The new pension contract: still work in progress
The government is currently working on a new regulatory framework for pension funds. The present rules, set out in the Financial Assessment Framework, will be revised due to the introduction of a new pension contract in 2015. This summer, State Secretary Klijnsma sent the preliminary draft of the revised Financial Assessment Framework to the Lower House of the Dutch Parliament for consultation. Crucial elements of the new assessment framework have been worked out in that draft. The consultation version of the Framework accommodates both the existing pension contract and the new contract. Following completion of the consultation, however, the State Secretary said to prefer one type of contract only in the final version of the Framework. This contract should be a mix between the existing and new contract. But the exact shape of such a combination is not yet clear. A draft bill will be submitted to the Lower House at the beginning of 2014. Looking at the consultation document, we believe that the new contract as is set out in this draft is better aligned with the interests of pension scheme participants. We focus here on explaining the differences between the existing and the proposed new contract.
The existing pension contract
The pension contract used by most Dutch pension funds is a so-called ‘benefit agreement’. Every month, employees are obliged to contribute a specific percentage of their pensionable salary in exchange for which they accrue a nominal pension entitlement (i.e. an entitlement in euros). The pension level is related to the average wage earned during their career. In principle, this is a ‘fixed’ nominal commitment; but many funds can decide to nominally curtail pensions in exceptional circumstances. In addition, funds seek to increase pensions in line with the price or wage level (indexation). Pension funds are subject to capital requirements which are designed so that pension funds can finance their nominal liabilities with a 97.5% degree of certainty over a one-year horizon. To comply with those requirements, pension funds need a (nominal) funding ratio  of around 130%.
The certainty with which pensions can be paid out has been significantly eroded. Owing to the decline in funding ratios in the wake of the dotcom crisis of 2001 and the financial crisis of 2008, many funds were no longer able to meet the required degree of certainty. Indexation of pensions has often not been applied in the past few years and therefore the purchasing power of future pension pay-outs is no longer as robust as many participants originally thought (Goudswaard et al, 2010). In addition, the way in which information is provided on expected pensions is confusing. The Uniform Pension Overview (UPO) states entitlements in euros, but this says virtually nothing about the expected purchasing power at retirement. That is because, without indexation, the accrued nominal pension will be eroded by inflation over time.
Pension funds wishing to achieve the degree of certainty referred to above will have to reduce risks in their investment portfolios and hedge a greater portion of the interest rate risk when funding ratios are low. It became clear during and after the financial crisis of 2008 that many pension funds did not follow this policy in practice. Quite the opposite, many funds have displayed an increased risk appetite to be able to recover from their low funding ratio (Boeijen et al., 2010). Moreover, incurring risk is necessary to generate sufficient returns to enable indexation of pension entitlements.
 The funding ratio is the ratio between the assets and the discounted value of the liabilities of a pension fund.
The new pension contract
In the pension agreement of 2010 the social partners expressed the wish to introduce a new pension contract alongside the present nominal contract (Piljic and Smid, 2012), allowing pension funds to choose which contract they wish to use. The new contract is also referred to as a ‘real ambition agreement’. The term ‘real’ as used here conveys the inclusion of indexation in the pension entitlement in order to explicitly target a pension linked to the price or wage level. The real contract differs in this respect from the present ‘nominal’ contract, in which indexation is only an ambition and pension rights are expressed in nominal terms. The differences between the nominal and real contract are also reflected in the calculation of funding ratios of pension funds. Since indexation is already incorporated in the pension entitlement in real contracts, the funding ratio is expressed in real terms as well. In nominal contracts, the nominal funding ratio is what matters, as only the nominal entitlements are guaranteed.
In the real contract, the pension entitlements are geared more directly to developments in financial markets and increases in life expectancy. Shocks can be spread out over a longer period (of at least three and at most ten years) than in the present contract (three years). Accordingly, while pension reductions occur more frequently in real contracts, their extent will be limited. The opposite is true for nominal contracts (Figure 1).
As windfalls and setbacks are accounted for immediately and in full in the real contract, the funding ratio will automatically return to 100% after a shock. For example, if a negative shock occurs, the curtailment in the pension entitlements in the current year as well as that for all other years within the recovery period is accounted for. The same mechanism is triggered in the case of a positive shock, except that the adjustment is upwards. In the existing pension contract, the nominal entitlements are guaranteed in principle and pension funds will seek to absorb any negative shocks as much as possible by discontinuing indexation. In the case of a maximum reduction of indexation, i.e. 0% indexation instead of 100%, the liabilities are only limited by the size of the price or wage increase (i.e. by no more than 3 to 4% per year) (Bovenberg and Van Ewijk, 2011). While funding ratios will in most cases be able to return to 100% within a few years, the difference compared to the real contract is that future cutbacks on pension entitlements are not communicated to participants in advance.
The discount rate is the interest rate that is applied to determine the present value of pension fund liabilities. Again, this calculation is different for the real contract (Box 1). Because indexation is part of the entitlement in the real contract and the entire entitlement is at risk (no guarantees are provided, contrary to nominal contracts), the discount rate accounts for both future inflation and financial market risks. Accordingly, the expected inflation and the risk premium are incorporated in the discount rate applied in real contracts. They are included in the calculation of the funding ratio and thereby affect the amount of pension entitlements. This differs from the nominal contract, in which the risk premium only plays a part in the calculations that pension funds make with regard to the expected return on their plan assets.
