Research and analysis

Pension charges survey 2020 – summary

Published 13 January 2021

Overview

This report provides the findings in the third in a series of studies commissioned by the Department for Work and Pensions (DWP), designed to assess the types and levels of charges across defined contribution (DC) trust-based and contract-based workplace pensions. In June 2020, the government committed to undertake a review of the default fund charge cap and standardised cost disclosure by the end of the year following the last review of the charge cap in 2017. Findings from this research report have been used to inform recommendation in the review.

To protect employees, workplace pensions are subject to a variety of rules, notably the government’s charges measures introduced in 2015 and 2016. One of these measures caps ongoing charges for pension schemes used for automatic enrolment (known as qualifying schemes) at 0.75%. Other rules prevent providers from levying charges that could be particularly inappropriate for people automatically enrolled into their employer’s scheme.

The research has 3 main objectives:

  • to measure average overall charge levels and the prevalence and level of different charge components
  • to measure the proportion of charges (if any) over the cap
  • to explore provider and trustee views of the current charge cap and whether the level of, or costs included within, the charge cap should change

The results are segmented by different scheme types and characteristics, such as the number of members of the scheme, and whether the scheme is a master trust, a trust-based or contract-based scheme.

Other workplace pensions (known as non-qualifying schemes) were also covered by the research to establish whether they indirectly benefited from the cap.

The study also collected information about a range of charges that, although outside the cap, are considered relevant, such as the prevalence of set-up fees being paid by employers.

Research context

In March 2014, DWP published a Command Paper, that announced a comprehensive range of charges measures designed to improve the value for money of DC workplace schemes.

In conjunction with these new charges measures, DWP commissioned the first wave of the Pension Charges Survey in 2015 to benchmark and monitor the impact of the charges legislation prior to the charge cap being implemented. This research was designed to capture the full range of charges that were applied to qualifying DC workplace pension schemes.

A second wave of research using the same methodology was undertaken in 2016 which reported that:

  • the charge cap had the intended impact on qualifying schemes. Qualifying contract-based and trust-based scheme charges had fallen across the board in almost all size categories, with average charges well below the cap
  • non-qualifying schemes had not benefitted to the same extent, with few changes since 2015. In particular, smaller non-qualifying schemes were not benefitting from scale, still paying annual fees as high as 0.92% on average

The 2016 research also covered unbundled pension schemes. These are schemes where trustees, often based at a single employer, work directly with separate administrators and investment managers to administer the scheme, as well as commissioning other specialists such as auditors.

Unbundled schemes’ charges varied depending on whether the scheme was open, closed or frozen. However, on average, charges in unbundled schemes were typically comparable to their equivalent bundled trust-based schemes, although a relatively small number of closed, non-qualifying schemes charged markedly higher than the average.

The Parliamentary Work and Pensions Committee has subsequently recommended that DWP review the level and scope of the charge cap, as well as permitted charging structures. The previous government agreed to this review, which is taking place in late 2020. To support the review, work on this third wave of research began in May 2020.

Key findings

  • all members in the qualifying schemes covered by this research are now below the cap, and the average charge of 0.48% across all members is significantly below the cap
  • comparing the pension providers who took part in both the 2020 and 2016 studies, and new market entrants, the overall average charge for qualifying schemes has barely changed, up 0.01 percentage points
  • the average charge for other workplace pensions not covered by the cap (non-qualifying schemes) is now 0.53%. Again comparing the providers participating in the 2 research waves, this represents a reduction of 0.20 percentage points. As a result, 88% of members of non-qualifying schemes are now below the level of the cap
  • charges for unbundled trust-based schemes remain at 0.49% on average, with all members of qualifying schemes within the cap
  • 72% of funds under management faced no additional fund manager expense charges (FMECs), a slight improvement from 70% in 2016
  • the average FMEC (including funds with no FMEC) is slightly down from 0.025% to 0.022%. Excluding the funds with no FMEC does not alter the trend noticeably (down from 0.084% to 0.080%)
  • the average transaction costs for remaining in the fund (also known as portfolio transaction costs or PTCs) is 0.069%. Excluding the funds with zero or negative charges increases the average to 0.083%. In 2016, insufficient providers had been able to give an estimate to calculate an average PTC so it is not possible to report a trend
  • approximately two-thirds of providers reported that they had zero direct investments in illiquids in their default fund(s). About a third had a small proportion, typically between 1.5% to 7.0%. All providers investing in illiquids mentioned property as their main class of investment. A small number also mentioned infrastructure, private equity and debt
  • the main barriers to investing in illiquids related to the high costs associated with these investments, with the unpredictability of charges also a concern

Methodology

The survey incorporates the full range of DC schemes, whether held directly with providers (bundled) or operated on an unbundled basis.

Reflecting the 2 types of workplace pension, the research used 2 different approaches:

  • we asked pension providers to report charges data using an Excel template designed by our research team with assistance from DWP and the Financial Conduct Authority (FCA)
  • telephone interviews with unbundled trust-based schemes, in which we collected information about each individual scheme’s charges

Since 2016, a significant number of new providers have entered the market. As a result, the number of pension providers returning completed templates increased from 14 in 2016 to 20 in 2020.

