Consultation response on proposed expectations on capital requirements, value extraction and investment restrictions for defined benefit (DB) superfunds.
Published: 18 June 2020
Background
1.1 In its March 2018 white paper, the government announced its policy intention to enable new forms of DB pension consolidation models to operate in the market. In December 2018, the Department for Work and Pensions (DWP) published a consultation paper setting out its plans for an authorisation and supervision framework in relation to models designed to sever the link with the covenant of a ceding scheme in exchange for the provision of upfront capital. These are termed superfunds.
1.2 We think consolidation within the DB sector could be beneficial, particularly among smaller schemes, which tend to be less well-governed and could benefit from economies of scale and access to a wider range of investment opportunities. Members of these schemes could also benefit from improvements in administration and stewardship. The addition of upfront contributions and investor capital would immediately improve the funding of DB pension schemes entering a superfund and remove the risk of the employer covenant deteriorating.
1.3 As part of the DWP’s proposals, they made it clear that a scheme transferring to a superfund must only do so where the trustees are convinced that members’ benefits will be more secure. The DWP also made it clear that superfunds should provide an alternative for schemes and sponsors who do not have a realistic prospect of being able to fund an insurance buy-out, either now or in the foreseeable future.
1.4 We agree with the DWP that schemes consolidating could provide better outcomes for their members and the PPF and improve funding of DB pensions. However, at the same time, superfunds represent a new form of DB pension provision. Given the absence of an ongoing employer covenant, the scheme is wholly reliant on its funding position and/or any capital provided.
1.5 The DWP’s consultation highlighted that existing legislation is not specifically designed to cater for these risks and vehicles such as superfunds. We agree, and we were supportive of the aims in the DWP’s consultation to legislate for an authorisation and supervision regime.
Our guidance and consultation
2.1 Superfund models could begin operations and complete transfers within the current legislative framework. Given the different nature of the risks they pose, and in light of our statutory objectives, we believe it is important that we set out clear expectations for this interim period before legislation. While much of our existing codes of practice and guidance is relevant to superfunds, we recognised that greater clarity in some areas was necessary. We published guidance in 2018, providing further clarity on our expectations prior to new legislation coming into force and before the market had begun to develop. Our guidance was largely based on the proposals set out in the DWP’s consultation.
2.2 We committed to updating our guidance as the market developed and where we believed additional clarity was needed. While new legislation specific to superfunds is expected, superfund provisions are not currently included in the recently introduced Pension Schemes Bill 2019-21. Market developments have also meant there is a risk that a significant superfund market starts to develop ahead of new legislation being in force.
2.3 There is a need for us to set clear expectations for superfunds prior to the introduction of new legislation. This will give greater clarity of our expectations to superfunds and give us a much stronger basis to intervene and take action in relation to a superfund, should this become necessary. During this interim stage, we will consider using our powers to ensure that requirements are met.
2.4 We developed some additional proposals where we strengthened our regulatory requirements, building on our existing guidance. We undertook a targeted consultation towards the end of 2019 in the following areas:
a) Capital adequacy
b) Value extraction
c) Investments
2.5 We described our proposals in relation to these three specific issues. We also explained our rationale for reaching those proposals.
2.6 We consulted with a number of interested parties about how we intended to operate our existing powers to regulate superfund activity. Given the size of this emerging market and its potential prudential regulation, we kept the number of the consultees focused on those we considered would be most affected by our proposed requirements. These were:
a) ABI
b) Christofferson Robb & Company
c) Clara Pensions
d) Legal & General
e) Pensions Insurance Corporation
f) Pensions and Lifetime Savings Association
g) Pensions Protection Fund
h) The Pension SuperFund
2.7 We have considered the views of respondents and arrived at the conclusions described in this response.
2.8 The consultation responses highlighted a number of challenges for us and some significant areas of concern were raised. As part of our consideration of how we might mitigate the issues raised while being aligned to the principles, we commissioned some specialist modelling and analysis to help us flesh out and develop our ideas.
