Evolution of occupational defined benefit schemes 2025

A 10-year forward look using the new TPR DB universe projection model.

---

On this page

Key Highlights

The defined benefit (DB) landscape has changed significantly over the past few years with significant improvements observed in scheme funding. The introduction of the Pensions Schemes Bill will provide more options and opportunities in the future. This means that trustees have a range of options available. They now need to make decisions about the future of their scheme, including:

• whether to consolidate, either through the traditional buyout (insurance market) or the superfund regime
• how to utilise any surplus, either when running-on or at the point of buy-out

Our analysis shows the following:

• the insurance market is able to absorb all schemes that may wish to buy out over the next decade, although there may be some short-term pressures. Over half of all schemes are expected to buy out. However, actual volumes in monetary amounts are uncertain, particularly with regards to the latter half of the decade.
• there is space in the market for both insurers and superfunds to operate.
• £150 billion of surplus could be shared between saver s and employers over the following decade, based on a combination of sharing surplus on an ongoing basis (whilst being able to maintain scheme security) and at the point of buy-out.

Now is the time for trustees to understand all the options and plan to determine the future of their schemes over the following decade. Doing so will ensure that pension schemes are well run, provide savers with the benefits they are expecting, and provide value for money for savers.

TPR puts evidence at the heart of its decision-making as it seeks to protect, enhance and innovate in savers' interests. We will increasingly publish our insights to help inform better practices in the market and enable further debate.

This report sets out the conclusions from the detailed modelling we have undertaken to anticipate how the occupational DB pension scheme universe may evolve over the next ten years. We will revisit these scenarios annually to identify emerging trends and changing views. It is timely to do so when matters such as surplus release terms and future consolidation vehicles remain under discussion as legislation continues to be developed.

While this publication focuses on the evolution of the occupational DB pension scheme landscape, we are also undertaking modelling of the occupational defined contribution (DC) pension scheme landscape. The introduction in 2012 of automatic enrolment, requiring employers and employees to save into a pension scheme, has significantly increased pension participation, with 11 million more people saving into occupational DC pension schemes, primarily master trusts, over the last decade.

We intend to publish this DC analysis alongside our DB modelling in the future. Including DC will provide a broader and more complete picture, helping to further contextualise the overall shift in the pensions landscape.

Background

Note referring to ‘schemes’: where a scheme is sectionalised or segregated, we have treated each section as if it were a separate scheme for the purposes of this report. We have not aggregated the segregated sections to the parent level.

The funding positions for DB schemes have seen material improvements over the last few years. The trend in the closure of DB schemes to new entrants and future accrual has continued, which means that the majority of DB schemes are gradually maturing towards the point where all savers are either close to retirement or have retired and are drawing their pension.

At the same time, we have seen a growing and burgeoning bulk purchase annuity (BPA) market. There were around 300 transactions last year resulting in the transfer of around £50 billion of assets from occupational DB pension schemes to insurers. In parallel, we have also observed the establishment and operation of the first superfund as an alternative vehicle for DB consolidation and a greater number of, often larger DB schemes, considering ‘running on’ in the immediate future.

The starting point of our projections is 31 March 2025. At this time there were around 4,700 private sector occupational DB schemes holding £1.1 trillion of assets and promising pensions to around 9m savers.

DB schemes are better funded than at any point in recent history, with material improvements in scheme funding positions over the last decade. On a buy-out basis, the number of schemes in surplus has increased from 2% to 52% of the DB scheme population over the period 31 March 2016 to 31 March 2025 , with the surplus for these schemes increasing from around £2 bn to £92bn.

[PLACEHOLDER FOR IMAGE Figure 1: historical funding position on a buyout basis as at 31 March – 2016 to 2025]

Read the text alternative of Figure 1: historical funding position on a buyout basis as at 31 March – 2016 to 2025

Our approach to determining plausible scenarios to model

Our work looks at a range of scenarios to understand how the future evolution of funding positions, and behaviour of trustees and employers in managing long-term strategies, affects the DB sector.

