Introduction
The Dutch pension fund industry is undertaking a once-in-a-generation shift from defined benefit to defined contribution pensions, that is reshaping one of the world’s largest retirement systems, under a recent reform. At the core of the reform is a policy judgement that the old system made unsustainable promises that had also become hard to explain. Low and volatile interest rates increased the cost and volatility of defined benefit promises, while rising life expectancy strained intergenerational risk sharing. Policymakers wanted a system that is fairer between generations, better aligned with a mobile labour market, more transparent for participants, and financially resilient across interest rate cycles. The Future Pensions Act of 2023 thus mandated a shift to a defined contribution system, under which investment risk will largely be shifted onto members, though with appropriate buffers to protect the value of their pension pots. This shift radically changes pension funds’ hedging needs for what concerns interest rate risk and is becoming a key theme for market participants following the long-end part of euro area yield curves. In this article, we reflect on the implications of the act, provide a timeline of events to follow, and elaborate on its potential effect on markets.
Defined Benefit versus Defined Contribution
To fully explain the implications of a shift from defined benefit (DB) pension schemes to defined contribution (DC) pension schemes, we must first explain the differences in the liability structure of the two scheme types.
Defined benefit schemes promise fixed periodic payments to members that will be paid after they retire for the rest of their life; in other words, they provide lifetime annuities. The magnitude of the periodic payments that will be made to each member depends on the contributions they have paid in the past. Once an estimate for the life expectancy of members has been made, a DB scheme can quantify and time its liabilities to its members as the present value of future payments that it will have to make once they retire.
Defined contributions schemes, on the other hand, do not promise a fixed periodic payment after retirement age. In defined contribution schemes, payments after retirement age will depend on the performance of the assets to which each member’s contributions are allocated. Depending on the scheme, members can have more or less choice over the asset allocation mix; certain members might be interested in choosing riskier or more conservative allocation solutions. If the asset allocation mix performs well, payments to pensioners will be larger, and vice versa payments to pensioners will be lower if the asset allocation mix they chose performs poorly. Different schemes offer a different degree of freedom with which members can withdraw cash from their pension pots. In a very broad sense, DC schemes either agree with pensioners on future withdrawals or communicate to pensioners the payments they can expect to receive in the upcoming years. As a matter of fact, DC scheme managers only know the timing and amounts of their short-term portion of liabilities and try to predict future pension payments for a time horizon ranging 5 to 10 years from the present. Indeed, this is the key difference with DB schemes, and it commands a difference in how pension schemes of each type hedge their interest rate risk.
The value of the liabilities of a pension scheme changes with the magnitude of contributions, but also when interest rates move and when life expectancy estimates change. The risk associated with changes in the value of liabilities due to life expectancy is called longevity risk. While DB schemes bear longevity risk, in DC schemes it is usually borne by members because their pension pot is finite. Depending on the country and model, DB schemes can choose to bear longevity risk on their balance sheet and set up ad-hoc capital buffers, or otherwise to contract it out to an insurer.
Pension schemes are also exposed to interest rate risk, for the portion of fixed payments that they’re committed to making in the future. Based on the liability structure described above, it’s easy to understand that DB scheme liabilities typically have higher duration compared to DC scheme liabilities. This leads to a key difference in how interest rate risk is hedged between DB and DC schemes, since they target buckets of the curve. DB schemes hedge interest rate risk by receiving fixed in long-maturity swaps and by buying long-maturity bonds. These hedges provide long duration exposure to compensate for the fact that pension liabilities increase in value when interest rates drop. DC schemes, on the other hand, only hedge the interest rate risk for the portion of predefined payments they will have to make, and thus rely on short-maturity instruments.
The Future Pensions Act of 2023 and Transition Plans
The Future Pensions Act of 2023 was approved by the Dutch senate in May 2023 and came into force in July 2023. With the aim of reforming the Dutch pension system, the act mandates that all defined benefit pension funds in the Netherlands convert to a defined contribution model. The act initially imposed the 1st of January 2027 as a deadline, but the Dutch parliament then extended the transition period to the 1st of January 2028.
The act introduces two new reference types of DC scheme:
- Solidary Premium Scheme (SPR) pools contributions into a collective portfolio, with the provision of a solidarity reserve that can buffer shocks and stabilise benefits. The reserve can be filled from contributions and or outperformance subject to caps defined in the law.
- Flexible Premium Scheme (FPR) gives participants more choice over investment risk and withdrawal patterns, while still operating with personal pension pots.
Policymakers argue that these structures make pensions more future proof by improving the match between assets, contributions and participant risk capacity, and by making the consequences of economic shocks more transparent.
In practice, while all funds are required to transition, they were also given some flexibility in how to achieve the transition of past contributions. Funds could move past accruals into SPR or FPR accounts, but they could also choose to keep accrued contributions as DB liabilities on their balance sheet, as long as all risks were fully hedged: this means hedging 100% of interest rate risk, longevity risk, and should inflation indexation promises be present, inflation risk. In addition, funds could also choose to spin off all risks of a DB book to an insurer via a buy-in (pension fund still has the liability on its balance sheet but buys an insurance contract to cover some or all of the liabilities) or via a buy-out (transfer of the liabilities to the insurer, which becomes the legal payer to members).
