Skip to main content

Your browser is out of date, and unable to use many of the features of this website

Please upgrade your browser.


This website requires cookies. Your browser currently has cookies disabled.

DC trust: presentation of scheme return data 2016 - 2017


This is the eighth edition of our annual statistics on occupational defined contribution (DC) trust-based pension schemes and memberships, including DC memberships of hybrid dual-section schemes.

This year's edition provides a high-level snapshot of the current landscape of occupational DC trust-based pension provision in the UK. It's based on a wide-ranging data set covering around 34,500 current schemes from the pension schemes register, with an effective date of 31 December 2016.

As with last year's release, we've included similar snapshot data from earlier versions of our database. These form a consistent time series, each with an effective date of 31 December for 2009 to 2016.

The information is based on data provided by schemes on returns we issued from July to December 2016 and relates to the levy year 2015 to 2016 or earlier. Therefore, the figures may not yet reflect all changes in DC memberships, including from automatic enrolment (AE) through to December 2016.

We provide an AE monthly report, based on information provided by employers when they declare their compliance, a quarterly compliance and enforcement bulletin, and we also publish an commentary and analysis on AE.

Executive summary

In this year’s statistics release, we've provided a textual description of the collection methods for the data contained in the Microsoft Excel (XLSX) files at the end of the page, together with information on what is covered by each of the different files.

  • Files 1, 2, and 3 are based on schemes with 12 or more members (with the exception of table 1 in file 1, which is based on all schemes with two or more members).
  • File 4 is based on all schemes with two or more members.
  • File 5 is based on schemes with 2 to 11 members.
  • File 6 shows relevant effective date information.

You can view and download the files in data tables.

Key findings

  • There are around 34,500 schemes with DC trust-based members, of which 32,000 are micro schemes.
  • DC schemes volumes are stabilising, with a reduction of just 3% in the past year, though the number of schemes with 12-plus members has declined by 10%.
  • In 2009, only 66% of schemes had members aged over 50, this is now 98%, and this represents around 20% of members, consistent since 2009.
  • There are 87 master trusts with DC members registered with us. Files 1 and 4 show the number of master trusts to be 70. This is because they cover a sub-population and exclude hybrid schemes with DC members and micro DC schemes. There is also an effect from rounding.
  • Less than half of schemes offer a default investment strategy, though more than two thirds of open schemes do this.
  • Membership has increased by 42% since last year, and by over 300% since 2009.
  • 55% of all private sector workplace pension members are in DC schemes, and 86% of all those currently saving are investing into a DC scheme.
  • 92% of members are invested in the scheme’s default strategy.
  • Assets increased by 17% this year, compared to 24% last year. The asset increase is made up of contributions (including transfers in) and investment returns, and is reduced by retirements and transfers out.
  • Asset values have almost doubled since 2010 (93% increase).
  • Contributions increased by 23% last year, compared to 28% the year before.
  • Average assets per membership are £4,700, a reduction from last year. This is driven by the huge influx of new memberships who started with no assets.
  • 750 schemes are being used for AE, up from 490 last year. Almost half of these schemes have fewer than 12 members.
  • 95% of members are in schemes being used for AE.
  • 72% of micro schemes are schemes formerly known as small self-administered schemes (SSASs). Relevant small schemes include schemes formerly known as SSAS and also executive pension plans (EPPs). We don’t currently have a full cycle of data for whether micro schemes are EPPs.

The landscape

The table below represents the private pension landscape in the UK, showing at a high level the different forms of employer-sponsored provision available within the private sector, and giving an overview of the size of each type of provision[1].

  DB  Hybrid: mixed benefit[2]
Hybrid: dual-section  DC (trust)   DC (workplace contract)[3] 
Schemes  5,170
Open schemes
26,240  2,400
Total memberships
Total active members

(Sources: The Pensions Regulator's data based on scheme returns 31 December 2016, Annual survey of hours and earnings (ASHE)[3] 2015 (published March 2016). Please note: the ASHE 2015 reports 4.4 million active members of DC trust-based schemes and 244,000 active members in schemes where the type was unknown.)

This statistics release focuses on the areas marked as DC (trust) and DC memberships of hybrid dual-section schemes.

We do not include data on assets held in DC sections of hybrid schemes, as we do not collect this information.