The 14 providers included in the 2016 research included 8 of the top 10 providers and 14.4 million pension pots. The providers returning templates this wave accounted for 29.3 million pension pots and included all of the 10 largest providers.

Each provider’s template was checked for consistency (both internally, with 2016 results and with public information). Where needed, we followed-up any issues concerning the template with the provider via email or telephone. An Excel pivot table model integrates all the data and allows tracking against previous waves.

A total of 35 qualifying unbundled schemes were also interviewed as part of the research, of whom 32 provided sufficiently complete data to include in the analysis. These unbundled schemes accounted for 432,000 members. Schemes accounting for 72% of these members were able to report upon the ongoing charge paid by the members and schemes accounting for 49% of members were able to report transaction charges for funds invested.

In addition, follow-up telephone interviews taking approximately one hour were undertaken with all providers and a representative sample of 15 unbundled schemes. These interviews clarified our understanding on charges and any challenges in their implementation and explored recent and future developments and the burden of study participation.

Findings Explained

1. Ongoing charges paid by members

The ongoing charge is levied by the provider in relation to administering the scheme, and is expressed as a percentage of funds under management per year. It includes any contribution charges or flat fees levied on the members.

For members in qualifying bundled schemes the average charge of 0.48% is significantly below the cap. Comparing just the pension providers in both the 2016 and 2020 research and new market entrants, the overall average charge for qualifying schemes has barely changed, having risen by 0.01 percentage points.

All members of qualifying schemes now have ongoing charges that fall within the charge cap, an increase of one percentage point since 2016. This reflects the phasing out of the handful of legacy qualifying contract-based and trust-based schemes that became inactive before 5 April 2015 and were, therefore, not subject to the cap.

In previous years, a higher level of ongoing charge was found among non-qualifying bundled schemes, which was attributed to them typically being older and sold in a less regulated and less competitive environment than qualifying schemes. In 2016, the average charge amongst members of non-qualifying schemes was still as high as 0.84%.

However, in 2020, charges for non-qualifying schemes have fallen sharply and the average charge is now 0.53%. Again comparing just the providers in the 2 research waves, this is a reduction of 0.20 percentage points. As a result, 88% of members of non-qualifying schemes are now below the cap level.

Charges for unbundled schemes remain consistent at 0.49% on average and all members of qualifying schemes are within the cap.

2. Active Member Discounts (AMDs), consultancy charges, initial commission and trail commission

‘Legacy’ charges that were banned under the charges measures include AMDs, consultancy charges and member-borne commission.

These had been eliminated from qualifying schemes by April 2016, and remained extremely rare even among non-qualifying schemes (where the charges measures did not apply).

Their usage has further declined and is now restricted to a small number of providers paying trail commission for contractual reasons, which is not passed onto scheme members.

3. Fund Management and transaction costs

Workplace pensions are also subject to a variety of charges relating to the funds they invest in, which can be divided into FMECs and transaction costs.

FMECs are charges paid by a member who invests in a particular fund, over and above the ongoing charge, to cover additional expenses incurred by the fund manager. Not all funds have additional FMECs associated with them: they normally apply to funds that require more active management.

72% of funds under management faced no additional FMECs, a slight increase from the 70% in 2016. This compares to the major change between 2015 and 2016 when FMECs became much rarer as a proportion of all members (in 2015, only 56% of all funds attracted no FMEC).

The average FMEC (including funds with no FMEC) is slightly down from 0.025% to 0.022%. Excluding the funds with no FMEC does not alter the trend noticeably (down from 0.084% to 0.080%). Only 2% of all funds faced charges of greater than 0.20% (down from 4% in 2016).

Many types of fund incur costs while assets remain invested, because underlying assets may be purchased or sold on an ongoing basis by the fund manager. They are usually deducted from members’ pension funds directly and are commonly known as portfolio transaction costs (PTCs).

The average PTC (including funds with zero or negative charges) was 0.069%. Excluding the funds with zero or negative charges would increase the average to 0.083%. In common with FMECs, funds with charges greater than 0.20% were rare, affecting only 3% of funds by fund value.

Transaction costs for unbundled schemes were estimated at 0.26% on average, slightly lower than the 2016 estimate (0.35%).

4. Illiquids

Illiquid investments are assets that are traded off-exchange or are otherwise less readily tradeable than cash, shares or money market funds. Examples of such investments include direct property investment, investment in infrastructure projects, private equity, equity or debt issued by very small listed firms, and venture capital.

Approximately two-thirds of providers reported that they had zero direct investments in illiquids in their default fund. About a third had a small proportion, typically between 1.5% to 7.0%. All providers directly investing in illiquids mentioned property as their main class of investment. A small number also mentioned infrastructure, private equity and debt.

Some barriers to illiquid investment for DC pensions included the following:

  • the nature of illiquids makes them more difficult to access than liquid funds. This lack of flexibility can be an issue for DC schemes, for whom daily liquidity and switching is required in order to meet the needs of individual members
  • illiquid investments attract higher fund management charges and, while they can offer better potential performance in the long run, this can make them appear uncompetitive in a market focussed on charges
  • coupled with higher costs for illiquid investments, the unpredictability of fund management charges and performance fees in regard to illiquids made providers nervous about the potential to exceed the charge cap
  • smaller providers typically were reluctant to accept the concentration risk associated with illiquid investments