2.9 Following a tender process, we commissioned Mercer Limited (Mercer) to perform some modelling and analysis for us. This was to inform our thinking on our capital adequacy and investment requirements. We used them as a sense check for some of our proposals as we developed them. The modelling covered the impact of adopting different funding bases and investment strategies. The results of their modelling are set out in the evidence document published alongside this consultation response. We refer to the slides throughout this document.
2.10 This modelling forms part of the evidence base behind the policy decisions contained with this response and set out in our guidance. While we have taken account of the analysis, including recommendations provided by Mercer in considering the implications and potential range of outcomes of our proposed approach, we have reached our conclusions based on all the relevant material, and the policy decisions we have made are the sole responsibility of TPR.
2.11 This response explains how we intend to use our current powers to regulate superfunds. We consulted on our expectations (eg the appropriate likelihood that a superfund can deliver good outcomes for members and over what timeframe). We are confident that our proposals are appropriate and that we are well-placed to exercise powers, if needed, to secure good outcomes.
2.12 We describe the principles and requirements that will guide our approach. We have chosen not to be prescriptive in every aspect and will be flexible in our approach to the circumstances as appropriate. We may expand our superfunds guidance in future.
2.13 Superfunds are complex and include numerous entities with different flows of obligation and benefit, although there is a broad distinction between (a) the commercial entities, and (b) the superfund pension scheme trustee. In this consultation response, we use ‘superfund’ to include these entities.
2.14 Sometimes our focus is clear in context, for example since the capital buffer is not a superfund pension trust asset on day one, it follows that many of our comments on this will centre on the commercial entities (although we will also need to understand the due diligence transferring trustees have carried out and how the trustee is comfortable with the arrangement). In other places (eg value-extraction) our comments will apply to both the trustee and the commercial entities. Broadly, we are concerned about risks to paying members’ benefits, so our comments will be relevant wherever those risks may manifest.
2.15 When it comes to some superfund models, the trustees will only gain control of the investment strategy and/or the assets in the capital buffer (or the share of the capital buffer appropriate to their section) in the event the funding level falls below certain funding level thresholds. In some instances, these thresholds will reflect the funding level triggers we are applying. There are also risks while the superfund market develops. During this interim period, as different superfund models emerge, we are mindful that some superfund models may not get market support, may not achieve critical scale, may fail or may decide to exit the market. We want the investments held by the scheme and the buffer to be appropriate (including realisable and transferable for full value) to enable 100% of members’ benefits to be protected to a high degree of certainty. It is for these reasons that we have taken the view that superfunds need their own tailored guidance.
2.16 The guidance contains how we believe the investment duties as well as the trustees’ funding and other governance requirements should apply to superfunds. The other aspects of this guidance set out how existing governance expectations have been adapted to apply to a superfund pension scheme’s circumstances.
2.17 We have a suite of current powers, which appear especially relevant to the regulation of superfunds. Some of the areas we will focus on include the following:
a) We anticipate we will be aware of proposed transactions with a superfund because of our expectation that clearance is sought. We also anticipate that ceding trustees will want to understand our position before making a transfer to a superfund.
b) We will want to understand the risks presented by a superfund model and this may involve asking detailed questions. We anticipate superfunds will want to use stochastic modelling to demonstrate that they meet our principles (eg on capital adequacy). We will want to interrogate superfund’s evidence such as modelling to check it is a suitable basis on which trustees can make decisions.
c) We will also want to understand the views of ceding and superfund (receiving) trustees. For example, we might want to be clear how they are comfortable that aspects of a superfund model do not pose unacceptable risks to members.
2.18 Our guidance clarifies what we consider to be superfunds and sets out that they are not within the ambit of our current DB funding consultation.
2.19 The government will continue to develop proposals for a new legislative regime for superfunds. The features of that regime will be informed by the responses DWP receives and our experience gained during the interim period. It is possible that new legislation will introduce greater prudence to funding requirements and standards compared to our guidance for the interim period prior.
