The content on this page has been converted from PDF to HTML format using an artificial intelligence (AI) tool as part of our ongoing efforts to improve accessibility and usability of our publications. Note:

No human verification has been conducted of the converted content.
While we strive for accuracy errors or omissions may exist.
This content is provided for informational purposes only and should not be relied upon as a definitive or authoritative source.
For the official and verified version of the publication, refer to the original PDF document.

If you identify any inaccuracies or have concerns about the content, please contact us at [email protected].

Technical Actuarial Standard 310: Collective Money Purchase Pensions: Feedback Statement and Impact Assessment

The FRC does not accept any liability to any party for any loss, damage or costs howsoever arising, whether directly or indirectly, whether in contract, tort or otherwise from any action or decision taken (or not taken) as a result of any person relying on or otherwise using this document or arising from any omission from it.

The Financial Reporting Council Limited 2024 The Financial Reporting Council Limited is a company limited by guarantee. Registered in England number 2486368. Registered Office: 8th Floor, 125 London Wall, London EC2Y 5AS

Executive Summary

1Following the legislative framework for Collective Money Purchase (CMP) pension schemes being introduced through the Pension Schemes Act 2021 and the Occupational Pensions Schemes (Collective Money Purchase Schemes) Regulations 2022, trustees of a CMP scheme are now able to apply to the Pensions Regulator (TPR) for authorisation. Technical actuarial work in relation to a CMP scheme will involve significantly different considerations for practitioners than existing work relating to defined benefit or defined contribution pensions. The Financial Reporting Council (FRC) consulted on 9 May 2023 on the proposed introduction of TAS 310: Collective Money Purchase Pensions. The FRC received 12 written responses to the consultation relating to TAS 310, which were supplemented by in-depth stakeholder outreach discussions. The responses were primarily from consultancies but also included trade and professional bodies. This executive summary draws out the key areas in which feedback was provided and highlights the main changes made to the exposure draft of TAS 310 v1.0 in response to that feedback. TAS 310 v1.0 will be effective for all technical actuarial work in scope completed on or after 30 September 2024.

General observations

2The responses were mainly supportive of the policy direction of TAS 310. They agreed that a new TAS relating to CMP pensions was needed and that this should set a high standard for advice. There was also broad support regarding:

The proposed timing for introducing TAS 310.
The proposed provisions relating to assumptions.
The decision not to define soundness in the standard.
The proposed information to be included in valuation reports (Appendix A).

3CMP is a new area of actuarial work, and best practice is still emerging. In the feedback stakeholders raised concerns in a number of areas based on practical experience of practitioners. In response to this we have made several amendments when finalising the standard to ensure practitioners are comfortable that they can comply with all parts of the standard. We understand from further stakeholder engagement that these changes would address the majority of their concerns. The main areas are summarised below.

Assumptions

4The FRC proposed a definition of 'central estimate' that built on the definition in CMP legislation in response to stakeholders' desire for guidance around how to determine a central estimate basis. Respondents expressed the view that an additional definition of central estimate would add unnecessary complexity. The standard now refers directly to the legislative definition.

Modelling

5Respondents commented that certain proposed provisions relating to modelling would require a disproportionate amount of work, compared to the benefit for users. These primarily related to:

An interpretation of the draft standard that stochastic modelling would be required annually.
A requirement to consider the likelihood of future failure of live running tests.

6The FRC has amended these provisions to ensure the requirements are reduced to targeted circumstances where they provide most benefit for users but do not increase the risk to high quality actuarial information.

Credible alternatives

7Most respondents raised concerns about valuation and modelling requirements to communicate outcomes under ‘credible alternative' assumptions or modelling methodologies. There were a number of different points raised:

Some respondents questioned the number of alternatives that need to be considered and communicated.
Some respondents suggested it would be impractical or create significant additional costs to show alternative modelling methodologies and that this would give little benefit to users.
Communication of what the benefit adjustment would be on credible alternative valuation assumptions could lead to a bias towards adopting alternative, more optimistic assumptions.

8In a CMP scheme, the practitioner will often not be the decision-maker, but will be providing advice to decision-makers (typically trustees). To make appropriate decisions trustees will need to be well informed, which includes them being aware of key credible alternatives, either to proposed assumptions or to the approach used in modelling. The responses show that stakeholders have taken an onerous interpretation on the requirement which is not the intention of the FRC given paragraph 1.6 of the standard specifically refers to the application of proportionality.

9The FRC has finalised TAS 310 v1.0 with amendments to emphasise the need to apply proportionality in following these provisions. The FRC has also revised the Proportionality Guidance to include scenarios relating to demonstrating credible alternatives to support practitioners interpreting how proportionality applies.

Viability assessments

10Many respondents raised concerns about the requirements for practitioners to consider items in addition to those set out in the legislation when signing a viability certificate, citing that this would be a barrier to practitioners to signing the certificate.

11The FRC agrees that it would be appropriate for the purposes of signing a viability certificate to only consider the specific matters set out in paragraph 11(2) of the regulations, and have in finalising the standard amended provision P5.1 relating to scheme soundness to exclude work carried out for viability certificates.

12Respondents commented that in provision P5.1a, they did not believe practitioners should give a view on intergenerational fairness, explaining that while practitioners may be well placed to identify potential areas of cross-subsidy between groups of members, it is not their role to take a subjective view of what is 'fair' and that this should be the role for trustees. The FRC agrees and in finalising TAS 310 v1.0 we have updated this to refer to 'intergenerational cross-subsidies'.

Valuation

13Many respondents suggested that provisions relating to considering and communicating post valuation date experience should only apply to very significant changes post valuation. That aligns with the FRC's intention. The provisions have been amended to clarify that post valuation date experience should only be considered, and its impact communicated, where the practitioner considers not doing so to be inappropriate.

Introduction and Background

1The FRC is the UK's independent regulator responsible for issuing and maintaining actuarial standards. The FRC keeps the Technical Actuarial Standards (TASs) and other actuarial standards under regular review and reconsiders them at least once every five years.

2The FRC published a Call For Feedback in May 2022 as part of the post implementation review of the sector specific TASs, which includes TAS 300: Pensions. This was followed by the publication of a position paper in February 2023 summarising the responses to the Call For Feedback.

3In May 2023, the FRC issued a Consultation Paper titled 'Technical Actuarial Standards for Pensions', which included an exposure draft of the proposed revised standard TAS 300 v2.0 and proposed introducing a new Technical Actuarial Standard, TAS 310 which would apply to technical actuarial work carried out in relation to CMP pensions. The exposure draft of TAS 310 v1.0 was issued alongside the consultation paper. The consultation closed on 4 August 2023 and was supplemented by further outreach with a number of respondents to the consultation.

4In December 2023, the FRC issued a final version of TAS 300 v2.0, alongside FRC's response to the consultation on proposed changes to TAS 300.

5This paper provides a summary of the feedback received in respect of the proposals for TAS 310, and sets out the FRC's response, summarising material amendments to the exposure draft of TAS 310 v1.0 following the consultation. The final version of TAS 310 v1.0 is issued alongside this paper.

6The final version of TAS 310 v1.0 consists of seven sections. Sections 1 (Data), 2 (Assumptions), and 3 (Modelling) are applicable to all technical actuarial work in scope of the standard, unless stated otherwise within specific provisions. Sections 4 (Scheme design), 5 (Viability assessment), 6 (Valuation) and 7 (Factors for individual calculations) are applicable only in specific circumstances as set out in the provision.

Summary of Responses

Responses to the public consultation

1The FRC received 12 written responses, 10 of which are not confidential and have been published on the FRC website. The table below summarises the number of responses by respondent type and a list of respondents is set out in Appendix 2.

Category of Respondent	Number
Government, professional and industry bodies	4
Consultancies / professional services firms	7
Pension schemes/providers	1
Total	12

2In addition, the FRC hosted a public webinar on 31 May 2023, and held a round table and seven meetings with individual stakeholders to discuss their responses to the consultation in more detail.

