build-up of the buffer will be relatively lower and this benefits the 1980 cohort under the FTK.

Most affected cohorts under DC lose about 10% due to the abolition of the average system

We will illustrate the effects of the transition from a progressive contribution percentage (an age -increasing contribution percentage) to a flat contribution percentage for defined contribution schemes. The choice for the contribution percentage determines the magnitude of the transition effects. In addition, higher and lower percentages are expected to impact the salary of a plan participant. The impacts of the transition from a progressive to a flat contribution are significant. We show the results from a 3% progressive contribution to a flat contribution of 16.7% based on a report published by the CPB (see here).

The graph on the left above shows that most affected cohorts will lose about 10% based on the gross benefit as well as based on the average replacement ratios. If the transition immediately takes place, then the age cohorts around 1970 suffer the most. If the transition will take place in 12 years (kanteling in 12jr), then the impact will be shifted by about 10years to the age cohort 1980. Due to the slower transition, younger plan participants will subsidize older plan participants. In case, the transition takes place in 12 years and a compensation is added, then the loss is reduced by about 3%-points under both the gross benefit and the average replacement ratios.

Default route would be to transfer pension entitlements to the new system

On Monday 22nd of June, the government minister overseeing the reforms, Koolmees, sent more information about the new contract to the Dutch parliament on the back of a request by the unions. However, it is rather unlikely that the House of Representatives will discuss the deal until the union FNV has approved it, which is scheduled for next week. By that time the cabinet will already be enjoying their summer recess. However, the Dutch government is willing to come back from their holidays and this clearly shows that the government feels the urgency to approve the deal by the end of this year for three reasons. Firstly, the cabinet will also decide about changes in the first pillar related to increasing the retirement age at a slower pace. Secondly, if the deal is approved, then most pension funds don’t have to cut pay-outs next year as the exemption rule will be extended for one year. Last but not least, General Elections will take place in March 2021. Consequently, we expect that the Dutch government will approve the deal before the end of this year and will subsequently start the process of changing the law to facilitate the transition as shown in the figure below.

We expect that the new law will be in place on the 1th of January 2022. Then the default route would be to transfer pensions to the new system in principle. Unless pension funds can justify why their fund should be exempt from this. For example, there are some pension funds where the employer is still liable to provide a top up if coverage ratios fall below certain levels (rather than decreasing the pension entitlement). In addition, for schemes that have already moved to defined-contribution system, there will be no benefit to move to the new system and therefore there is no money available to compensate employees which are 40 years and older as explained under the abolition of the ‘Doorsneesystemtatiek’. Between the second half of 2023 and 2025, we expect pension funds will move towards the new contract, with an exception for insurers and defined contribution schemes