Across the Dutch pension system, the decisive actors are ABP (Government and Education employees’ fund), PFZW (Care and Welfare workers’ fund), PMT (Metal and Technical workers’ fund), PME (Metal and Electrical engineering employees’ fund), bpfBOUW (Construction employees’ fund), BOV (Professional Road Transport employees’ fund), and PGB (Creative and Media employees’ fund). Each has now pinned down a switch window under the Future Pensions Act. We report assets under management as reported by De Nederlandsche Bank in their 2025 Q1 release.
- PFZW (AUM: €9bn) has confirmed a 1 January 2026 go-live in a solidary contract, and in which accrued rights will be converted, subject to supervisory approval. Along with PFZW, also BPF Bouw (AUM: €66.0bn) and PMT (AUM: €84.2bn) will switch on 1 January 2026. PMT has stated to the press it will reduce its interest-rate hedge over roughly six months after the switch, using the minister’s new flexibility to tidy hedges post go-live.
- ABP (AUM: €8bn) has published its transition plan and targets 1 January 2027, also under a solidary contract. PME (AUM: €56.6bn) has also targeted the same transition date to switch to a solidary scheme.
- PGB (AUM: €5bn) is yet to release a draft of the transition plan, but they have made agreements with social partners and employers which feature a transition to a solidary scheme for most sectors. BOV (AUM: €34.0bn) is yet to release information regarding its transition plan.
- As for all other funds, the transition dates for the bulk of AUM are almost evenly split between 1 January 2026 and 1 January 2027, with relatively more AUM transitioning on 1 January 2027, and a few outliers with intermediate transition dates.
The Future Pensions Act may have an impact on markets through two channels. The first channel is associated with how future contributions will be managed and paid out. The second channel is linked with accrued contributions and how the payouts associated with those contributions will change. One might argue that the first channel of transmission of this shock is likely to deliver its effects in the long run, while the second channel is likely to deliver more immediate, short-run effects.
Once Dutch pension funds begin collecting and managing contributions according to the rules of a defined contribution model, they might be less interested in seeking long-dated bonds and swaps than before, as a result of the change in their hedging needs. While it’s difficult to establish what asset classes they will re-allocate to, it’s reasonable to expect structurally less demand for instruments in the long end of the curve coming from these pension funds, which would then translate into higher long end rates both in swap and European government bond markets.
When DB schemes transition past contributions into FPR or SPR schemes, their hedging needs change radically. After the transition, only a portion of their liabilities is likely to be hedged against interest rate risk, and it’s primarily composed of short-term pensioner liabilities. PIMCO estimated that a large majority of total Dutch pension fund liabilities are non-pensioner liabilities and thus are likely to be hedged differently after the transition. Pension funds that transfer past contributions accrued in defined benefit schemes into defined contribution schemes thus face the problem of how and when to unwind their long-duration interest rate risk hedges. While certain funds may gradually unwind their hedges in the months leading up to their transition dates, other funds may prefer to hedge more interest rate risk leading up to the transition dates in order to protect their funding ratios.
This is because lower funding ratios make it difficult to convert DB promises into DC pension pots at their full value: should funding ratios drop abruptly, political pressure for a further delay of the transition may occur, in an effort to safeguard the integrity of pension pots, and to avoid more drastic measures such as freezing indexation or cutting pensions.
What could this shift mean for rates markets?
After having introduced the fundamentals of the event, we can now speculate on how they will impact markets. In order to unwind their hedges against interest rate risk, Dutch pension funds would have to sell long-dated bonds and enter fixed payer positions in long-dated swaps. Critics have argued that a large chunk of the unwinding of these positions is likely to take place in swap markets rather than bond markets due to liquidity concerns.
The fundamentals seem to support a case for curve steepener trades on the euro swap curve. Outright fixed payer positions may also offer exposure to the unwinding of pension funds’ fixed receiver positions but are exposed to shifts in the overall level of the curve. In order to bet on long-end swap rates rising relative to those of a shorter tenor, one would have to receive fixed in the shorter-tenor swap and pay fixed in the longer-tenor swap.
When selecting the buckets of the curve for a steepener trade, two challenges arise. The first challenge is understanding what points of the long end of the swap curve will be most affected by the unwinding of interest rate hedges by Dutch pension funds. Will Dutch pension funds unwind mostly via 20-, 30-, or 40-year swaps? The second challenge is represented by a trade-off between liquidity considerations and having targeted exposure to the flow dynamic presented. The 40-year or 50-year buckets may best isolate the effect of pension fund unwinds, but liquidity considerations will have to be made; by contrast, trading the 20-year and 30-year buckets would benefit from better liquidity at the expense of less targeted exposure. While these considerations apply to the choice of tenor for the fixed payer leg of the trade, similar ones apply for the choice of tenor of the fixed receiver leg of the trade. If we compare a 5s30s steepener and a 10s30s steepener the main difference does not lie in liquidity, but rather in exposure to front-end drivers like short-term macro.