3In this section we summarise the points raised in the written submissions and provide comment on the FRC position, which is informed by the outreach meetings.

Question 10: Do you have any comments on our intention to have an effective date for TAS 310 of within one year of the first CMP scheme being in operation? Is there an alternative timing that would be more appropriate? Please provide any supporting evidence for alternative timings

410 of the 12 respondents answered this question.

5Most respondents agreed on the need to have TAS 310 in place as soon as practical. Of these, many respondents commented that work is ongoing around the potential design of future CMP schemes, and it would be appropriate for TAS 310 to be in place for this work. A small number of respondents disagreed with the need for an actuarial standard in respect of CMP at the present time, suggesting that the standard should not be issued until the market had developed further.

6While most respondents were supportive of the need for TAS 310 to be put in place as soon as practical, around half also highlighted that given comments they raise on other elements of the exposure draft of TAS 310, there would need to be adequate time to consider and address these before the standard is finalised.

FRC response

7The FRC agrees that it would be appropriate for TAS 310 to be in place as soon as possible given there is actuarial work being carried out in this area. Through our post-consultation engagement, we have understood and considered comments made by respondents on the exposure draft. These are discussed further in this paper.

8The FRC disagrees with the suggestion that a Technical Actuarial Standard for CMP schemes is not currently required. CMP is a new area of actuarial work and the work of actuaries is central to ensuring sound schemes are set up and maintained. The legislation relating to CMP schemes places specific responsibilities on scheme actuaries, and it is necessary that the standard of actuarial work in these areas is appropriate to protect scheme members.

9The FRC has therefore finalised TAS 310 with an effective date of 30 September 2024, being earlier than the anticipated 1-year anniversary of the first CMP scheme being in operation.

Question 11: Do the proposed provisions provide sufficient clarity of requirements for practitioners to set central estimate assumptions? Please set out any areas of setting CE assumptions you believe require further provisions, including reasons for these.

1010 of the 12 respondents answered this question.

11Most respondents supported, or did not comment on, the provisions in section 2 of the exposure draft relating to setting central estimate assumptions. Around half of the respondents commented that the central estimate definition in the exposure draft being different from that in the legislation could cause confusion. Some respondents also commented that the definition of central estimate in the exposure draft referring to the likelihood of experience being higher or lower than the assumption was not compatible with some demographic assumptions.

12Many respondents suggested the definition of central estimate in TAS 310 should either be the same as in legislation or refer directly to the legislation.

13Around half of the respondents commented that the provisions in section 2 of the exposure draft did not reflect that practitioners would be more likely to be advising on assumptions rather than setting them and suggested that the wording should be amended accordingly.

14One respondent commented that reading the provision P2.3 which referenced the term structure of assumptions together with the consultation paper paragraph 3.9 could be seen to be endorsing gilt-based assumption setting and suggested that more neutral wording be adopted. No other respondents raised a concern on this provision.

15One respondent commented on provision P2.4 relating to the requirement to use 'as much relevant information as is sufficient' when setting central estimate assumptions. They suggested that in some cases sufficient information may not be available, and so the provision would not be possible to comply with, unless further reference to proportionality were added.

FRC response

16A number of respondents to the Call For Feedback requested additional guidance on how to determine a central estimate basis. In response to this, the exposure draft introduced provisions P2.2 to P2.4 and a definition of central estimate that expanded on the definition set out in legislation. The responses to the consultation have demonstrated that the existence of the expanded definition has the potential to cause confusion. The FRC has therefore finalised the standard such that the definition of central estimate refers to the definition in legislation. We consider that the definition as set out in legislation, in combination with provisions P2.2 to P2.4, is sufficient to ensure advice relating to central estimate assumptions is appropriate and is compatible with demographic assumption setting.

17The FRC recognises that typically the role of the practitioner would be to advise on setting assumptions, but there may be circumstances in which the practitioner is responsible for setting assumptions. We have amended provision P2.2 and P2.3 in finalising the standard to allow for either situation.

18Provision P2.3 in the exposure draft requires practitioners to consider how to make allowance for the term structure of assumptions. Although a gilt-based approach was discussed in the consultation paper as an example of considering term structure, TAS 310 does not specify this approach is required. Recent experience has demonstrated that the view on assumptions such as inflation may be very different in the short and medium-term compared to the long-term. In addition, it would be typical for mortality assumptions to make different assumptions for short-term and longer-term rates of improvement. It would be inappropriate for a practitioner not to consider how any differences in the short and long -term may impact their advice in relation to setting assumptions. They may still, after appropriate consideration, choose to make no specific allowance for the term structure. The majority of respondents to this question did not raise concerns with this provision. We have finalised the standard retaining this requirement.

19The majority of respondents did not raise any concerns with provision P2.4. The FRC expects practitioners to apply proportionality in determining what level of information is sufficient for setting central estimate assumptions. The application of proportionality is encouraged through paragraph 1.6 of the standard which references the relevant guidance. We have finalised the standard with no changes to provision P2.4.

Question 12: What are your views on the proposed provisions in relation to CMP modelling? Do you expect the proposed requirements on communication to support intended users in making relevant decisions based on modelling? Do you believe there are further items where additional requirements would be appropriate?

2011 of the 12 respondents answered this question.

Relevance for deterministic modelling

21One respondent noted that provisions P3.1 and P3.2 in the modelling section appeared more suited to stochastic modelling and were inappropriate for annual valuations which would be expected to use deterministic modelling. This observation was also made in two of the meetings we held with respondents during the consultation period. It was suggested that this section of the draft standard be amended to directly relate to stochastic modelling.

Modelling uncertainty

22Seven respondents commented that the wording in provision P3.2 suggesting models should be able to 'identify scenarios (including probabilities)' relating to certain events happening, appears to mix deterministic modelling (which may consider specific scenarios) and stochastic modelling (which considers probabilities of events).

23The majority of these seven respondents also commented that provision P3.10 as written in the exposure draft refers to 'explaining scenarios' rather than following the wording 'identify scenarios including probabilities' used in the equivalent provision P3.2. Respondents queried the reason for this difference and suggested a consistent approach be followed.

24Provision P3.2 required practitioners to use models which are able to demonstrate the impact of assumptions not being born out in practice, and identify the likelihood of certain events as listed in subclauses (a), (b) and (c). Most respondents commented on one or more of these conditions:

Two respondents did not believe it appropriate to require modelling of the likelihood of negative real benefit adjustments (as per P3.2.b), as these are a design feature of CMP and it should be expected that such events would be a common occurrence.
Five respondents suggested that asking practitioners to model when ‘live running tests are failed' (as per P3.2.c) was disproportionate and of limited use to trustees after a CMP scheme had received initial authorisation.
Three respondents recommended that provision P3.2 should be amended to refer to the time period over which future events should be considered, although of these, two commented that it should be the role of the practitioner to select an appropriate time period.
Two respondents suggested that paragraphs P3.2 a, b and c should set out examples of events to consider modelling rather than specifying that each should be modelled.

Annual stochastic modelling

25Six respondents commented that provision P3.3 implies that stochastic modelling is required annually. It was commented that this provision P3.3 was inconsistent with:

Guidance from the Pensions Regulator.
Paragraph 3.18 in the consultation paper stating that alternative approaches to stochastic modelling are permissible.
Provision P5.3 in the exposure draft which covers circumstances when stochastic modelling has not been updated.

Communication provisions

26Most respondents commented on the use of the term 'credible alternatives' in provision P3.5 (P3.6 in the final standard). Respondents gave the view that the requirement to communicate 'how recommendations and decisions could be materially different under credible alternative modelling methodologies or modelling assumptions' would be impractical or create significant additional costs with little benefit for users. The questions raised on this point were in respect of:

The number of different alternatives which may need to be communicated, given a practitioner may believe there to be a large range of possible alternatives that could be viewed as credible.
Whether it would be necessary to understand and comment on the detailed modelling methodologies used by other firms (for instance their asset projection models), in order to satisfy this requirement.