2.20 Our expectations for this interim period prior to legislation have been incorporated into our superfunds’ guidance, along with additional clarifications.
Key principles underlying our approach
3.1 Our approach during this interim period has been developed in light of our assessment of the key risks we have identified. We have been guided by our statutory objectives in relation to protecting the benefits of members of occupational pension schemes and reducing the risk of situations arising that may lead to compensation being payable from the PPF. The proposals have also been informed by the direction of travel for new legislation set out in the DWP’s consultation
3.2 Superfunds are new and emerging business models which carry particular risks. We are conscious of the potential for new superfund models to enter the market and take a different approach to those currently in the public domain. We are seeing a number of new market models emerge, which are looking to offer consolidation benefits and ‘end-game’ solutions to pension schemes. As such, in developing our interim approach, we have been mindful that new and divergent models are emerging.
3.3 There are two main principles underlying our approach to the interim period:
a) There should be a high degree of certainty that members’ benefits will be paid in full: This leads to a low risk appetite for members’ losses during the period before legislation is in place and the need for our approach to recognise the risks flowing from the ‘for profit’ nature of business models.
b) The approach should reflect the direction of travel of future legislation: Aligning our interim approach with the direction of new legislation as far as we are able do will help reduce the risks for superfunds operating during any transitional period (eg having difficulties adapting to the new standards). This may well create risk to members and the PPF. This leads to our interim approach being influenced by the proposals set out in the DWP’s 2018 consultation.
3.4 During the interim period, we want a high degree of certainty that a superfund will be able to pay 100% of members’ benefits if, for example, the superfund:
a) does not achieve its business aims (for example by not achieving sufficient scale or market traction)
b) suffers a reduction in its funding level (for example through materialisation of asset risks)
c) is unable to meet our requirements (for example in relation to governance, systems or processes, or other matters posing risks to members’ benefits) on an ongoing basis
3.5 Our interim approach communicates a clear basis when we will intervene and take action in relation to a superfund. This will help to ensure that members’ benefits are protected and risks to the PPF are minimised, while providing a framework for emerging superfund models to operate under ahead of legislation.
3.6 As well as presenting potential risks to members’ benefits and the risk of claims being made on the PPF, a superfund failure ahead of legislation would also risk undermining the DWP’s policy intent to facilitate a market in this area. Our interim approach should help to reduce those risks and help to ensure that existing superfunds are appropriately resourced and governed in this period.
Scope of our guidance
4.1 The market for superfunds/consolidator vehicles and other new business models facilitating risk-transfer is developing rapidly. The superfund model involves some form of severance or change of the employer’s covenant and liability towards a scheme. That employer is then replaced by a special purpose vehicle (SPV) or the liability of the employer is replaced by an employer backed by a capital buffer. The replacement employer backed by a capital buffer will usually be supporting a consolidator scheme.
4.2 We think it is right, as the DWP set out in its consultation, that these vehicles adhere to strong standards across a number of areas. As the DWP also set out in its consultation, it is important that the regulatory regime is alert to and captures vehicles with similar features to superfunds.
4.3 Since we initiated our consultation in October 2019, we have seen a number of new business models starting to come to market. These include arrangements where a SPV (or equivalent) arrangement is put in place. In these models, severance from the employer may not happen at the point of the SPV being put in place, but the model envisages the SPV acting as the main or sole employer at some point in the future. This could include, for example, after an insolvency event in order to avoid the scheme going through the usual process of entering PPF assessment or through the employer extinguishing its liability to the scheme through another method. This arrangement may involve the introduction of investor capital alongside a contribution or premium paid by the employer or scheme.