We consider the scenarios modelled to be plausible based on discussions held with industry stakeholders. They are intended to prompt further debate and imply nothing about how TPR considers future government policy should develop . We stand ready to support the successful implementation of any future government policy.

We engaged with actuarial consultancies and other important stakeholders when determining a range of plausible scenarios. [INSERT LINK] Read the Industry views section of our DB universal projections report for more information about the discussions.

The accompanying report provides greater detail regarding the underlying assumptions and approach that we have adopted in this regard. Note that long term strategies are likely to be correlated with investment returns and that our model currently has one deterministic financial scenario and cannot draw out the implications of that.

Below, we highlight the main observations from the potential progression of the DB universe over the 10 years to 31 March 2035:

1. Over 75% of schemes can potentially buy out over the next decade

Overall, absent any scheme exits from the current DB universe, we expect the size of assets under management (AUM) in the occupational DB pension scheme universe to increase over the 10-year period in nominal terms. However, after adjusting for expected inflation over the period, this leads to a reduction in real terms of around 10% from £1.1 trillion to around £1.0 trillion. The reduction in size of through the ongoing payment of pensions and lump sum cash payments to savers.

Simultaneously, we expect that the number of schemes being able to buy out without recourse to requiring additional employer contributions to increase from 52% of DB schemes at 31 March 2025 to. The breadth of schemes in a strong position is therefore expected to increase accordingly.

[PLACEHOLDER FOR IMAGE Figure 2: projected percentage of schemes in surplus and deficit on buy-out basis assuming no exits]

Read the text alternative of Figure 2: projected percentage of schemes in surplus and deficit on buy-out basis assuming no exits

2. DB schemes are likely to remain a material segment of the occupational pension scheme market

Our projections illustrate that there will likely be a significant reduction in the number of occupational DB pension schemes over the next 10-year period, with around half of schemes transferring to the insurance sector and the other half remaining in the occupational pension scheme sector.

The vast majority of the schemes expected to transfer to the insurance sector are ‘small’ schemes, which we have defined as those with AUM under £100m. There is likely to continue to be a financially significant DB sector, with AUM in 2035 estimated to be around £0.6 trillion to £0.7 trillion (in real terms).

For context, the AUM of all occupational trust-based DC schemes, including those of master-trusts, has increased from £71 billion to £205 billion over the past five years (according to TPRs report Occupational defined contribution landscape in the UK 2024) and is expected to continue to grow over the next 10 years.

3. Over 2,000 schemes and more than £200 billion of assets could transfer to the insurance sector

There are now more options for DB schemes when considering how to secure their future. For some, buy-out remains the preference. The insurance market is able to meet current demand, and new competitors in the market mean that capacity is growing. We expect schemes that wish to buy out, and can afford to, will be able to do so over the projection period.

Based on our assumptions about the number and size of schemes who want to and are sufficiently funded to undertake a BPA transaction, combined with annual insurer capacity limits of around 350 schemes and around £50 billion (real) of liabilities, our projections show around 2,400 to 2,600 schemes with around £200 billion to £400 billion of assets are expected to participate in insurance transactions over the 10-year period.

[PLACEHOLDER FOR IMAGE Figure 3: cumulative volume of insurance transactions]

Read the text alternative of figure 3: cumulative volume of insurance transactions

As at 31 March 2025, there were around 3,650 small schemes (equivalent to 77% of schemes by number). The vast majority of the BPA transactions under all our scenarios are undertaken by these small schemes .

The majority of transactions are undertaken over the first half of the decade, where we have modelled that the insurance sector would absorb around 300 small schemes every year, equivalent to around 1,500 small scheme BPA transactions over the next five years and around 2,100 over the 10-year period.

The reduction in the number of transactions over the latter half of the decade is due to there not being enough schemes who wish to buy out having sufficient assets to be able to proceed with a transaction based on insurer pricing as at 31 March 2025. For many of the remaining schemes, it is possible that additional funding from the employer would enable the transfer to an insurer, although this is not something that we have incorporated into our modelling.