27Four respondents requested that provision P3.6b (P3.7b in the final standard), requiring practitioners to communicate which variables had been modelled stochastically and why, be extended to also cover which variables had not been modelled stochastically.

Other comments

28Four respondents commented that in provision P3.1 it was not necessary to refer to the 'complexity' of scheme benefits, as this paragraph should apply equally for schemes with relatively simple benefit structures.

29A further respondent questioned why provision P3.1 required practitioners' models to demonstrate the level of uncertainty in future benefit adjustments, suggesting that that the recent history of benefit adjustments would indicate the uncertainty from year to year.

30Six respondents commented that provision P3.4 was unnecessary. Five of these respondents stated that they are not sure what the provision is intending to achieve, and that 'clearly changing the underlying model could result in significantly different modelling results but simply confirming that this is the case (which would appear to satisfy this requirement) would not be of particular benefit'. Two of these respondents suggested the requirements of this provision were already addressed within TAS 100 but did not reference specific provisions other than all the requirements on assumptions and modelling.

31One respondent was supportive of the requirement that users understand the implication of material modelling assumptions used but suggested that since provision A7.3 of TAS 100 already covers this, provision P3.5 (P3.6 in the final standard) was unnecessary.

32A number of respondents suggested a focus on downside risks in provision P3.2 and elsewhere in the draft standard could introduce bias towards prudence into the decision-making process in relation to modelling.

FRC Response

Relevance for deterministic modelling

33While there will be many situations where models used for projections of CMP schemes will need to consider the uncertainty of future outcomes, there will be circumstances, often for annual valuations and recommending benefit adjustments where a deterministic model is appropriate.

34To provide advice a user can place reliance on, the FRC considers that models used for assessing the soundness of a scheme or those employed during benefit design should be able to consider the uncertainty of future outcomes. We have updated provisions P3.1 and P3.2 to only apply in work for testing soundness or benefit design.

Modelling uncertainty

35The nature of stochastic modelling is that outputs are in terms of probabilities of particular future events happening. The FRC's intention with provision P3.2 was for practitioners to supplement this stochastic modelling by identifying, in general terms, the types of scenarios that may lead to these outcomes and communicating these to the intended user to support their decision-making. For example, certain negative scenarios may be more likely to occur as a result of adverse experience of one of the stochastic variables. The feedback from the consultation suggests this was not clear in the exposure draft.

36During the stakeholder outreach we outlined the policy intent covered above. Practitioners indicated they were comfortable with requirements for identifying broad themes of scenarios which lead to specific events occur, in addition to carrying out stochastic modelling to estimate probabilities of those events occurring. Practitioners did, however, indicate that it would be important to show a clear separation of these two exercises in any revised standard.

37We have updated provision P3.2, and the equivalent part of P3.10 (P3.5 in the final standard) to clarify our intention that estimating the probabilities are the primary focus and that only indicative scenarios need to be considered. This change also addresses the inconsistency between provisions P3.2 and P3.10 (P3.5 in the final standard) discussed in paragraph 23 above.

38As set out in paragraph 3.16 of the consultation paper, the value members expect from CMP pensions will be tied to the purchasing power of the pension, and so it is important from an actuarial perspective to consider how this may develop. For example, a 5% benefit adjustment increase when inflation is 10% has an equivalent impact on the value of a pension to a member as a 3% decrease when inflation is 2%. The FRC considers it important that decision-makers understand such interactions, which cannot be explored by only considering nominal pension amounts. Actuaries carrying out modelling of CMP schemes will be best placed to consider the potential future uncertainty in real values of future pensions, and this should be a part of their consideration and communication to decision-makers. To consider this, models should have the capability to provide output comparing benefit adjustments and inflation as set out in provision P3.2b.

39The majority of respondents did not raise any concerns with the reference to negative real increases in provision P3.2b. From discussions on this point in our stakeholder outreach, participants did not disagree with the need to consider benefit adjustments relative to inflation. The FRC has retained the wording as per exposure draft.

40As discussed in paragraph 3.13 of the consultation paper, the intent of provision P3.2c is to ensure that trustees understand the circumstances under which a CMP scheme may not be able to continue in its current form. However, the FRC accepts that once initial authorisation of a CMP scheme has taken place, there may be little practical action the trustees can take based on knowing the risk of future failure of the live running tests. The FRC agrees it would be disproportionate to require this after initial authorisation. The FRC has finalised the standard with amendments to:

Limit the scope of provision P3.2c of the exposure draft to initial viability assessments.
Relocate the content of provision P3.2c in the exposure draft to the viability assessment section of the standard as a new provision P5.4 to reflect the scope of this provision only being relevant to a viability assessment.
Make corresponding scope and location changes to the equivalent communication provision P3.10c in the exposure draft (now as provision P5.6).

41We do not consider that it would be appropriate for the FRC to specify time periods that should be used for modelling of CMP schemes. To maintain a proportionate standard we do not consider it necessary for TAS 310 to specifically ask practitioners to consider the time period used in their modelling, as this is sufficiently addressed by Principle 5 in TAS 100 requiring models to be fit for purpose.

42The FRC considers that subject to the change to P3.2c described above, the circumstances described in P3.2 a and b should be identifiable in any modelling of CMP schemes to enable decision-makers to make fully informed decisions relating to scheme soundness. As such we do not agree that they should only be expressed as examples. If there are specific circumstances meaning that any of P3.2 a and b are not complied with, this must be justified by the practitioner. The FRC has retained the wording as per exposure draft.

Annual stochastic modelling

43It is not the FRC's policy intention to require stochastic modelling of CMP schemes to be carried out annually, and we agree this would be inconsistent with TPR's code of practice. The responses have indicated that provision P3.3 as drafted could be interpreted as requiring annual stochastic modelling. In finalising the standard the FRC has amended P3.3 to provide clarity of our regulatory expectation that stochastic modelling would be required for initial authorisation and scheme design. Stochastic modelling may or may not be required for subsequent testing of soundness after consideration of provision P5.3 (which allows for circumstances where it is not necessary to update the stochastic modelling).

Communication provisions

44As set out in paragraph 3.21 in the consultation paper, the FRC expects practitioners to provide sufficient information for intended users to understand the output and uncertainty of any modelling carried out through provisions P3.5 to 3.10. In order for intended users to make an informed decision based on advice received from a practitioner, it is important that the intended users understand the limitations of that advice. This would include understanding the dependence of the output of the modelling both on key input assumptions (for example central assumptions for asset returns in a stochastic model) and any implicit assumptions in how the model itself has been built (which could include for example the distribution of stochastic variables, or any mean reversion). The FRC does not expect the communication requirements in provision P3.5 (P3.6 in the final standard) to create additional work or costs beyond that required to comply with provision P3.4 in making decisions appropriate for the modelling.

45The intention of provision P3.5 (P3.6 in the final standard) is not for practitioners to present the intended user with a complete list of credible alternatives, nor to present detailed results under alternative modelling methodologies (unless requested by the intended user) as it is unlikely to be compliant with TAS 100 P7.3¹. The provision as drafted allows for proportionality to be exercised by the practitioner by use of the word "materially different”. Furthermore, the FRC considers certain points relevant to this application of proportionality:

P3.5 (P3.6 in the final standard) does not explicitly require communication of modelling results under alternative methodologies or assumptions, and the provision could be met by descriptive commentary on key alternatives.
Paragraph 1.6 of the standard refers to consideration of proportionality.
There is no stated requirement to consider how another firm may carry out equivalent modelling.