4.4 The risks these models pose to members and the PPF are similar to superfunds in that there is no further recourse to a sponsor. We therefore believe it is important that we adopt a consistent approach to these new models to ensure that the level of security provided to members’ benefits is consistently applied. Additionally, it is important to control the development of this emerging market in a manageable and robust way. Therefore, our guidance is relevant to a business model where a SPV is put in place so it could result in employer replacement at some point in the future. The requirements apply at the point when the SPV or other applicable vehicle becomes the sponsor to the scheme, and so our arrangements need to be set up so our guidance can be complied with when this happens. Before entering such a transaction, trustees and employers will need to understand the ramifications of this requirement.
4.5 However, in contrast to some superfunds and consolidator vehicles, some business models do not offer services beyond those required for capital adequacy purposes and do not plan to consolidate schemes (specifically, where there is no change to the scheme in terms of trustees or the level of the control the trustees have over governance and other management and administration aspects). In these scenarios, it may not be necessary to apply all the additional requirements that our guidance sets out for superfunds beyond capital adequacy.
Capital requirements (scheme and capital buffer funding)
We said
5.1 We believe it is appropriate that a tailored solution in relation to the funding requirements for superfunds is developed to reflect the different nature of the risks they pose.
5.2 As highlighted above, a key aspect of superfunds is that the covenant of a ceding scheme is replaced by additional capital, or that additional capital is being provided to a scheme where the employer covenant has already been lost. This means the scheme is wholly reliant on the level of funding, the level of provided capital, the investment strategy and the requirements around risk management. It is, therefore, important that there are clear requirements around these aspects set in the context of an overall risk appetite.
5.3 We proposed an overall level of capital for superfunds at which full member benefits would be protected to a high degree and the potential for calls being made on the PPF would be limited pending the introduction of legislation. This was consistent with our statutory objectives.
5.4 The DWP’s 2018 consultation proposed a 99% likelihood of members’ benefits being paid or secured in full. We aimed to set our approach so that the chance of members’ benefits being paid or secured in full was equally high. As such, we proposed to focus on the ability of a superfund to transfer its benefit liabilities to an insurer if problems arose and to support that transfer with a risk-based capital buffer to give a defined level of confidence that the superfund would have sufficient assets to do so.
5.5 Our proposal was that the scheme assets plus capital buffer should at least: equal an estimate of the buy-out price and produce at a 1-in-100 Value at Risk (VaR) over a one-year period. The buy-out estimate should be calculated by the scheme actuary, and the VaR should be calculated by the scheme actuary or investment adviser and evidenced to us. This would then be measured and reported quarterly and topped up to the required level as needed.
You said
5.6 While almost all respondents expressed agreement and support for our guiding principles, responses to our proposals were mixed and somewhat split between those with a focus on the occupational pension industry and those with a focus on the insurance industry – although there were some areas of overlap and agreement.
5.7 Those with concerns highlighted three main areas:
a) Firstly, that it set too high a bar for the capital required and the level of capital required would mean that the proposition would not be investible, meaning capital providers will not enter the market. Some argued that it would set a capital requirement above that of insurers under the Solvency II Directive, especially when combined with some of the other aspects of our proposals. This would mean that a superfund market would not develop, and the policy intent set out by the DWP (that superfunds provide an alternative solution to schemes that cannot access the buy-out market) would not be met.
b) Secondly, aligning the proposals to a buy-out figure and setting any intervention triggers on that basis would lead to superfunds investing in the same limited pool of assets as insurance companies, which could affect their availability and price (and so increase the price of risk transfer to those insurance companies looking to purchase those assets). This is because superfunds would look to hedge their liabilities on a buy-out basis and would need to invest in a way that mirrors insurance companies, which is a key driver of the buy-out price.
c) Thirdly, accurate assessment of the buy-out price on an ongoing basis would be very difficult in practice, as insurance companies have no obligation to disclose their pricing bases, there is not a transparent market and no objective measure for assessing the price on a consistent basis. Insurance companies would also be unlikely to spend time and resource accurately quoting for schemes which were not planning on transacting with them. The buy-out price would also be driven by a number of factors that would be out of the control of superfunds. It was highlighted that this would make it very difficult for superfunds to accurately measure and assess their funding position and to hedge and manage their risks on this basis. It was also highlighted that this would create similar issues for our ability to effectively supervise superfunds’ compliance with this principle.