There are many varieties of hybrid schemes, but they can broadly be classed as either mixed benefit or dual-section:

  • a mixed benefit scheme offers one set of benefits which has elements of both defined benefit (DB) and DC schemes, such as a DC scheme with an underpin on a DB basis[2]
  • a dual-section scheme has two sections, one offering DC benefits and the other offering DB benefits

DC work-based personal pension schemes, also known as contract-based schemes, are excluded from this release, other than in the above table. The term ‘scheme’ here refers to the number of products offered by providers, and not to the number of employers who have bought into those products. These types of scheme consist of:

  • group personal pensions (GPPs)
  • group self-invested personal pensions (GSIPPs)
  • group stakeholder pensions (GSHPs)

DC contract-based schemes are subject to a different regulatory framework from DC trust-based occupational schemes. We work jointly with the Financial Conduct Authority (FCA) in the regulation of contract-based schemes. While contract-based schemes are obliged to register with us, they provide significantly less information than trust-based schemes. As a result, the above table is able to provide a figure for the total number of contract-based schemes. Furthermore, in order to provide landscape context, a figure is sourced from the Office for National Statistics (ONS) for the total number of active contract-based members. Beyond that, all types of DC contract-based provision, both group and individual, are outside the scope of this statistics release.

Relevant small schemes are subject to fewer regulations than other occupational DC schemes, and they include schemes formerly known as SSASs, and EPPs. We’ve recently added a question to the scheme return to identify EPPs, but this question has not yet been through the full cycle as schemes with fewer than 12 members only complete a return once every three years. As a result, this data is not included here and we are not currently able to provide statistics on the population of relevant small schemes. We’re continuing to provide figures in respect of schemes formerly known as SSASs.

We request information from schemes which we use to assess whether these schemes are (or were) small self administered schemes. We ask schemes to confirm the following:

  • This scheme has no more than 11 members (YES / NO).
  • All members are appointed as trustees to the scheme or all members are directors of the company which is the sole trustee of the scheme (YES / NO).
  • The scheme's provisions provide that any decisions to be made by the trustee(s) are made by unanimous agreement of the trustees who are members of the scheme / directors of the corporate trustee who are members of the scheme (as applicable) or the scheme has a trustee / corporate trustee has a director (as applicable) who is independent in relation to the scheme for the purposes of section 23 of the Pensions Act 1995 and is registered on our trustee register (YES / NO).

Pensioner members are not included in this data release. This is because our figures for pensioner members do not necessarily represent the number of members who have retired but remained in the scheme. Therefore, we cannot be confident that the number of pensioner members represents the total number of retirees from that scheme, including those who transferred out. We do however report the number of members for whom the scheme is providing (self-annuitisation) or facilitating (lifetime annuities) the annuity payments.

The data tables include figures for the total amount of master trust schemes (and members within those schemes), excluding micro schemes and hybrid schemes. The below table provides totals including micro schemes and hybrid schemes. It also shows these figures split according to whether the scheme has reported to us that it has achieved master trust assurance and would like to be added to the published list. It should be noted that there may be some schemes that achieved assurance but did not want to be added to our published list.

  Master trust (not on published assurance list)
Master trust (on published assurance list)  All master trusts with DC members 
Total schemes
Open schemes
Active DC memberships
Reported assets (excludes hybrids)
DC members in schemes which reported assets
Total DC members (includes hybrids)

There will be a new legal definition of master trust, which will be introduced by the Pensions Schemes Bill 2016-17, which is still going through the parliamentary process and has not yet commenced. The intent is that this definition will be further developed by regulations under the Bill. For the purposes of this year’s statistics release (including the table above), we have retained our previous working definition of master trusts which can be found in governance structures. As the legislation develops we will adjust the reporting of the data to reflect the final position.

The first three files of this statistics release file 1: schemes, file 2: memberships and file 3: financials refer to occupational DC trust-based schemes with 12 or more memberships (with the exception of table 1 in file 1). These schemes account for 5% of the total universe of DC trust schemes[4] but more than 99% of the memberships.

File 4: AE schemes refers to all schemes with two or more DC members.

File 5: micro schemes is concerned with schemes of 2 to 11 memberships. File 6: effective date information includes information on the effective dates[5] for data submitted on memberships, contributions and assets.