46During the stakeholder outreach, respondents understood the FRC's intention in P3.5 around communicating the impact of different modelling assumptions or methodologies as stated above and examples were provided of this being done in practice. Respondents were, however, unclear how to comply with this provision in a proportionate way. In order to provide further clarity, the FRC has finalised TAS 310 v1.0 with amendments to provision P3.5 (P3.6 in the final standard) to emphasise the need to only consider key credible alternatives. In addition, we have revised our Proportionality Guidance to include scenarios relating to this situation to support practitioners interpreting how proportionality applies.

Other comments

47The FRC included provision P3.6b on communicating which variables were modelled stochastically to allow the intended user to understand decisions made on the modelling approach. We agree that there is merit to also including information on the material assumptions which have not been modelled stochastically. The FRC has finalised TAS 310 with provision P3.6b (P3.7b in final standard) amended accordingly.

48In the exposure draft, communication provisions P3.5 and P3.10 related directly to communication of work described in provisions P3.4 and P3.2. While the feedback received suggests the relationship between provisions P3.2 and P3.10 was clearly understood by respondents, the responses covered in paragraphs 30 and 31 above surrounding the intention of P3.4 suggest the relationship between provisions P3.4 and P3.5 was not well understood. In order to provide greater clarity in how the communications requirements follow from the modelling provisions, the FRC has finalised TAS 310 v1.0 by relocating provision P3.10 to the start of the communication subsection (becoming P3.5 in the final version) so that the ordering of the 'Communication' subsection follows that of the main 'Modelling' section. Provisions P3.5 to P3.9 in the exposure draft have then been renumbered P3.6 to P3.10.

49With regards to the comments on the reference to the 'complexity' of scheme benefits in P3.1, while the FRC considers CMP schemes will often be complex, there is not a need within the technical standard to reference the level of complexity, and it is sufficient simply to require that models should reflect the scheme benefits. The FRC has amended the text in P3.1 accordingly in finalising the standard.

50The FRC does not agree a record of benefit increases over previous years is sufficient without other forward-looking analysis and modelling to demonstrate the uncertainty in a CMP scheme and this view was only held by one respondent to the consultation. We have retained the requirement in P3.1 that models 'demonstrate the level of uncertainty in relation to future benefit adjustments'.

51There appears to have been widespread misunderstanding of the policy intention of P3.4. As mentioned in paragraph 3.20 of the consultation paper, the FRC expects the practitioner to have a clear understanding of the decisions in relation to modelling methodology and modelling assumptions to which the output is most sensitive, and proposed provision P3.4 requires practitioners to consider the impact on their model output. There is no communication requirement to provision P3.4 - practitioners should follow the requirements of P3.4 to understand their model, and the uncertainties in the modelling. Provisions P3.5 and P3.6 of the exposure draft (P3.6 and P3.7 in the final standard) then relate to communicating this to the intended user. To improve clarity, the FRC has amended the text in provision P3.4 to more clearly align with the communication requirements of P3.5 (P3.6 in the final standard).

52All technical actuarial work carried out by Institute and Faculty of Actuaries (IFoA) members within the geographic scope is within the scope of TAS 100 v2.0. The sector specific TASs provide particular applications in the relevant sector to clarify our expectations of practitioners in meeting the reliability objective. While respondents did not specify the specific provisions in TAS 100 v2.0 which they believe make provision P3.4 unnecessary, they did make references to models and assumptions being fit for purpose and the need to identify material bias. The FRC's intention is that while the requirements of P3.4 in TAS 310 align with Principle P2.2 in TAS 100, P3.4 of TAS 310 extends this directly to the implications of alternative approaches on the results of their modelling. The FRC considers that this is an important requirement for CMP as users are expected to place significant reliance on the results of this modelling.

53The FRC considers that understanding the impact of modelling methodologies or modelling assumptions on recommendations and decisions is of such significance to the reliability of the actuarial information relating to CMP schemes that it is appropriate to include requirements relating to this in provisions P3.4 and P3.5 (P3.6 in the final standard).

54The standard does not prevent practitioners from considering any other risks to CMP schemes that might arise from upside scenarios. We expect practitioners apply judgement to consider any such risks they feel are relevant in addition to those mentioned in the standard. We have not added reference to upside scenarios in provision P3.2.

Question 13: What are your views on the proposed provisions in relation to Scheme design? Do you envisage any difficulties in meeting the requirements of these provisions. Please provide details to accompany your response.

5510 of the 12 respondents commented on this question.

56The majority of those who responded to this question stated that the requirement to use data 'which is as comprehensive as possible' could create an onerous requirement, and may be interpreted as a requirement to collect all conceivable data. Respondents suggested a more proportionate approach should be followed with the level of data needed at preliminary design stages being significantly less than for a scheme approaching authorisation.

57A number of respondents gave the view that it was not appropriate to distinguish between when work was considered scheme design and when it proceeded to be part of the preparation for the application of authorisation, which would be subject instead to requirements under Section 5 of the standard.

FRC response

58The intention of the requirements in section 4 of the exposure draft, as set out in paragraph 3.28 of our consultation paper, was to set expectations for how a practitioner working on scheme design should deal with limitations in the data available at that stage. Available data may be significantly more limited for preliminary analysis exploring a potential CMP scheme than for a scheme in the process of applying for authorisation. Our expectation was that data obtained should be as complete as possible considering the stage of the work and may, for instance, include only limited or projected membership details. This is distinct from the position where a scheme is seeking authorisation, where there is likely to be more complete data on membership and investment strategy.

59Responses provided show that the term 'as comprehensive as possible' has been interpreted as suggesting a higher expectation, possibly requiring practitioners to acquire all conceivable data. In light of these responses, the FRC has amended provision P4.1a in finalising the standard to clarify the intention, relating the requirement to the stage in the design process, and the decisions to be made, thus making it more proportionate.

Question 14: What are your views on the proposed provisions on completing assessments of scheme viability and certifying soundness? Do you consider it is appropriate to require practitioners to consider areas beyond those outlined in legislation when certifying soundness? Please give reasons for your response.

Question 15: Do you agree that the considerations for a practitioner certifying scheme soundness via a viability certificate are the same as those a practitioner should communicate to trustees in their own consideration as to whether the design of the scheme is sound for their viability report?

6011 of the 12 respondents answered questions 14 and 15. We have combined responses to the two questions as they closely relate to each other.

Certifying and advising on soundness

61The majority of the respondents agreed with the FRC's decision not to define the term soundness in the standard.

62In response to question 15, a number of respondents commented on the proposed approach in provision P5.1 of combining the requirements for work carried out by practitioners in certifying scheme soundness with those for work supporting the trustees in their own opinion of soundness in their viability report:

Most of the respondents suggested that the viability certificate should be limited to purely actuarial matters, whereas it may be reasonable to include non-actuarial matters in the separate advice to the trustees. Some of these respondents referenced provisions in the IFOA's Actuarial Profession Standard APS P1² as the framework for how the viability certificate and additional advice to trustees should work.
Respondents did not believe it appropriate in considering signing a viability certificate to include any matters beyond those set out in paragraph 11(2) of the regulations³ and commented that requiring this was beyond FRC's remit.

63Respondents also made a number of comments in relation to the specific requirements set out in P5.1 a to c:

A third of the respondents suggested providing a list of issues to consider was unnecessarily restrictive and could encourage herding by practitioners.
Other respondents did not think the items listed necessarily represented soundness and expressed a preference for them to be removed. These respondents did, however acknowledge that the drafting of the exposure draft allowed practitioners to consider the factors listed under P5.1 a to c and decide they are not relevant.
Over half of the respondents commented on the inclusion of intergenerational fairness in provision P5.1a. A number pointed out that since intergenerational fairness was not defined there would be various interpretations and one respondent suggested that the trustees rather than the actuary should make judgements on intergenerational fairness. In addition, some respondents noted that in any CMP scheme design there would always be some cross-subsidies between age groups and this would not necessarily mean a scheme design is not sound.
One respondent stated that comparing observed experience in benefit adjustments to the original scheme design, as required in provision P5.1b, was not appropriate given that it is to be expected that CMP scheme increases will vary over time.
A small number of respondents commented on provision P5.1c comparing expected benefit adjustments with price inflation, stating that CMP schemes are designed to have increases that fall below inflation at times and if this occurs it does not indicate that a scheme design is not sound. This opinion was repeated during our meetings with respondents where it was suggested that this could result in practitioners being unable to sign viability certificates in such cases.