5.8 Contrary to this, some respondents were supportive of our approach but wanted greater clarity over the details. Some also asked us to set out a ‘standard’ methodology for our requirements. They highlighted the need for greater clarity in relation to the financial models and metrics used to ensure the framework was consistent.
5.9 Some were supportive of the buy-out pricing link but highlighted the difficulties in measuring this, and the associated implication that this would require hedging of the price. One respondent stated that estimates of the buy-out price could be provided by any suitable firm or actuary engaged by the superfund.
5.10 One organisation asked how risks which had already been hedged by a superfund would be appropriately recognised where the valuation basis was based on an assessment of the buy-out price.
5.11 Several respondents were supportive of our proposal that superfunds should be able to secure a buy-out with an insurer in the event of a superfund failure or market exit during the interim period. However, one respondent suggested that this requirement was less appropriate once a superfund achieved a level of scale. Another respondent suggested that the requirement to wind-up and transfer to an insurer in the event of a commercial failure of a superfund was not necessary as benefits could continue to be paid in full under a suitable low risk funding strategy.
5.12 Other comments included:
a) Irrespective of the approach, the link to the ceding scheme’s employer should be retained for all superfund transfers.
b) No consolidation business should operate ahead of legislation as the impact on members of the longer-term legislation would be unclear.
c) Our proposals should not limit the market only to superfunds that aim to act as a ‘bridge to buy-out’ only. Specifically, those not aiming to move schemes to buy-out in the short-to-medium-term should be included and catered for.
d) Our proposals should open the door to weaker schemes so the potential benefits from consolidation could reach a broader market.
e) There was a risk that the high bar that was set for the interim period to mitigate the additional risks arising during a period of emerging superfund models and market practices would become the baseline for the permanent regime.
f) The requirement to provide top-up capital would discourage superfund investors who would not want to agree to open-ended commitments.
g) Clarity was needed on the operation of triggers for TPR intervention through use of existing powers, including engagement and enforcement.
Our response
5.13 We believe the respondents raise valid concerns regarding our initial proposals. In particular the problems in relation to accurately assessing the buy-out price - given the lack of a transparent market - and the implications such an approach would have for superfunds and our supervisory approach. We also recognised the potential for our proposals to set too high a bar and that our aims could still be achieved through alternative approaches.
5.14 We are persuaded that the concerns raised could undermine the policy objective that the DWP set out in its 2018 consultation, as well as our ability to effectively supervise. We have considered ways we might adjust our approach, while still continuing to meet our aims.
5.15 To manage the risk of our capital requirement being tied to the buy-out price (which is not transparent and driven by factors out of a superfund’s control), we considered alternatives based on a specified level of funding that could be easily managed and reported. To achieve this, our approach should be based on clear and observable market parameters and be measured against an appropriate low-risk basis. Consistent with common practice for DB pension schemes, we believe an approach with discount rates set with a suitable spread relative to UK government bonds would be appropriate and meet our aims.
5.16 In judging the level of security our regime provides for members, it is important to focus on the overall level of funding and capital required. However, in order for superfunds to have a suitable exit strategy, if risks materialise and the funding level deteriorates, we believe a prudent technical provision approach would be for this funding basis to be set so a scheme could run on with a high probability of securing or paying members’ benefits in full. We refer to this as our ‘low-risk basis’, and it should represent the minimum standard for the technical provisions of the superfund scheme.
5.17 In addition to these prudent technical provisions, we continue to believe that a superfund should hold additional capital and that these capital requirements should be risk-based. The level of additional capital required to provide security for the superfund scheme should not be based on a ‘fixed addition’ as a capital buffer but rather one that is sensitive to the risks to the scheme reflecting the underlying asset and liability risks of the particular superfund. This is important to ensure there is adequate capital for the risks being run and that this is set at a consistent level we deem appropriate and consistent with our aims of a low risk to members’ benefits and the PPF. This will also help ensure that risk-taking has appropriate mitigation in place and gives credit to superfunds that have reduced the risks they are running.
