Response to our consultation on the code of practice for the authorisation and supervision of collective defined contribution pension schemes.
Published: 9 June 2022
You can save the contents of this page as a PDF using your web browser. Open the print options and make sure the destination/printer is set as 'Save as PDF':
Introduction
The workplace pensions market continues to evolve. Currently the UK’s workplace pensions framework enables employers to offer a defined benefit (DB) or a defined contribution (DC) scheme. The government believes that there is also a place for a third type of scheme, a collective money purchase scheme (commonly called a collective defined contribution (CDC) scheme), and we welcome this innovation.
Background
The Pension Schemes Act 2021 (PSA21) introduced an authorisation and supervisory regime for CDC schemes. In addition, the Occupational Pension Schemes (Collective Money Purchase Schemes) Regulations 2022 set out what we must take into account in authorising a scheme, and the information we must be given as part of an application and during supervision. The draft code of practice that came out of this legislation sets out the process for authorisation, including what is more likely to satisfy us that the legal requirements have been met. We consulted on the draft CDC code from 25 January to 22 March 2022.
The government’s intent is that only CDC schemes that are well run and built on sound foundations should be allowed to operate. The authorisation and supervision regime is designed to protect members and build confidence in CDC as a new form of occupational pension. The PSA21 and subsequent regulations will enable CDC schemes to become a reality, and require savers to be appropriately protected through an authorisation and supervision regime, overseen by The Pensions Regulator.
Consultation responses
We received a total of 24 responses from stakeholders, including advisers, industry associations and employers. We would like to thank all those who participated, as their feedback has helped us to clarify some aspects of the code.
Key themes
Many respondents commented on the level of detail in the code, and questioned if this was proportionate for a single employer scheme, or whether the detail was more suitable for multi-employer schemes.
We also had feedback that more guidance would be useful as the regime expands.
We’ve changed the final version of the code in response to some of the feedback we received. In the next section we explain how we’ve responded.
Comments on the ‘general’ and ‘introduction’ sections and applying for authorisation questions
We asked:
- Do you consider that any important areas of the authorisation criteria have been missed in the new code?
- Is the level of detail we have set out appropriate?
- Are there areas where supporting guidance would be useful?
- Is it clear what constitutes a section and when you must divide a scheme into multiple sections?
- Is it clear how the authorisation fee will be set for schemes with multiple sections?
You said:
Several respondents had concerns that the authorisation criteria were too onerous, and suggested they should be more proportionate to avoid creating barriers to entry for multi-employer and smaller schemes.
Some asked for further information on the application process, and others commented on the high cost of applying for authorisation, asking for clarification on how the fee structure would work for schemes with more than one section.
Most respondents felt the code was comprehensive, if onerous, and provided suggestions on how to improve it to make it even clearer, including requesting guidance on the application process and fees for multiple sections.
Our response:
While we understand that the code may appear to be onerous, we also believe that this level of detail will be necessary and useful to any employer considering establishing a CDC scheme, and to the trustees of a CDC scheme applying for authorisation.
We’ve amended the code to better explain its role in the context of the legislative framework it sits in. The PSA21 sets out that all CDC schemes must be authorised before they can operate, and mandates the authorisation criteria that form the basis of our decision. The regulations then set out, for each of the authorisation criteria, what we must take into account in deciding whether or not we are satisfied that a scheme should be authorised. The PSA21 makes it clear that we must authorise a scheme when we’re satisfied that it meets the authorisation criteria, and must not authorise if we’re not satisfied.
The function of the code is to set out the matters that are more likely to satisfy us that the scheme meets the authorisation criteria. It doesn’t, however, create additional criteria beyond those set out in legislation. We’ve added additional explanation into the introductory section of the code to better explain this.
The legislative requirements for sectionalisation appeared to be well understood, although a number of respondents felt that additional guidance on fees would be useful, and we will look to publish this as soon as possible.
We were pleased to see that a number of respondents had considered how the code might need to change as the legislation expands. While we anticipate we will need to amend the code to address any future legislative changes, we hope that much of the existing code will be applicable in the future.
We also want to thank respondents for their suggestions for future guidance which we will consider as we work with the Department for Work and Pensions (DWP) and industry to move forwards with the expansion of CDC. In the meantime we will look to publish guidance on the fitness and propriety criteria and on how fees are to be calculated for authorisation of additional sections of a CDC scheme.