64A small number of respondents identified potential for misinterpretation of provision P5.2a. It was suggested that the wording 'all member communications that they consider are relevant' be changed to 'the member communications that they consider are relevant' to avoid the risk of practitioners interpreting this as requiring all member communications be reviewed.

65There were other comments suggesting minor drafting changes in provisions P5.1a, P5.3 and P5.4d. Several respondents commented that provision P5.4d referred to 'running tests' rather than 'live running tests' and suggested this should be amended.

Communicating risks to future soundness

66The majority of respondents commented that it would be difficult to meet the requirements of the communication provision P5.4e, relating to scenarios where the scheme would no longer be sound. Respondents had interpreted this as a requirement to comment on all the possible events which could impact soundness and so offered amended wording to the provision to narrow the requirements to more likely or plausible scenarios.

67A number of respondents also provided comments on provision P5.4f as follows:

A significant proportion of the respondents suggested that provision P5.4f be amended to address upside risk as well as downside. This was consistent with comments made more generally on the draft standard regarding a focus on downside events as mentioned in paragraph 32 above.
One respondent also commented that the requirement to determine likelihood of material risks in provision P5.4f may not be feasible for some risks where the likelihood is not quantifiable.
A separate respondent commented that forecasting future failures of live running tests as set out in provision P5.4f, is not required by legislation.

FRC response

Certifying and advising on soundness

68The FRC recognises the comments made regarding the difficulty practitioners may face in signing viability certificates and the potential impact this will have on the development of the CMP market. The FRC agrees that it would be appropriate for the purposes of signing a viability certificate to only consider the specific matters set out in paragraph 11(2) of the regulations. However, we consider that in advising trustees in their own consideration of soundness that it is not appropriate for the practitioner to limit themselves to matters in paragraph 11(2). Further, consideration of other matters as set out in provision P5.1 is necessary to meet the reliability objective and ensure that the intended user of the advice can make a fully informed decision.

69The FRC has amended the scope of provision P5.1 in finalising the standard to remove work relating to the viability certificate, leaving P5.1 only applying to practitioners' work considering soundness in advice to trustees relating to their viability report. The standard leaves room for practitioners to apply their judgement as to what sound means to them and communicate this with their intended users. It is possible that there may be evolution in what represents soundness as more CMP schemes come into existence and the level of accrued funds and benefits increases.

70The FRC does not consider that under the revised scope for P5.1, considering 'all relevant matters' including the items in a to c within advice to trustees would inhibit market growth. In our view such consideration is necessary in providing advice to trustees to meet the reliability objective and high quality actuarial work in this emerging sector is key to public confidence in support of growth. The FRC does not consider the list in P5.1 a to c will encourage herding of actuarial advice as the provision is not written to make the list exhaustive and P5.1 requires consideration of 'all relevant matters' extending beyond the specified list. It is also clear that a practitioner may consider matters a to c and, as a result of this consideration, decide that one or more of these is not relevant to soundness. The FRC has finalised the standard with no further amendments on these matters.

71Through engagement with respondents following the consultation, we understand that the amended scope of P5.1 in the final standard to remove the work in relation to viability certification addresses many of the concerns raised by respondents.

72Over half of the respondents commented on the inclusion of intergenerational fairness in provision P5.1a. During engagement following the consultation, stakeholders gave more detailed views on intergenerational fairness in the context of advising trustees on soundness:

Stakeholders were consistent in stating that while the scheme actuary may be well placed to identify potential areas of cross-subsidy between groups of members, it is not their role to take a subjective view of what is 'fair' and that this should be the role of trustees.
Intergenerational fairness may be viewed from different perspectives, for instance between different ages at the current time (for example a comparison of the value of one year's benefit accrual between a 30 year-old and a 50 year-old) or between current and future generations at the same age (e.g. the same comparison between a 50 year-old today and a 50 year-old in 20 years' time). The view given was that it would be for the practitioner to consider which comparisons may be relevant in the given circumstances.

73It is not within the FRC's remit to define “intergenerational fairness” in the absence of such a definition in legislation. The FRC acknowledges that there can be different perspectives on what constitutes “intergenerational fairness” and practitioners may devise relevant measures accordingly.

74The FRC agrees with stakeholders that the determination of what is 'fair' is a matter for the intended user, and that the role of the practitioner should be to provide decision-makers with appropriate information rather than making their own judgement on the criteria of fairness. The FRC's intention is for practitioners to consider where there are risks of material intergenerational cross-subsidies which may be relevant to their view on soundness of the scheme and to inform the intended users who would be in a position to take a view on the 'fairness' of these cross-subsidies. The FRC considers that the practitioner providing advice will be best placed to understand current and potential future risks of intergenerational cross-subsidies occurring as a result of the scheme design or future experience.

75We have finalised TAS 310 with the amendment to P5.1a to refer to 'any material intergenerational cross-subsidies'. Under the revised wording practitioners will still be in a position to apply judgement on how this is tested and on whether this is a relevant factor in determining soundness.

76The FRC's intention of P5.1b was to ensure that where observed experience of benefit adjustments in the scheme differs materially from expectations, practitioners should consider whether the scheme in its current form, when considered alongside how it is communicated to members, remains sound. We agree the point raised that over time the original intention of the scheme design becomes less relevant. We consider that the key point is that if the benefit adjustments evolve over time, it is important that there remains a clear connection between the expected benefits from the scheme and how the scheme is communicated. We have finalised TAS 310 by removing references to the original scheme design in provision P5.1b to aid clarity.

77In advising the trustees in forming an opinion on soundness, practitioners should consider the benefits in real terms as well as nominal terms. The FRC's intention with provision P5.1c is that practitioners consider whether their modelled expectations of benefit adjustments compared to inflation are consistent with members' expectations which are likely to be represented by member communications either on joining or in any updated communications issued. P5.1c does not suggest a scheme is unsound if expected future increases do not match inflation, and we agree the nature of CMP schemes is such that this may be a relatively frequent occurrence. It is important, however, that the expected increases relative to inflation are well understood and considered alongside other aspects of the scheme, such as expectations given to members through ongoing communications. If in the circumstances of the scheme the practitioner does not believe the ability to provide benefit increases at least in line with inflation is relevant to the consideration of soundness, then provision P5.1c allows for this (so long as this view is documented). We have, therefore, finalised the standard with no changes to provision P5.1c.

78The FRC's intention with provision P5.2a was to provide scope for practitioners to judge themselves which communications are relevant to their view of soundness. We consider that the reference to 'member communications that [practitioners] consider relevant' achieves this and that the potential misinterpretation highlighted in the feedback is unlikely to occur. We have however finalised the standard with the word 'all' replaced by 'the' in provision P5.2a to clarify our policy position.

79We have amended P5.4d (P5.5d in the final standard) as suggested to clarify that this relates to live running tests.

Communicating risks to future soundness

80In relation to P5.4e (P5.5e in the final standard), the FRC's intention is not for practitioners to communicate all possible scenarios, but only likely scenarios. We have amended the provision accordingly in finalising the standard.

81In support of the reliability objective, the standard aims to ensure decision-makers understand the potential risks that may have a negative impact on members' benefits (especially in the case of CMP schemes their benefits relative to expectations that have been given). As a result, it is appropriate there will be a focus on negative events that could cause benefits to fall below members' expectations. However, we also recognise that there could also be upside scenarios that could impact on the practitioner's view of the soundness of the scheme. We have finalised the standard with provision P5.4f (P5.5f in the final standard) updated to address both upside and downside risks.