It goes without saying that this trend has caught the attention of markets. As an example of why this is the case, we mention a key political event that played in favor of the steepener trade. On 20 May 2025, the Dutch lower house of parliament rejected an amendment proposal to the Future Pensions Act that would have allowed pension scheme members a choice to keep their accrued rights under a DB system. On the day of the announcement, the 10s30s slope increased by 4.11 bps, and continued in this direction for the rest of the week. Relative to a 10-year history of daily 10s30s movements, such a shift represents a 2.28-standard-deviation event.
Rather than attempting to quantify the impact in terms of volumes that will be unwound – which, based on the considerations presented thus far, can easily be understood to be large and likely to move markets – we will provide an overview of factors that have driven the thesis and how their evolution in the next few months affects it.
What’s next? Things to watch ahead of the deadlines
An interesting theme to watch ahead of the transitions is the evolution of hedge ratios relative to interest rate risk. De Nederlandsche Bank periodically releases useful data on individual pension funds, including a measure of how much interest rate risk each fund hedges, which is calculated as the ratio of the DV01 of assets relative to the DV01 of liabilities. While one might expect funds to gradually unwind their interest rate hedges in the lead-up to the transition so as to minimize market impact, many funds have increased their hedge ratios in order to minimize exposure to interest rate risk and preserve their funding ratios. As we get closer to the key dates of the transition, a continuation of this trend is likely to put flattening pressure on the long end of the swap curve, and later on, to lead to larger position unwinds later.
Indeed, De Nederlandsche Bank’s next data release is scheduled for 18 September 2025, and it should include sector assets, policy funding ratios and each fund’s supervision data. It matters because it will anchor where interest-coverage stands going into the 2026 and 2027 switch windows. ING estimates the current average funding ratio of about 68% could fall toward 40% after the reform; as stated earlier, if the sector keeps adding hedges beforehand, the eventual DV01 to unwind grows (ING’s base-case simulation points to an unwind of around €500m DV01, concentrated in the 30-year bucket of the curve). That makes 18 September the first hard checkpoint.
Operational updates can still move the long end on short notice. Two smaller funds (veterinarians’ pension fund and physiotherapists’ pension fund, managing a collective AUM of around €5bn) postponed their 1 July 2025 switches with only two weeks’ warning, underlining tight workflows at administrators. If a large plan such as PFZW were to delay from 1 January 2026, we could expect some degree of flattening pressure at the ultra-long end as markets push out the timing of receiver unwinds. In short, legal processes and go-live readiness have direct curve impact.
Flow risk clusters around two windows. The first is 1 January 2026, when PFZW, PMT and bpfBOUW plan to move. The second is 1 January 2027, when ABP and PME plan to switch. Sell side and industry trackers point to a very large share of AUM concentrated on those dates, although slippage is possible; regulators have also granted funds that transition to DC up to 12 months after their transition dates to temporarily deviate from transition-related investment policy and to finish hedge adjustments, precisely to avoid cliff-edge flows.
Autumn indexation decisions are another practical watch item. If policy funding ratios permit higher indexation in Q4, some funds may lift near-term pensioner cash-flows and keep a bit more short-dated receiving in place, even while preparing to unwind the far end after go live, temporarily lifting demand in less than 20Y tenors without changing the structural steepening bias out in 30-50Y.
Finally, expect the narrative to shift from hedging to allocations as plans are executed. Each quarterly factsheet and implementation update reveals how “return” and “matching” sleeves are being rebuilt under solidary contracts. According to ING, ABP has already signalled a reduction of €25bn in government bonds in its strategic allocation, which ING scales to €100bn for the sector on a rough, age-mix-adjusted basis. Together with a structurally lower need for long-dated receivers, that’s why most sell-side research on the topic warns of long end swap led steepening and possible volatility around the switch windows.
References
[1] De Nederlandsche Bank; Statistics Data Search
[2] IPE, Dutch Pension Funds to get 12 extra months to review interest rate hedges
[3] PIMCO, The End of the Dutch Defined Benefit Model: A Steeper Euro Swap Curve Ahead, 2023
[4] ING, Dutch pension reforms: 4 things to watch, July 2025
[5] Achmea Pensionservices, “Wtp-transitie Pensioenfondsen Dierenartsen en Fysiotherapeuten uitgesteld”, June 2025,
[6] ABP Transition Plan, 2024
[7] PME Transition Plan Summary for holders of accrued rights, November 2024
[8] PFZW Implementation Plan, May 2025
[9] BPF Bouw, Implementation Plan, June 2025
[10] PMT Transition Plan, July 2025
[11] Pensions Europe, “Road to DC: Understanding the shift”, April 2024
0 Comments