Comments on the fitness and propriety questions
We asked:
- Is the level of detail we have set out appropriate?
- Are there areas where supporting guidance would be useful?
- Is it clear what the expected level of trustee competence is to successfully govern a CDC scheme?
You said:
Most respondents felt this section of the code was clear and the requirements were a sensible approach to assess persons as ‘fit and proper’. There were some concerns about individual competence assessments, as these were felt to be a barrier for new trustees and member-nominated trustees. Additionally, several respondents felt that the assessments could discourage inexperienced trustees and become a barrier to making trustee boards more diverse.
Our response:
We were pleased that many respondents felt that the code was broadly clear on the requirements for fitness and propriety. We’ve clarified our approach in respect of individual competence for trustees to set out our different expectations for a new or inexperienced trustee, and an experienced or professional one. Our intent was not to limit the pool of CDC scheme trustees to professionals and this should now be clear. A number of respondents also asked for clarification on who would need to be assessed, and we will set out guidance to address this.
A small number of respondents said that they didn’t believe we should be assessing individuals who appointed or removed trustees, however we are required to do this by the PSA21 and so have retained this in the code.
Comments on the systems and processes questions
We asked:
- Is the level of detail we have set out appropriate?
- Are there areas where supporting guidance would be useful – for example preparing a governance map?
- Is it clear what constitutes an effectively administered and governed CDC scheme?
You said:
Most respondents agreed with the requirements for the systems and processes needed to effectively run a CDC scheme. Some raised issues around whether it was necessary for the systems and processes to be in place and live at point of application, or whether it would be sufficient for them to be in place but still under development.
Our response:
We have sought to clarify in the code our expectations on the state of readiness for systems and processes at the point of application. We also clarified a number of minor points that were raised, in particular the role of ‘accountable people’, and have amended the wording to ‘responsible people’, adding further narrative to our expectations for clarity.
Comments on the member communications questions
We asked:
- Is the level of detail we have set out appropriate?
- Are there areas where supporting guidance would be useful?
- Do you see any barriers in meeting our expectations for effectively using members’ feedback and communicating how benefits may potentially vary from target?
You said:
Most respondents agreed that member communications were important in explaining how the scheme works and highlighting the risks involved. However, a significant number of respondents commented that the frequency of the reporting requirements to trustees and members was too onerous. Many also said that it would be virtually impossible to demonstrate that all their members understood the communications.
Some respondents were concerned about information overload and focus on the downside risks of CDC schemes, which could have the negative effect of disengaging members.
Several respondents also felt that the quality assurance process should be limited to ‘key communications’, such as those that explain the mechanics of the scheme and annual benefit statements.
Our response:
We consider communicating with scheme members to be an integral part of running a well-governed scheme. As CDC schemes are new and more complex than standard DC schemes, we feel there should be an even greater focus on ensuring that members understand how the scheme operates and this is required by the legislation.
Therefore, schemes should be regularly consulting with their members to ensure that their communications are clear, easy to understand and not misleading. We expect member communications to be a standing agenda item for regular trustee board meetings, so that their effectiveness can be properly monitored.
We agree that quality assurance doesn’t need to be carried out for all member communications (for example standard administrative letters such as change of address etc) every year. The focus of the quality assurance should be on communications that explain how the scheme operates, the benefits, and how the benefits can rise and fall.
We’re not expecting schemes to develop a significant number of member communications as this could lead to overload and, as a result, a negative effect on member engagement. However, we expect communications to provide enough information for members to understand how the scheme operates, and easily access more information if they choose (for example by signposting to a scheme microsite or online member portal).
Many respondents were concerned about the reporting requirements for member communications, especially the requirement to report on member feedback to trustees. We recognise that this has been viewed as having to actively seek feedback from members every quarter. While we haven’t changed our position on this, we’re happy to clarify that we don’t expect schemes to actively seek member feedback quarterly.
The feedback could include any complaints related to member communications, any employee forums/focus groups that have provided feedback on member communications, and any surveys the trustees may have carried out. These are just examples of the type of feedback we’re looking for, and some reports may simply confirm that there was no feedback that needed to be discussed that quarter.