5.18 In addition, in setting the capital requirements, it is vital that there are in place appropriate trigger points for action to be taken (or not taken) if funding falls below certain prescribed limits. This is a common feature of supervisory regimes.
5.19 In line with the DWP’s consultation proposals, we are adopting three triggers:
a) New transactions: A section in a segregated superfund that wishes to transfer in a new scheme must be funded to at least the level set under our capital requirements (ie technical provisions plus capital buffer). A non-segregated superfund must be funded to at least this level to accept a transfer. This is to ensure that the security of members of ceding schemes is not weakened by being integrated into a superfund that is funded below our capital requirement. It also acts as an incentive for superfunds that wish to transfer in additional schemes to maintain their capital position.
b) Low risk trigger: Set at 100% of our technical provisions basis. If overall funding in the scheme plus the capital buffer falls to this level, in the absence of an additional capital injection, all capital buffer funds flow to the scheme and come under the control of the scheme trustees so it can run on with a suitable investment strategy or be transferred out. This is designed to ensure that there is a point at which the trustees gain full control of the assets and they could continue to run the scheme on with a high chance of member benefits being paid in full.
c) Wind-up trigger: Set at a level of 105% of the value of PPF liabilities measured in line with Section 179 Pensions Act 2004. This provides protection against a claim for compensation being made on the PPF. This is in line with the DWP’s consultation proposals.
5.20 In practice, through our supervision (including ongoing monitoring and reporting) we would expect to be in regular contact with superfunds and be informed in advance if a superfund is approaching any of these triggers. We expect appropriate plans to be put in place to ensure the trigger is implemented effectively.
Setting the strength of these requirements
5.21 There are two key elements that define overall strength of our approach: the level of the capital buffer which reflects the confidence level and the time period over which that confidence is derived; and, the level at which we have set our low risk trigger and minimum level of technical provisions.
Capital buffer
5.22 The capital buffer replaces the ceding employer’s covenant and, while it is not an asset of the pension scheme, it forms part of the longer-term security of the scheme, which can be called upon when needed. The capital buffer serves as a key mechanism to support the scheme against adverse experience in the absence of a sponsoring employer. Therefore, it is important that it is set at a level that provides for a high level of confidence that is sufficient to achieve this. Specifically, that it can provide for a high level of confidence that the scheme will be funded at or above the technical provisions and low risk trigger.
5.23 As part of the analysis to inform our approach, see slides 8 to 16 in the evidence document (PDF, 2.4MB, 69 pages), we have stochastically modelled differing investment strategies and capital buffer requirements. This has informed our decision on what level of confidence would achieve a high level of protection and be a reasonable and appropriate level for superfunds, which would be consistent with our policy aims.
5.24 At this stage it is not clear when there will be superfund legislation. We have set out our approach mindful of this uncertainty, and we believe that looking over a period of five years is a reasonable approach. This period is also short enough that modelled results are more robust and less sensitive to model drift.
5.25 We, therefore, expect the confidence level for the capital buffer to be measured at the end of the five-year period following transfer. This is the case for any superfund coming to market or existing superfunds looking to transfer in new schemes to a section of the superfund.
5.26 As can be seen from the analysis, for the investment strategies we have modelled, a confidence level of 99% that the scheme is funded at or above technical provisions in five years would mean a capital buffer of between 15-25% above technical provisions, before any allowance is made for longevity risk. In practice, the individual investment and risk management strategies of the superfund will determine this level. We believe a confidence level of 99% represents a sufficiently high level of security, while remaining a practical and manageable level for superfunds to operate under.