While the majority of schemes are ‘small’, the combined AUM for all small schemes accounted for only around 7% of the whole universe as at 31 March 2025. The overall level of transactions in pound amounts, particularly over the latter part of the decade, is highly dependent on the behaviour of ‘large’ schemes with AUM in excess of £250 million and, in particular, those schemes with AUM in excess of £1 billion.

We can see this in the large variation of projected transactions in the billions of pounds (real). Our estimates range from around £200 billion to around £400 billion in real terms of BPA transactions in respect of our lower and higher scenarios modelled. Whether large schemes decide to run-on or transfer to an insurer will determine how the market evolves over the latter part of the decade.

Based on industry feedback, our modelling has incorporated an increase in the time that it takes the initial BPA transaction to convert to a full buy-out. We have used a time period of three years in our modelling for all scenarios and have assumed that this remains constant over the 10-year period.

There was much debate over whether the recent increase in time required for insurers to fully onboard the scheme was just a temporary phenomenon or whether the continued demand for future transactions might actually increase this further. At the same time, we are aware of innovation in this area which may help to close this gap. We will keep this issue under observation.

4. Surplus of £150 billion could enhance saver benefits or be returned to employers

The proposed return-of-surplus legislation will enable regular distributions of assets to both savers and employers. However, the proportion of schemes that will choose to run on and extract surplus is uncertain. Our discussions with industry suggest that around half of large schemes are considering running on in the short to medium term.

Based on these opinions, we have adopted an initial approach which assumes half of schemes with assets over £ 250 million decide to run-on and release surplus. Under this scenario, we make the following observations:

• it could potentially provide savers and sponsoring employers of schemes that are closed and wish to run on with around £65 billion (real) surplus refunds.
• for ‘open’ schemes – assuming that instead of returning surplus they utilize their surplus in order to cover the cost of future benefit accrual – it enables a saving to savers and employers of around £30 billion (real).
• for schemes that decide to buy out, around a further £50 billion surplus is available for savers and sponsors from schemes that undertake a BPA transaction from the surplus that exists in excess of the cost of buyout.

Together, these represent a total of around £150 billion (real) assets available for savers and employers of defined benefit schemes over the next 10-year period.

[PLACEHOLDER FOR IMAGE Figure 4: annual surplus distribution combined £ billions (real)]

Read the text alternative of figure 4: annual surplus distribution combined £ billions (real)

Industry feedback indicates a range of views, but trustees are generally expected to take a cautious approach to surplus extraction under a run-on model. Many are likely to prefer maintaining a buffer above low dependency, with some schemes likely to target solvency plus an additional margin.

There is also uncertainty about how surplus assets may be shared. Our approach to the modelling assumes that, for schemes running on, the trustees retain a buffer before releasing any surplus at a level of 5% above the cost of buyout. This could be considered at the more conservative end of the range. We also assume any surplus distribution is shared 50% between savers and employers.

For open schemes, we have assumed that they use all surplus in excess of full funding on their technical provisions basis without any buffer. In reality, there will be many factors that trustees and employers will need to be considered when determining the funding basis, the level of any buffer to retain and how any surplus is shared and distributed, which we have not considered further in our report.

To understand the sensitivity of the level of potential surplus available to savers and employers, we have looked at schemes having a higher and lower appetite for running on rather than buying out. We have also considered the impact of higher and lower buffers at the point that ongoing surplus is released.

5. Superfund market exists alongside BPA market

The Pension Schemes Bill has introduced a timescale for the introduction of legislation for superfunds. One commercial superfund is already operating, and we are aware of interest from other potential interested parties to enter the market.

At the moment, it is still too early to make any detailed predictions for how the superfund market may develop.

We have considered the impact of a superfund market being an attractive alternative to the BPA market as a long-term solution for some schemes. For the purposes of our modelling, we assume that 33% of schemes that are under £1 billion in size and may be looking to buy out at some time in the future, instead transfer to a superfund. We combine this with schemes over £1 billion in size with weak covenants for whom a superfund transaction could improve saver security.

This scenario shows transactions of around 350 schemes, with AUM transferring into the superfund market of around £35 billion (real).