Comments on the financial sustainability questions
We asked:
- Is the level of detail we have set out appropriate?
- Are there areas where supporting guidance would be useful?
- Is the structure of the Costs, Assets and Liquidity Plan reasonable for this type of model?
You said:
We received various responses on our proposals for our expectations of how schemes should demonstrate financial sustainability. The majority focused on the Costs Assets and Liquidity Plan (CALP), either in relation to the principles on which it is built, or the practical complications of producing it. Some respondents queried whether it was necessary to hold reserves. This was either in the short term, or where they expected that income from scheme assets (or the assets themselves) would be sufficient to cover the costs of a triggering event. Other respondents accepted the need for reserving but questioned whether the reserves could be built up over time. Some asked whether it would be acceptable to move from employer guarantees towards cash reserves as time progressed.
A few respondents looked forward to potential future developments. They were interested in the ways in which reserving and the CALP might work for multi-employer schemes.
One respondent asked us to clarify some of the language we use about the accounts we expect from the scheme’s employers.
Our response:
Our expectations of financial sustainability seek to create a framework that allows only well-supported new schemes to be authorised. This may mean that small employers are less able to set up their own CDC scheme. However, recent announcements from the Pensions Minister have made it clear that government intends to put legislation in place to allow smaller employers to participate in multi-employer CDC schemes.
Some respondents were interested as to how the framework established here might adapt to those multi-employer schemes in the future. Our expectations in this code are based on the legislation as it stands, which does not allow for schemes with multiple unconnected employers, and it is too early to speculate what might happen in future iterations of legislation.
Respondents often cited the information burden of our expectations alongside the financial ones as a concern. While we require the CALP to be submitted as part of the application, most of the contents are set out by legislation. This requires trustees to either disclose that information, or for us to take certain matters into account when considering whether we are satisfied with the scheme.
We can’t choose to waive these requirements or, as suggested, water them down in certain circumstances. Even if it were not a legal requirement, we can’t imagine a scenario where it would be acceptable for otherwise comparable schemes to face a different bar for authorisation. Any dilution of our expectations could expose savers’ benefits in those schemes to greater risk following a triggering event.
In a similar vein, one respondent suggested that we could receive information over time to build up a picture of a scheme. Their intention was to ease the burden on smaller potential entrants. The legislation sets out requirements for our assessment, and for a valid application. It’s simply not possible for us to approve a scheme on a provisional or contingent basis.
The question does, however, highlight an important point. Authorisation of a scheme is only the first step. Once authorised, a scheme then moves into supervision. Trustees will have to keep us satisfied that the authorisation criteria are met throughout the life of their scheme. One aspect of the scheme that is most likely to change over time and require updates is financial sustainability.
Over time, a scheme’s financial sustainability resources may change not just in scale, but also in composition. For example, a CDC scheme might start off being supported by guarantees from their employer(s). As that scheme grows and matures it could transition to holding asset-backed reserves held in dedicated accounts.
Some respondents claimed that the reserving requirements could be based on a rough guess of the likely cost. They took the view that a triggering event was likely to occur many years in the future and need not be an immediate concern. While it is likely that any triggering event for a newly authorised scheme would be some time away, it’s not guaranteed. We expect schemes to be able to demonstrate that they have appropriate financial reserves at all times. Any estimation of future costs is likely to be unreliable, and to expect trustees to forecast costs decades into the future is unrealistic. Our expectation of the period covered by a forecast is only three to five years, at the trustees’ discretion. This is a timeframe that allows forecasts to be based on actual experience and more reliable estimates.
While we expect such forecasts to be updated regularly, we don’t anticipate seeing significant swings in the amounts needed in reserves. Trustees must still expect to be challenged by us on their calculations, assumptions and estimates, both at the point of application and throughout their supervision.
The financial reserving requirements are put in place to ensure trustees have access to the necessary resources to deal with a triggering event, whenever that might occur. A group of respondents suggested that reserving would be unnecessary as costs of dealing with a triggering event would be small and could be taken from scheme assets within the charge cap. We don’t agree. We believe that the costs of dealing with a triggering event, especially one that leads to the scheme winding up, would be much greater than normal running costs. The law (section 45 of the PSA21) prohibits increasing or introducing administration charges after a triggering event, so we can’t consider this suggestion.