5.27 Some respondents suggested we should use the Solvency II standard formula approach to set the capital requirement but adjusted to the desired confidence level. However, one respondent expressed concern at the potential for scope creep of the insurance regulations to occupational pension schemes.
5.28 There are existing methodologies and calibrations developed for Solvency II. However, we are not convinced that such an approach is wholly necessary or appropriate at this stage. The time horizon we are looking to adopt is not in line with that of Solvency II, as we are focusing on the risks and level of security during the interim period before legislation. We are also targeting a different confidence level. Significant analysis would be needed to recalibrate the Solvency II standard formula and ensure it was fit for our purposes across an adjusted confidence level and time horizon.
5.29 We want to incentivise superfunds to carry out appropriate due diligence and assessment of their risks, and it is important that we have clear oversight of the models and approaches used by superfunds through our assessment process. We believe that our aims during this interim period can be best achieved by placing the onus on superfunds to justify their models and approaches to our satisfaction as to their appropriateness and compliance with our principles. This needs to be coupled with controls around any changes to those models or approaches and a regulatory approach that allows use to independently, objectively and consistently assess to what extent we can accept the outcome of the superfund’s modelling.
5.30 We expect the modelling for the scheme and capital buffer to include all the relevant and material risks that the superfund is exposed to. This must include market risks (including inflation) consistent with the confidence level and timeframe set out above.
5.31 A reserve for longevity must also be included. We believe that the longevity risk is most likely to manifest itself whereby the rates of future improvements in longevity are greater than those assumed. While our aim for a longevity reserve using this approach is to be consistent with the confidence level we are applying for market risks, we are aware that there is no universally agreed approach to how this would apply.
5.32 To help determine the appropriate reserve, we have reviewed the impact of applying an immediate and permanent shift to the mortality base tables and also through applying a stochastic longevity assumption to an asset liability stochastic model. These alternative approaches result in a reserve equivalent to broadly 3% of technical provisions, for a scheme made up of 50:50 pensioner and deferred members.
5.33 Based on the above, we consider that the capital buffer should include a longevity reserve at least equivalent to the increase in liabilities caused by changing the assumption for the long-term rate of improvement in the mortality assumptions to 2.0% p.a. While this is expected to result in a 3% reserve for a 50:50 scheme where technical provisions were set with a 1.5% improvement rate, in practice the reserve will be dependent on the membership profile and mortality assumptions adopted. We believe this is a pragmatic and prudent approach for superfunds during this interim period.
5.34 Any risk mitigations in place, such as hedging in relation to interest and inflation rates or hedging of mortality risks, should be accounted for in the modelling.
5.35 We expect the superfund to demonstrate that the approach they have taken satisfies our requirements. Our guidance sets this out in more detail.
Low risk trigger and minimum standards for technical provisions
5.36 A superfund needs to be well enough funded so that, if our low-risk trigger applies, the scheme can still run on with a high chance of paying benefits in full. In deciding what funding level may be appropriate to set our low-risk trigger at, we have reviewed modelling outcomes and analysed the success probabilities for different funding levels and investment strategies on differing funding bases.
5.37 As can be seen in slides 17 to 22 and 28 to 33 of the evidence document (PDF, 2.4MB, 69 pages), we modelled a scheme running on when fully funded using Gilts+0.5% p.a. as a discount rate, under a number of different investment strategies. We have projected the probability of the scheme paying full benefits, the average proportion of benefits met across all scenarios and the probability of the scheme reaching full funding on a proxy buy-out basis under all modelled scenarios.
5.38 The results of this modelling show that at this level of funding, there is a high probability of members receiving full benefits under three of the four strategies modelled. However, Strategy C1 is expected to produce the most positive outcome for members – with a c.93% chance that members’ benefits would be paid in full and an average proportion of benefits met across all scenarios of 98.8%.
5.39 At this point, the assets would be under the full control of the trustees and there would be no further recourse to additional capital.