As at 1 January 2017, 66% of schemes have an effective date for their membership data after 1 January 2014 and this represents 92% of members.

For contribution data, 77% of 12-plus member DC schemes (those with 12 or more members and are not hybrid) have provided information whose effective date is 1 January 2014 or later, and this covers 99% of reported contributions.

For asset data, 77% of schemes that have 12 or more members and are not hybrid have provided information whose effective date is 1 January 2014 or later, and this covers 99% of reported assets.

For micro schemes:

  • 90% have an effective date for their membership information of 31 December 2012 or later, and this covers 92% of memberships
  • 79% of schemes have provided contribution information, and 99% of these provided information with an effective date of 31 December 2012 or later

Data collection

Pension schemes in the UK with two or more memberships are required to register with us. At the point of registration, an occupational scheme provides us with registrable information, which includes the following information:

  • scheme name and contact details
  • scheme status (eg open, closed)
  • membership information
  • details about the relevant employer(s), trustees and service providers

In addition to registration, we also operate a scheme return process.

The scheme return is how we collect information about occupational pension schemes. We are required to collect scheme return information from each scheme no more than once a year and no less than once every five years. The current practice is for DC schemes with 12 or more members to complete a scheme return annually and schemes with 2 to 11 members to complete once every three years.

Scheme returns are submitted using our online reporting tool, Exchange. This offers several advantages, one of which is that schemes can log in to alter their details at any time using the scheme maintenance process.

We ask for more information for occupational DC trust-based schemes with 12 or more memberships in the scheme return than for smaller schemes, such as scheme leaver financial amounts, default strategy and decumulation options.


In most cases, the data we show here is a direct summary of the data we hold. However, in some cases we've performed some simple calculations to provide more useful information. In such cases, we've also included the direct data from which we made the calculations, in a separate tab in the file labelled 'calculations'.

Total schemes year on year

In order to create the time series of total schemes year on year, we start with the total number of DC schemes with two or more memberships on the register as at 31 December 2016. We then use a table showing the total number of registrations each year, and the total number of schemes winding up each year. By removing the schemes established during 2016, and adding the schemes that wound up that year, we create an estimate for the number of schemes as at 31 December 2015.

We continue this process back to 1997, the year when the Occupational Pensions Regulatory Authority (OPRA), our predecessor, took responsibility for maintaining the pension schemes register. In the calculations tab the number of wind ups and registrations per year is shown dating back to 1988, as there is a significant peak in registrations that year.

Market value of scheme assets

While the significant majority of relevant schemes has provided us with their market value of scheme assets, it's not completed by every scheme. Within the data files we've shown data for the amount of assets that have been reported to us, as well as the memberships in schemes that have reported an asset figure. We limit this to reporting those asset values whose effective date is on or after 1 January 2014.

In order to allow users to estimate the total amount of assets within these schemes we've provided information to enable average assets to be calculated per member for each scheme size and status banding (where size is based on the number of members) for those schemes who have supplied the information.

Annual contributions (by employer and employee)

The issue for market value of scheme assets is also present for annual contributions. Within the files we've shown data for reported contributions and active members in schemes which have reported contributions. We limit this to reporting those contribution values whose effective date is on or after 1 January 2014.

We've used the same method to enable an extrapolated estimate for the total amount of contributions to trust-based DC schemes as for our estimate of assets in such schemes. The difference with respect to contributions is that we provide the active members, rather than all members, as only active members are contributing.

The active members figure could include members who were only active for part of the year.

Governance structure

The governance structure theme is designed to show the number of master trusts, schemes with non-associated employers, a growth area following automatic enrolment, and also to separate out those schemes which have been created by employers (unbundled) from those which have been purchased from a pension provider (bundled), and have all services provided by related corporate entities.

The scheme return doesn't have a question which provides the bundled / unbundled split so we made an estimate by analysing data provided on the insurer or administrator. Where the same company is listed as both administrator and insurer, we classed this as a bundled scheme. Where the companies are different we classed them as unbundled, and if the administrator is not present we labelled them 'unknown'.

Last year we started incorporating email addresses into this data cleaning process. So if a scheme provides an individual’s name as their administrator, we then check the email address for that individual and, if relevant, we assign that scheme to the appropriate company. This reduces the number of unknown schemes.