Our analysis also indicates that the development of such a superfund market does not affect the BPA market over the short term, given that there are excess numbers of schemes who are sufficiently funded and want to insure all benefits. Over the 10-year period, the competition for consolidation in the DB market is expected to reduce the level of BPA transactions by around £10 billion (real) . It should be noted that we have not included any allowance for a superfund ‘bridge to buy-out’ model within our analysis, which would increase the number and size of the BPA market in those scenarios.

6. Gilt holdings reduce across the sector

Our analysis suggests the level of gilt holdings that underpin the DB liabilities of the 4,700 initial schemes reduces by around 35% to 45% from around £570 billion to between £320 billion (real) and £380 billion (real), dependent on our higher and lower run-on scenarios.

[PLACEHOLDER FOR IMAGE Figure 5: volume of gilt holdings £ billions (real) - DB schemes and insurers combined]

Read the text alternative of figure 5: volume of gilt holdings £ billions (real) - DB schemes and insurers combined

The reduction in gilt holdings is a consequence of schemes selling assets to pay savers’ pensions and provide surplus distribution, and for those schemes who buy-out, our assumption regarding how insurers invest the assets that are transferred. In particular, schemes currently have around 50% of their asset holdings in gilts, while we assume that insurance companies only hold 20% of their assets in gilts in respect of our proposed transactions.

Our expectation is that the overall gilt allocation for the occupational DB sector is unlikely to change materially. Therefore, the extent to which the above reduction applies will depend primarily on the investment appetite of the insurance sector for those schemes that transfer.

Our analysis also indicates that the average ‘duration’ of gilts held by DB pension schemes is expected to reduce over time, caused primarily due to schemes maturing and the average duration for pension scheme liabilities reducing.

Summary of results

If the projections below occur in practice, we can expect substantial activity across the DB and insurance markets within a relatively short time. We will continue to monitor how the market evolves to ensure it continues to meet savers’ needs.

For the purpose of this report, ‘small schemes’ means schemes with assets less than £100 million as at 31 March 2025 and ‘large schemes’ means schemes with assets more than £250 million as at 31 March 2025.

Summary of scenarios modelled

Run 1: base projection

The base projection assuming all schemes run-on with no scheme exits and no surplus extraction.

Run 2: technical provisions surplus used to fund accruals

Technical provision surplus is used to fund accruals for open and open to accrual schemes, and no further schemes exit or permit surplus extraction.

Run 3: run-on scenario 1 (middle estimate)

50% of large schemes (greater than £250 million) run on for 10 years, surplus extraction is 50% benefit uplift and 50% paid out immediately as lump sum whilst none of the small schemes (less than £100 million) run on except those that are open or immature and open to accrual.

Run 4: run on scenario 2 (lower run-on estimate)

25% of large schemes (greater than £250 million) run on for 10 years, surplus extraction is 50% benefit uplift and 50% paid out immediately as lump sum while none of the small schemes (less than £100 million) run on except those that are open or immature and open to accrual.

Run 5: run on scenario 3 (higher run-on estimate)

75% of large schemes (greater than £250 million) run on for 10 years, surplus extraction is 50% benefit uplift and 50% paid out immediately as lump sum whilst none of the small schemes (less than £100m) run on expect those that are open or immature and open to accrual

Run 6: superfund scenario

From 2026, around 350 schemes conduct a superfund transaction. These are broadly 33% of schemes that we assume are not running on and are less than £1 billion in size with buyout funding level between 80% to100% plus a small number of £1 billion plus schemes.

[PLACEHOLDER FOR IMAGE Summary of results]

Read the text alternative of summary of results

[Insert link] Read the full report for further details about the modelling

Appendix 

Appendix 1

[PLACEHOLDER FOR TEXT]

Appendix 2

[PLACEHOLDER FOR TEXT]

Appendix 3

[PLACEHOLDER FOR TEXT]

Appendix 4

[PLACEHOLDER FOR TEXT]

Appendix 5

[PLACEHOLDER FOR TEXT]

Appendix 6

[PLACEHOLDER FOR TEXT]