Box 1: Determining the discount rate
The discount rate that pension funds use to measure their liabilities is based on the risk-free discount rate. In practice, the interbank swap rate is used for this purpose. To counter disruptions in the interest rate market, the regulator decided to artificially increase long-term discount rates by an Ultimate Forward Rate (UFR) (Smid, 2012).
The discount rate for nominal and real contracts is determined as follows:
In a real contract, therefore, two components are added to the discount curve: an indexation abatement and a risk premium. The former is applied because indexation to the price or wage level is part of the pension entitlement, the latter because developments in the financial markets are more directly reflected in entitlements. A uniform indexation reduction has been adopted in the revised Financial Assessment Framework. The risk premium is the same for funds with similar participant bases and is independent of the risk profile applied in the pension fund’s asset mix.
For both the UFR and the risk premium, ‘green’ funds, i.e. funds with mostly long-term liabilities, benefit from a higher discount rate compared to the ‘grey’ funds. This reflects the idea that the degree of ‘certainty’ by which pensions can be paid out should be greater for pensions with a short duration. Figure 2 compares the existing nominal yield curve with the real risk-weighted discount curve applying to funds introducing the new (real) contract (under the parameters currently applied by the Dutch Central Bank). The real risk-weighted yield curve is significantly below the present nominal yield curve, due to the low short-term nominal interest rate and the limited short-term risk premium. This would mean that funding ratios for funds that introduce the real contract would fall, forcing many funds to apply pension cutbacks immediately.
Figure 2: Nominal and real risk-weighted yield curve
 Another reason why individual cutbacks are lower is that a new recovery period is opened for each shock. Again, this differs from the nominal contract, in which new shocks have to be absorbed during the ongoing recovery period.
 The parameters for wage and price changes are only relevant for a limited number of contracts with unconditional indexation.
Does the real contract offer effective solutions?
Pension funds that introduce the real contract will not have to scale back risks in response to falling funding ratios, as the nominal guarantee is abandoned. This is more closely in line with funds’ and participants’ own wishes, as funds are barely hedging more risks after the financial crisis compared to before. Abandoning the nominal guarantee boosts recovery capacity and making the pension entitlement explicitly risky is also a fairer way of communicating the level of pension rights to participants. For in the past few years, pensions have not exactly proven to be a gilt-edged guarantee. Communication about the expected pension level will also become more meaningful in the real contract. Pension funds no longer have to explain to participants that nominal certainty is not the same as real certainty, as the real contract already takes future inflation into account.
Yet the real contract does not eliminate all problems. A frequently-voiced criticism of the institutional design of the Dutch pension system is that participants’ ownership rights in pension funds are not clearly defined (Bovenberg and Mehlkopf, 2011; Teulings and De Vries, 2003). Pension funds absorb risks in an anonymous buffer instead of settling shocks directly with individual participants. The uncertainty about who owns a funding surplus or who should bear a deficit may give rise to distribution conflicts. This applies to both nominal and real contracts. One of the most significant risks is that the ageing participant base of pension funds may compel social partners to settle distribution conflicts in their favour or that the regulator may be excessively receptive to this group’s interests. The regulator is already applying various measures in an effort to mitigate the pension woes of existing retirees as much as possible. For instance, ahead of the introduction of the new pension contract in the ‘transitional year’ 2014, pensions funds no longer have to apply a curtailment if new shocks occur. An additional risk is that, in determining the discount rate applying to real contracts, overly optimistic assumptions regarding inflation and the risk premium are made.
The new pension contract eliminates several crucial shortcomings of the present contract, such as the limited recovery capacity of pension funds with low funding ratios and the way of communicating about pension entitlements (which currently happens in nominal terms, without accounting for indexation). Further-reaching reforms of the pension system are necessary however, such as greater transparency in terms of ownership rights. A more direct allocation of risks to individual participants, or clearly demarcated groups of participants, may offer a way forward. An important condition for achieving this is a transparent set of rules prescribing how pension funds should treat funding surpluses and deficits. Those rules need not be the same for all pension funds, as long as they serve to ensure that participants know in advance where they stand and support for these rules is sufficiently broad. The role of politics would be to commit itself as firmly as possible and to ensure that future governments are not tempted to adapt the rules in favour of particular generations.
Boeijen, D., N. Kortleve and J. Tamerus (2010), Van toezegging naar ambitie, Netspar Nea Paper 35, June 2010
Bovenberg, A.L. and C. van Ewijk (2011), Naar een nieuw pensioencontract, Netspar Design Paper 01, July 2011
Bovenberg, A.L. and R. Mehlkopf (2011), Hoe snel moeten pensioenfondsen herstellen?
Een aantal overwegingen, TPEdigitaal 2011, volume 5(2), pp. 69-82
Bovenberg, A.L., T. Nijman and B. Werker (2012), Voorwaardelijke Pensioenaanspraken, Netspar Occasional Papers, 2 April 2012
Goudswaard et al (2010), Een sterke tweede pijler: naar een toekomstbestendig stelsel van aanvullende pensioenen, Commissie Toekomstbestendigheid Aanvullende Pensioenregelingen
Piljic, D. and T.H. Smid (2011), Pension agreement for the future?, Rabobank Special Report 2011/09, July 2011
Smid, T.H. (2012), A realistic discount rate, Rabobank Special Report 2012/16, September 2012
SZW (2012), Hoofdlijnennota herziening financieel toetsingskader pensioenen, Ministry of Social Affairs and Employment
Teulings, C.N. and C.G. de Vries (2003), Generational Accounting, Solidarity and Pension Losses, IZA Discussion Paper No. 961, December 2003
The closing date of this study was 28th November 2013