We’ve sought to clarify some elements of our expectations. For example, the legislation requires an application to contain details of set-up costs. For many schemes, much of the cost of setting up a new CDC scheme will have been incurred by the point that the trustees apply for authorisation. For this reason, we can confirm that any setting-up costs that have been incurred and settled by the employer need not be included in an application. However, applications must still contain details of any costs incurred but not yet settled, and any still expected to be incurred.
A further relaxation has been to remove requirement from the list of ‘calculation of financial reserves’. This called for trustees to maintain reserves greater than the already prudent reserves that they would have calculated. As this expectation created further additional prudence on already prudent calculations and asset haircuts, we agreed that this measure was excessive.
Comments on the continuity strategy questions
We asked:
- Is the level of detail we have set out appropriate?
- Are there areas where supporting guidance would be useful?
- Is it appropriate for a CDC scheme to not plan for continuity option three (closure) when it first comes for authorisation and are there any risks to this?
You said:
We received various responses on our expectations for preparing a continuity strategy following a triggering event. In the main, comments were supportive of our approach but emphasised that actions required by a triggering event are likely to be highly context-specific, and therefore not necessarily capable of detailed planning in advance.
Respondents largely appreciated that there was little point in preparing for continuing to run a scheme closed to new contributions or members until it was a financially viable option, but some were keen that the collective benefits of CDC were not lost by a lack of consideration that could lead to winding up becoming the default option. A few respondents emphasised that it should be possible to model in advance when running as a closed scheme would become a viable option and be aware of how the scheme design could, in such an event, have a significant effect on members’ expectations for future, and perhaps past, service.
Our response:
Although the draft code stated that we don’t expect the continuity strategy to be exhaustively detailed, we appreciate that some respondents thought that it might be too onerous if schemes had to plan for all possible outcomes. We’ve clarified our expectation further by distinguishing between the principles of strategically planning for what you would do when dealing with a triggering event, and an implementation plan to manage the consequences once one has occurred.
We confirm that we don’t expect new schemes to plan for continuing to run as a closed scheme at application, but we do expect trustees to assess when it could become a viable option unless the scheme’s rules do not permit it.
Comments on the scheme design questions
We asked:
- Is the level of detail we have set out appropriate and are there further matters we should consider?
- Are there areas where supporting guidance would be useful?
- Is it clear what we expect with regard to testing and modelling, and are there any additional issues or factors which could be relevant?
- In regard to testing and modelling, is it appropriate to expect schemes to conduct asset liability modelling (ALM)?
- Are there any other aspects of the trustees’ stewardship of the investment strategy that we should be assessing in support of a scheme’s design being sound?
- In respect to the gateway and live running tests, are there any further matters we should consider, and is it clear what we expect?
- Our plan is to provide a standard template for the viability certificate to ensure what is being certified is standardised across CDC schemes. Is this helpful?
You said:
We received various responses on our proposals for our expectations of how schemes should demonstrate sound scheme design.
Responses in the main focused on:
- further guidance on what constitutes soundness
- trustees' stewardship of investments
- restrictions on the investment strategy imposed by the scheme rules
- our expectations for stochastic ALM modelling
- the need for guidance on member options
- adjusting for post-valuation experience
- more clarifications in the Viability Certificate
Our response:
Soundness
The majority of respondents noted that we hadn’t provided a definition of a sound scheme design. While the regulations and the code of practice set out some tests that must be taken into account, they leave discretion for us to exercise our judgement. We believe this approach allows for the flexibility needed to encourage a market to develop. At this point in the development of CDC schemes, it would be inappropriate to define soundness tightly, and given the intent reflected in the regulations, it would be inappropriate for us to narrow the DWP’s policy intent. We have taken the approach that reasoned conclusions on the tests in regulations, supported by a sufficiently robust evidence base, should provide evidence of a sound scheme design, and have provided further clarity on this.
Investment
Almost all respondents agreed with the approach we’ve taken to trustees’ stewardship of investments. We’ve sought to take on board respondents’ comments and have made it a more explicit expectation for the actuary and investment advisers to engage with each other.