82The FRC notes that the view expressed surrounding the use of the term 'likelihood' in provision P5.4f by one respondent was not shared by other respondents and so we expect that practitioners are comfortable applying proportionality in complying with this provision. Further, we would expect practitioners to be providing similar information under the risk identification principle of TAS 100 and application A7.2⁴, and so provision P5.4f is a specific example of the application of this principle in the case of CMP pensions. We have therefore finalised the standard without further amendments to this provision.

83We note that modelling future failures of live running tests is not required by legislation, but descriptions of scenarios that could lead to failed live running tests would be a useful part of practitioners ensuring users understand risks to future soundness. The inclusion of the live running tests in legislation mean that practitioners will carry out modelling in this area and we expect them to apply proportionality when considering how to communicate the risk of their failure when complying with provision P5.4f. We have therefore finalised the standard without further amendments to this provision.

Question 16: Are there any other areas in relation to soundness (including practitioners' communications of their work on soundness) which require further standards? Please provide as much detail as possible.

84One respondent answered this question.

85It was suggested that it would be helpful for trustees to periodically survey members' understanding of soundness in order to gauge whether existing soundness measures are connecting with members.

FRC response

86While the FRC agrees that it would be beneficial to understand members' interpretation of soundness, we consider that this is a trustee matter, so it does not directly affect the work of practitioners for CMP schemes. This means it is out of the scope of TAS 310 and we have made no changes to the exposure draft for this comment.

Question 17: What are your views on the proposed provisions on actuarial valuations for CMP schemes? Are there other key areas of judgement beyond the central estimate assumptions? Are there further areas you would expect to be included? Please give reasons for your response.

879 out of 12 respondents answered this question.

Credible alternatives

88The majority of respondents expressed concern around considering and communicating 'credible alternatives' in provisions P6.1b and P6.2a:

Some respondents questioned the number of credible alternatives that may need to be considered and communicated. For example, some respondents asked if the intention of the provisions was for practitioners to look at credible alternatives to each assumption.
A number of respondents commented that the requirement to show what the benefit adjustment would be on credible alternative valuation assumptions could lead to pressure on practitioners to adopt alternative, more optimistic assumptions.
Some respondents questioned the benefit of providing such information for trustees.
Some respondents suggested that the requirement relating to 'credible alternatives' should be replaced with a requirement around illustrating sensitivities to material assumptions.

Post valuation date experience

89Most of the respondents to this question raised concerns with the proposed provisions P6.1c and P6.2b which relate to considering and communicating the impact of post valuation experience. The feedback fell into the following areas:

It would be practically difficult, and potentially add to costs with little benefit, if practitioners are always required to consider allowing for post valuation date experience since it is a constantly moving target. It was noted that this may risk valuations not being completed annually. It was suggested that post valuation date experience should in most circumstances instead be captured in the following valuation.
The use of the word 'material' in provision P6.1c, as defined in the glossary to TAS 310, suggests that any post valuation date experience which may lead to a different level of benefit adjustment should be considered. It was commented that the legislation allowing for considering post valuation date experience had been drafted with extreme market movements in mind. Several respondents suggested amending the wording in this provision so that it applies only when allowing for post valuation date experience would lead to a large change to the benefit adjustment.
A small number of respondents suggested that the requirements of P6.1c should not be explicitly included in TAS 310 but rather left up to practitioners' judgement.

Other comments

90All of the respondents to this question disagreed with, or suggested amendments to, the requirements of provision P6.1a to consider the consistency of actuarial valuation assumptions with those used for the first gateway test. Respondents stated that as time progresses the assumptions for the first gateway test become less relevant and that the provision, if included, should instead simply seek consistency with the most recent valuation basis (or the first gateway test only when carrying out the first annual valuation of the scheme).

FRC response

Credible alternatives

91The FRC considers that the use of the term 'credible alternatives' does not create any additional requirements to approaches practitioners routinely follow when satisfying the judgement principle of TAS 100 P2.2⁵ and A7.3b. We would expect practitioners to apply proportionality when considering the range of credible alternatives to be reflected when complying with provisions P6.1a (previously P6.1b) and P6.2a.

92The FRC understands the risk highlighted by respondents of users of actuarial work choosing to use what they consider to be a more 'favourable' set of assumptions than the practitioner's 'central estimate'. We are, however, conscious that the assumptions used in a valuation are the decision of the trustees, and that in order to make appropriate decisions they need to be well informed. This includes their understanding of key credible alternative assumptions, and we consider that the presentation of the sensitivity of results to assumptions without information on how credible they are, would be insufficient to meet this objective. We expect that a practitioner complying with provisions P6.1 and P6.2 will explain to the user the context in which the key credible alternative assumptions are being presented and the appropriateness and potential impacts of adopting them compared to the practitioner's recommended central estimate assumptions.

93Given above two points, the FRC has amended provisions P6.1 and P6.2 in the final standard to emphasise the need to only consider key credible alternatives for material assumptions and the addition to the Proportionality Guidance mentioned in paragraph 46 above will provide additional support on the application of proportionality.

Post valuation date experience

94The FRC's intent is for practitioners to consider only post valuation date experience that would result in significantly different benefit adjustments when advising on post valuation date experience. We agree with the comments provided that the use of the word 'material' in provision P6.1c as currently drafted does not achieve this. Practitioners should apply judgement when deciding whether to recommend the trustees consider post valuation date experience considering factors such as the size of the estimated impact on the recommended benefit adjustment and the period to the next annual valuation effective date. The requirements of P6.1c and P6.2b form part of providing trustees with sufficient information to fulfil their duties with regards to post valuation date experience as required by the legislation⁶.

95The majority of participants in the post-consultation outreach suggested they were comfortable with the reference to post valuation date experience in P6.1 and P6.2 so long as it was made clear this would only apply in the case of experience that the practitioner considers to be significant.

96In finalising the standard, the FRC has amended provisions P6.1c (now P6.1b) and P6.2b to clarify that post valuation date experience should be considered, and its impact communicated, only in situations where not doing so is considered by the practitioner to be inappropriate.

Other comments

97The FRC agrees that over time, the relevance of the assumptions adopted for the first gateway test will diminish. Provision P6.1a was intended to avoid any inconsistency between valuation assumptions and the first gateway test. Following discussions with respondents in stakeholder outreach meetings, we understand that this risk is minimal as:

gateway tests and annual valuations are to be carried out on 'central estimate' assumptions, which are unlikely to differ when the scheme is established; and
subsequent to this first valuation, application of TAS 100 A7.5b⁷ would require the practitioner to consider and communicate consistency of assumptions between valuations.

98The FRC has removed P6.1a when finalising the standard and adjusted numbering of subsequent paragraphs accordingly.

Question 18: Do you agree the required content of the valuation report set out in Appendix A is reasonable for CMP schemes? Is there further content which should be included?

9910 of the 12 respondents answered this question.

100Respondents were either supportive of the content in Appendix A or suggested possible drafting changes. One respondent gave the view that including Appendix A was not consistent with the principles-based approach to Technical Actuarial Standards but did not disagree the content of Appendix A. A separate respondent suggested the items in Appendix A should be included in the relevant CMP legislation. No other respondents opposed including Appendix A within TAS 310.

101Comments made on Appendix A as drafted were:

A small number of respondents suggested that the explanation of experience covered in point f should be extended to cover the quantification the impact of experience items.
Many of the respondents commented that item h in under the 'risk and uncertainty' heading should be restricted to material risks. One respondent suggested the word 'risk' in this point was inappropriate as it implies an emphasis on downside events and that item h should be deleted.
Respondents asked for clarification that the requirements of Appendix A were subject to the guidance on proportionality.

FRC response

102The FRC considers Appendix A to be equivalent to similar provisions in TAS 300. With CMP scheme valuations being a new area of work, we consider that the appendix will help drive high quality actuarial work in this area and ensure sufficient information is included to meet the reliability objective. The FRC notes that the feedback was broadly supportive of the content of Appendix A, and considers that, in the absence of specific requirements within CMP legislation it is reasonable for TAS 310 to include these requirements. The FRC has retained Appendix A in the final standard.