In addition, there has also been a change to the assumption made. Previously, where a scheme did not provide an insurer name they were classed as ‘unknown’. For the purpose of these statistics it is assumed that all bundled schemes will have an insurer, so where a scheme does not provide an insurer name it is classed as ‘unbundled’.

Amounts leaving schemes

As with assets and contributions, the amounts leaving schemes has not been supplied by all schemes. In the data files we show information for the reported amounts leaving schemes, and the number of members in schemes which have reported their leaver amounts.

Transfers into schemes

As with assets, contributions and leaver amounts, not all schemes have supplied relevant transfers in data. In the data files we show information for the amounts reported to be transferred into schemes, and the number of active members in schemes that have reported transfers in.

It should be noted that where members or amounts are flagged as being transferred this means they have been moved from another workplace pension scheme.


Governance structures

Master trust

For the purposes of this publication, a master trust is an occupational trust-based pension scheme established by declaration of trust, which is or has been promoted to provide benefits to employers which are not connected, and where each employer group is not included in a separate section with its own trustees. For this purpose, employers are connected if they are part of the same group of companies (including partially-owned subsidiaries and joint ventures).

Please note that as part of the current Pension Schemes Bill it is expected that a legal definition for master trusts will be introduced, but that definition has not currently completed the parliamentary process and received royal assent and so is not used here.

Industry-wide master trust

An industry-wide master trust is a master trust scheme that is only offered to employers within a specific industry.

Bundled / unbundled scheme

A bundled scheme is one where the majority of services are provided by related corporate entities, and an unbundled scheme is one where services are provided by different companies.

Relevant small schemes (RSS)

A scheme with fewer than 12 members where all the members are trustees of the scheme and either decisions have to be made unanimously or there is an independent trustee appointed to the scheme who appears on our trustee register. Where there is a sole corporate trustee, the above conditions apply in relation to the directors of the corporate trustee.

Executive pension plan (EPP)

An EPP means a scheme:

  • in relation to which a company is the only employer and the sole trustee
  • that the members of which are either current or former directors of the company and include at least one third of the current directors

Decumulation options

A scheme offering a lifetime annuity is one where at the point of retirement the scheme purchases an annuity in the name of the trustees, and then facilitates payment of that annuity to the member.

A scheme offering self-annuitisation will calculate an annuity payment for the member and make the payment to the member directly from scheme funds.

Should the member use the fund they have built up to purchase an annuity in their own name then they cease to be a member of the scheme.

You can find the results of these two questions in file 1: schemes and file 2: memberships.

Scheme status 

Open to new joiners and future service
Closed to new joiners, open to future service  Closed to new joiners and future service  Winding up 
New memberships are allowed to join the  New memberships are not allowed to join the scheme.
New memberships are not allowed to join the scheme.
Existing memberships (or employers) are allowed to continue contributing.
Existing memberships (or employers) are allowed to continue contributing.
Existing memberships (or employers) are not allowed to continue contributing.
Existing memberships (or employers) are not allowed to continue contributing.
Funds continue to be invested.
Funds continue to be invested.
Funds continue to be invested.
Funds continue to be invested.
Funds continue to attract charges.
Funds continue to attract charges.
Funds continue to attract charges.
Funds continue to attract charges.

Member status

Active  Deferred  Pensioner 
A scheme member who is continuing with future service.
A scheme member who has ceased future service, but has built up benefits which remain in the scheme.
A scheme member who is being paid a pension. An annuity has been purchased on their behalf in the name of the trustees of the scheme, or the scheme is paying a retirement income directly from scheme funds.
Funds continue to be invested.
Funds continue to be invested.
Funds can remain invested if the scheme operates a decumulation option.
Funds continue to attract charges.
Funds continue to attract charges.
Funds can continue to attract charges if the scheme operates a decumulation option.

The number of memberships does not equate to the number of individuals, as one individual can have multiple memberships.

In DC schemes which operate self-annuitisation or lifetime annuities, a member may have the option to purchase a joint life annuity. In this case, a nominated dependent will receive a proportion of the member’s retirement income for the rest of their life, should the member die before the dependent.

Membership figures collected reflect those members who are relevant for calculation of the pension levy. As dependents are not included in this they are not reported to us and so dependents are not included in the statistics in this release.

Data tables