We’ve also provided more information on our approach to schemes where trustee investment decisions are governed by scheme rules. We expect to see justification for any prescriptions on investment policy that are set out in scheme rules. We also need to know how and when the prescriptions are capable of being amended to allow the trustees to be able to react appropriately where necessary, including during a triggering event period. We believe that, in some contexts, investment prescriptions that seek to support and protect the scheme’s design could provide additional comfort that the aspiration and nature of the scheme can be achieved.
We’ve removed the expectation on trustees to have defined contingency plans to respond to periods of poor investment performance. However, we’ve added a more explicit explanation of the need for robust monitoring and governance to support the ongoing justification of investment approach and ensure the scheme’s design remains sound.
Modelling
Most respondents supported some use of stochastic asset liability modelling at inception to support understanding of the scheme’s design. Respondents suggested that a timescale of three to five years for re-running stochastic ALM would be appropriate. While this period might be broadly appropriate, we’re conscious that we don’t want to drive unnecessary re-running of modelling when there is no need. Consequently, we’re not specifying a timescale for ALM to be updated, but we do expect trustees to use their judgement to understand when it would be appropriate in the context of their scheme and broader circumstances. To support this judgement, we’ve provided some examples of significant changes that could indicate a need to undertake new stochastic ALM.
Some respondents felt that the extent of testing and modelling expected could be clearer, and provided a range of helpful suggestions, including clarifying when central estimates should be used, and circumstances outside of market or economic shifts that should be examined to check that a scheme’s design remains sound. We’ve taken this feedback on board and sought to refine the expectations in this area.
Member options
Some respondents asked for the code or additional guidance to provide more detail on how to approach member options in CDC schemes. Member options are a matter for the trustees and scheme rules, and will be approached differently depending on the scheme’s design and the wishes of the employer and trustees. We’ve therefore provided additional expectations concerning trustees’ understanding of the impact of the member options offered, both on the scheme’s design and on members. We will monitor the approach to member options in CDC, both as the single-employer regime evolves and as the broader CDC landscape develops.
Post-valuation experience
Some respondents wanted us to restrict the option for benefit adjustments to reflect post-valuation experience to exceptional circumstances only. We believe this is a matter for the trustees and scheme actuary, and that it may generally only be appropriate in exceptional circumstances. However, we don’t wish to set a restriction at this stage.
Viability certification
Some respondents requested further clarity on the scheme actuary’s role in member communications. We think the code provides sufficient clarity already, although we’ve sought to make explicit which parts of member communications the scheme actuary is required to examine.
Many respondents agreed that provision of a template for the viability certificate was helpful. A number of respondents proposed additional notes to be included in the template to aid understanding of the nature of the certification, and set out what the actuary’s opinion has been based on. Respondents proposed an explicit statement in the certificate to confirm that the actuary’s provision of a certificate does not preclude the trustees from winding-up or closing the scheme, or that benefits won’t be reduced in the future.
We agree with respondents that the viability certificate template does not form part of a code of practice, and we’ll be publishing the final template soon, after further discussion with interested parties. This final version will take on board respondents’ suggestions for additional notes that seek to provide clarity on the nature of CDC certification.
Appendix: List of respondents to the consultation
We received 24 responses in total. 21 of the respondents gave their consent to be listed, with three either requesting confidentiality or not specifying consent at all.
List in alphabetical order
Peter Williams on behalf of Aon
Peter Williams (Chair of ACA Pension Schemes Committee) on behalf of Association of Consulting Actuaries
Association of Pension Lawyers
Association of Professional Pension Trustees
Philip Bennett, Visiting Professor of Pensions Law at Durham University
Adrian Boulding
Dave Ward and Terry Pullinger on behalf of Communication Workers Union
Equiniti
Eversheds Sutherland
First Actuarial LLP
Kathryn Fleming on behalf of Hymans Robertson LLP
ICAS
Iain McLellan on behalf of Isio
Steven Taylor on behalf of Lane Clark & Peacock LLP
Matthew Hall on behalf of Railpen
Redington
Royal Mail Group and the prospective trustees of the Royal Mail Collective Pension Plan
David Pitt-Watson and Dr Harinder Mann co-chairs of the RSA (Royal Society of Arts) CDC Pensions Forum
The Society of Pension Professionals
Squire Patton Boggs (UK) LLP (Organisation only)
Willis Towers Watson