103The FRC agrees that quantification of experience items that have led to the recommended level of benefit adjustment (as an expansion of paragraph f) would be an example of good practice, but have not adopted this in finalising the standard to ensure the standard is targeted and proportionate. Practitioners are free to apply their judgement in how best to explain experience to users of actuarial work, and quantifying amounts is one such approach.

104The FRC expects practitioners to apply judgement and consider proportionality when complying with Appendix A. The responses suggest that those who responded to the consultation had not interpreted these provisions as such and were not comfortable with applying proportionality and materiality. The inclusion of the word 'material' in the opening sentences of Appendix A supports this, but we accept that for item h in particular, respondents are seeking explicit reassurance on applying materiality. In finalising TAS 310 we have set the defined term 'material' as bold to provide clarity, and added the word 'material' to item h to clarify our expectations.

Question 19: What are your views on the proposed provisions in relation to factors for CMP schemes? Do you envisage any issues complying with provision P7.4 regarding selection risk? Are there certain groups of members you believe this may disadvantage? Please provide reasons for your response.

10510 of the 12 respondents answered this question.

Cost neutrality

106Around half of the respondents noted that scheme rules may require that actuarial factors follow an approach other than central estimate and recommended provision P7.2 clarify that the requirements are subject to scheme rules.

107A number of respondents expressed the view that, as drafted, provision P7.2 could be interpreted as requiring that factors are advised on or set in a manner that is cost neutral in every aspect. Through outreach with respondents, we understand that this could be interpreted to require frequent reviews of factors to maintain cost neutrality as market conditions change. Some respondents suggested alternative wording such as referring to 'principles of cost neutrality'.

108A small number of respondents requested that the standard clarifies that the references to cost neutrality in provisions P7.2 and P7.5 do not require the consideration of sex-specific factors where trustees wish to adopt unisex factors.

109One respondent was not supportive of including P7.2 in the standard, suggesting that setting an expectation of cost neutrality was narrow when compared to the approach required in TAS 300 relating to setting factors for defined benefit pensions of illustrating a range of different bases. This respondent suggested that specifying a preference for cost neutrality was outside the FRC's remit.

110A small number of respondents suggested the reference to a 'share-of-the-fund basis' in provision P7.3 was unnecessary as legislation already requires cash equivalent transfer values for CMP schemes to be calculated on a share of the fund basis.

Selection risk

111Two respondents commented on the appropriateness including selection risk in provision P7.4:

One respondent commented that the consideration of selection risk should apply more broadly than just in setting actuarial factors, extending this to work relating to soundness and annual valuations, and that this is sufficiently addressed in TAS 100 Principle 1 Risk identification.
The other respondent suggested that CMP schemes would not have sufficient evidence to support making an allowance for selection in actuarial factor setting.

Other comments

112One respondent suggested that in P7.1 the FRC should reference a typical period between factor reviews in a similar manner to that stated in TAS 300 for defined benefit schemes. They suggested that the FRC's expectation stated in the consultation paper that annual valuations and viability certification exercises should drive regular reviews should be monitored as evidence emerges.

FRC response

Cost neutrality

113P7.2 is a regulatory expectation rather than a mandatory requirement denoted by the use of the word 'should'. If scheme rules require that factors are set using an approach other than a central estimate basis, then a practitioner would be required to follow the scheme rules, and this would be suitable justification for a deviation from the requirements of provision P7.2. We have finalised the standard with this provision unchanged from the exposure draft on this aspect.

114The FRC does not intend for provision P7.2 to be interpreted as requiring factors to be cost neutral in every aspect. We expect practitioners to consider proportionality when considering how to apply provision P7.2 and form their own judgements in appropriate areas, for instance whether to allow for a member's sex in setting the factors, or how to allow for varying market conditions. We note that provision P7.5 requires practitioners to communicate the reasons for any deviations from cost neutrality when this is the case, which implies the expectation of practitioners exercising judgement on cost neutrality while providing sufficient information for decision-makers. The FRC has finalised the standard with minor changes to provision P7.2 to emphasise the need to apply proportionality.

115The FRC does not consider P7.2 to be too narrowly targeted at cost neutrality and that this goes beyond the FRC's remit. This view, from a single respondent is not aligned with most respondents' expectations. Use of factors which are not calculated on a cost neutral basis may to lead to individuals receiving a benefit which differs from their share of the fund and may create advantages and disadvantages to groups of members. This requirement is consistent with legislative requirements relating to cash equivalent transfer values as covered in paragraph 116 below. We have retained this provision P7.2 in the final standard with the minor changes on proportionality mentioned above.

116The FRC's intention with provision P7.3 was to ensure practitioners consider fully the practicalities involved in calculating cash equivalents based on a share of the fund, when the total fund value itself can change over time. We agree that the legislation on CMP schemes and that on Transfers Values mean that all CMP transfer values will be calculated based on, or with reference to, a share of the fund basis. We also consider that were circumstances to arise where an approach other than share of fund was followed, the provisions as drafted in P7.3 would be equally applicable. We have amended this provision to remove reference to share of fund when finalising the standard.

Selection risk

117Selection risk has potential to result in members not receiving an appropriate share of the CMP fund, to the extent that members choosing to exercise options have a different demographic profile than the average of the scheme. The FRC considers that selection risk is of sufficient importance in the context of CMP schemes to require explicit consideration within TAS 310, and practitioners would be expected to consider proportionate approaches to obtaining evidence to support their approach. We have retained provision P7.4 as drafted when finalising the standard.

Other comments

118The suggestion by one respondent of specifying a preferred period of time between factor reviews was not repeated by any other respondents. As described in paragraph 3.48 of the consultation paper, the FRC expects that the annual exercises for viability certificates and actuarial valuations will drive regular factor reviews. The FRC has not made amendments to P7.1 in finalising the standard. The FRC will review how practice evolves in setting factors for CMP schemes in future as part of our regular review of the standards.

Question 20: Do you agree with our impact assessment? Please give reasons for your response.

1199 of the 12 respondents answered this question in relation to TAS 310.

120A significant number of respondents stated that the impact assessment did not adequately recognise what they considered to be substantial additional work arising from provisions in the draft standard that extended beyond what was reasonably required.

121In these responses, specific examples highlighted were:

Requirements to consider the risk of future failure of live running tests (provisions P3.2c and P3.10c).
Requirements to consider and report on credible alternative assumptions and modelling methodologies (provisions P3.5 (P3.6 in the final standard), P6.1b (P6.1a in final standard) and P6.2a).
Requirements relating to post valuation date experience (provisions P6.1c (P6.1b in final standard) and P6.2b).

FRC response

122Having considered written responses and stakeholder engagement, the FRC has made amendments in finalising the standard which address the areas where respondents suggested a disproportionate amount of work was required. The FRC have made amendments to:

Remove or reduce requirements, for example for provisions P3.2c and P3.10c (P5.4 and P5.6 in the final standard) not to apply beyond initial authorisation.
Clarify the need for applying materiality or principles of proportionality in the review, for example in provisions P3.5 (P3.6 in the final standard), P6.1b (P6.1a in the final standard) and P6.2a, and included scenarios relating to 'credible alternatives' in our Proportionality Guidance.
Update the requirements to be appropriate for the requirements of users of actuarial work, for example in provisions P6.1c (P6.1b in the final standard) and P6.2b regarding post valuation date experience.

123We would expect practitioners to consider the FRC's guidance on proportionality as referenced in paragraph 1.6 of TAS 310 when applying the standard.

124Our engagement with respondents following the consultation suggests that the changes and clarifications made will address the areas where a disproportionate amount of work was expected to be required. Following these changes, we expect any costs arising from the introduction of TAS 310 to be due to the regulatory requirements for this new type of scheme, rather than arising from FRC requirements in this TAS. The updated provisions within TAS 310 are a proportionate approach to ensure decision-makers on CMP schemes receive actuarial advice meeting the reliability objective. On this basis, we consider that the Impact Assessment as set out in the consultation paper remains appropriate.

Impact Assessment

Benefits

1The introduction of TAS 310 is a result of the emergence of CMP schemes as a new way of providing pensions.

2The introduction of TAS 310 provides technical actuarial standards relating to new activities required in supporting the provision of CMP benefits following their introduction. The content of TAS 310 aims to ensure that members of CMP schemes are treated fairly and receive appropriate information on their benefits:

Scheme actuaries and trustees appropriately review the soundness of CMP pension schemes, to reduce the risk of inappropriate schemes being established, and ensure risks are communicated to scheme members appropriately;
Valuations of CMP schemes are carried out in appropriate and fair manner, and the assumptions used are appropriate; and
Members exercising options such as transferring from a scheme receive an appropriate share of the assets of the scheme.

Costs

3The introduction of TAS 310 in respect of CMP schemes has arisen following the responsibility imposed by the new legislation and TPR regime. Any costs on the actuarial profession which arise from the introduction of TAS 310 are due to the regulatory requirements for this new type of scheme. The Department for Work and Pensions set out their cost benefit analysis in their consultation on CMP regulations.

4To date, there has been only one CMP scheme authorised by TPR. With the CMP market in its infancy, it is not possible to estimate with any degree of accuracy, the frequency that the additional work may be required.

Appendix 1 – Amendments to the exposure draft

Section	Issue	Change
Section 2	Assumptions provisions should apply to practitioners advising on assumptions and not just setting them	P2.2 and P2.3 - added 'or advising on' to highlight that practitioners could be advising on assumptions
Section 3	Provisions are inappropriate for deterministic modelling used in valuations	P3.1 and P3.2 - scope of provision amended to advice relating to 'viability assessments and scheme design'
Section 3	Unnecessary reference to complexity of benefits	P3.1 - reference to complexity of benefits removed
Section 3	Mixing requirements for stochastic and deterministic modelling	P3.2 and P3.10 (P3.5 in final standard) - wording amended
Section 3	Modelling live running test failure disproportionate	P3.2c and P3.10c - scope of provisions reduced to initial assessments of soundness and relocated to be provisions P5.4 and P5.6 respectively
Section 3	Potential interpretation of P3.3 that stochastic modelling is required annually	P3.3 scope of provision amended to clarify it covers scheme design and initial viability assessments
Section 3	Relevance of P3.4 not clear to practitioners	P3.4 - wording amended to align with P3.5 (P3.6 in final standard)
Section 3	Concern over the number of credible alternatives to be considered	P3.5 (P3.6 in final standard) - worded amended to emphasise the need to consider only 'key' alternatives
Section 3	Communicating variables which have not been modelled stochastically	P3.6b (P3.7b in final standard) - amended to cover variables which have not been modelled stochastically and added reference to materiality of variables and relevance of statistical measures
Section 3	P3.10 and P3.2 wording not consistent	P3.10 (P3.5 in final standard) - amended to be consistent with the revised P3.2
Section 3	Lack of clarity over how provisions P3.1 to P3.4 related to communication provisions P3.5 to P3.10	Reordering of the communication provisions: P3.10 -> P3.5 P3.5 -> P3.6 P3.6 -> P3.7 P3.7 -> P3.8 P3.8 -> P3.9 P3.9 -> P3.10
Section 4	'As comprehensive as possible' data requirement could be interpreted as excessively onerous	P4.1a - amended to relate to the stage in the design process and decisions to be made P4.1 and P4.2 – amended wording to add clarity in respect of when these provisions apply.
Section 5	Provisions not appropriate for practitioners signing a viability certificate	P5.1 - scope reduced to cover soundness assessments to support trustees with their viability report only
Section 5	Difficulty interpreting intergenerational fairness	P5.1a - amended to use the term 'intergenerational cross-subsidies' and to clarify materiality
Section 5	The need to continue to compare benefit adjustments to the original scheme design	P5.1b - removed reference to the original scheme design
Section 5	Possible misinterpretation of P5.2a requiring practitioners to consider all possible member communications	P5.2a - the word 'all' replaced by 'the'
Section 5	P5.4d missing the word 'live' when referring to live running tests	P5.4d (P5.5d in final standard) - added the word 'live'
Section 5	P5.4e could be interpreted as onerous	P5.4e (P5.5e in final standard) - added the word 'likely'
Section 5	Perception that only downside risks should be considered in provision P5.4f	P5.4f (P5.5f in final standard) - added the words 'upside or downside'
Section 6	The requirement to consider the consistency of valuation assumptions with those adopted for the first gateway test is not necessary	Removed provision P6.1a
Section 6	Difficulty in interpreting and applying 'credible alternatives' in provisions P6.1 and P6.2	P6.1b (P6.1a in final standard) and P6.2a - wording amended to highlight the need to apply proportionality
Section 6	The requirements for post valuation date experience are not restricted to significant experience	P6.1c (P6.1b in final standard) and P6.2b - redrafted to reflect our expectations and provide clarity on materiality
Section 7	Suggestion that factors are required to be cost neutral in every aspect	P7.2 - added the word 'broadly'
Section 7	Suggests CETVs can be on a basis other than share of fund	P7.3 - removed reference to share of fund basis and rephrased
Appendix A	Interpretation that item h was not restricted to material risks	Added the word 'material' to point h
Glossary	Different definition of central estimate from that in legislation	Central estimate definition - revised to reference legislation

Appendix 2 - List of respondents to consultation

List of respondents to consultations

The FRC received 12 written responses to the consultation on proposed changes to TAS 310, 10 of which were not confidential and were published on the FRC website. The respondents were as follows:

Aon plc
Association of Consulting Actuaries
First Actuarial LLP
Government Actuary's Department
Isio Group Limited
Lane Clark & Peacock LLP
Mercer Limited
Society of Pension Professionals
Universities Superannuation Scheme Limited
WTW GB Retirement Team

Financial Reporting Council 8th Floor 125 London Wall London EC2Y 5AS +44 (0)20 7492 2300

www.frc.org.uk

Footnotes

The practitioner’s communications must exclude information that is not material if that information obscures material actuarial information, unless the inclusion of such information is a regulatory requirement. ↩
Para 3.2 – the [actuary] must draw the Trustees' attention to any matters which he/she believes the Trustees should bear in mind before taking any action associated with that certification. ↩
Paragraph 11(2) of the Occupational Pension Schemes (Collective Money Purchase Schemes) Regulations 2022 https://www.legislation.gov.uk/uksi/2022/255/regulation/11/made ↩
https://www.frc.org.uk/documents/4287/TAS_100_General_Actuarial_Standards_Version_2.0.pdf A7.2 In support of the risk identification principle, practitioners' communications should state the nature and significance of each material risk or material uncertainty faced by the entity in relation to the technical actuarial work and explain the approach taken to the risk. ↩
P2.2 'Practitioners exercising material judgement must consider credible alternative methodologies, models, data and assumptions'. ↩
19 (2) and 19(3) of the Occupational Pension Schemes (Collective Money Purchase Schemes) Regulations 2022 ↩
A7.5 In support of the assumptions principle, practitioners' communications should: b) where there was a previous exercise carried out for the same purpose, describe any change to a material assumption used in the previous exercise with an explanation of any material difference, and description of any change in the rationale underlying that material assumption. ↩

File

Name Technical Actuarial Standard 310: Collective Money Purchase Pensions: Feedback Statement and Impact Assessment

Publication date 24 May 2024

Type Feedback paper

Format PDF, 365.9 KB

Download original

Technical Actuarial Standard 310: Collective Money Purchase Pensions: Feedback Statement and Impact Assessment

File

Is this page useful?