5.40 We modelled differing funding bases for our technical provisions, see slide 27 of the evidence document (PDF, 2.4MB, 69 pages), using investment Strategy C. Strengthening the technical provisions discount rate to Gilts+0.25% would improve the probability of members receiving full benefits by c3% to 96%. However, a stronger technical provisions discount rate of Gilts+0.25% would be more onerous for superfunds. It would require the technical provisions to increase by around 4-5% more and would risk undermining facilitating a market in this area without sufficient member risk reduction in receiving their full benefits.
5.41 Our modelling illustrates that an alternative weaker technical provisions requirement of Gilts+0.75% would reduce the probability of members receiving full benefits by c7% to 86%. This would make our proposal more affordable for superfunds. However, this is a material reduction in the probability that members receive full benefits and would in downside scenarios significantly reduce the average proportion of benefits that members would receive. This could undermine our aim that there is a low risk of members not receiving their benefits in full.
5.42 Any consideration of overall security for members’ benefits needs to take account of the capital buffer as well. However, the results for Strategy C provide us with comfort that a funding basis for technical provisions set with a discount rate set at Gilts+0.5% combined with an appropriate investment strategy would represent a prudent and low risk approach. In our guidance, we also set out our minimum expectations in relation to other key assumptions that form part of these technical provisions.
5.43 We are aware that changes in market conditions may mean that, at particular points in the future, were we to repeat this analysis, then outcomes consistent to those that we have found from the modelling - and that have formed a basis for our decision on the appropriate funding basis – may require an adjustment (up or down) to the Gilts+0.5% discount rate.
5.44 We believe it is appropriate to retain a fixed Gilts+0.5% basis for this interim period. We believe this to be a reasonable basis which acts a suitable long-term proxy in which to achieve our aims. Although we will monitor the appropriateness of this basis, we do not plan to continually adjust this basis as it represents a long-term view. We expect it to be included in a superfund’s legal documentation in respect of our low-risk trigger.
5.45 However, we recognise the material implications for superfunds in the situation where they are approaching being funded only to this level. Specifically, that when our low-risk trigger applies, all funds from that section would flow to the scheme and be under the full control of the trustees.
5.46 Market conditions at that point could be materially different to the point in time at which the scheme was taken on by the superfund - given that the level of funding required above technical provisions is set with a very high level of probability that funding would not fall to this level. Therefore, we believe it is appropriate that we retain some discretion to ensure that the Gilts+0.5% remains the appropriate benchmark for a scheme in that specific scenario, given the prevailing market conditions.
5.47 Our decision will be based on whether for that scheme, a change in the Gilts+0.5% basis would be anticipated to achieve a similar level of prudence, security and expected future outcomes that was initially targeted and if it were to run on consistent with our aims as set out above. This will be the driving principle behind our assessment and decision.
5.48 The onus of demonstrating this to our satisfaction will lie with the superfund. We will expect that the superfund will include within its governing documentation that the low-risk trigger will apply at the Gilts+0.5% level, unless otherwise agreed by the trustees and TPR. As part of any arguments made to make an adjustment at that point in time, we would expect to see all modelling of the investment strategy and funding basis that the superfund believes justifies any adjustments. We will expect any such submission to be fully supported by independent advice and we will assess that submission in line with our approach to assessment of the superfund’s modelling for the capital buffer. We will need to be convinced that the modelling is appropriate and robust and leads to the outcomes that are consistent with the basis under which we have set the Gilts+0.5% discount rate.
Taking the low-risk trigger and capital buffer together
5.49 The most important aspect to focus on is the overall level of security provided by the assets covering our minimum standards for the technical provisions and capital buffer when taken together (the “total stack”).
5.50 As part of our analysis, we have carried out similar modelling as we undertook in assessing the different levels for the low-risk trigger – namely running a superfund on over the long term when they were funded at the total stack. Our overarching objective is that when all these elements are taken together, that the probability of members’ benefits being paid in full is high.
5.51 As can be seen in slides 23 to 26 of the evidence document (PDF, 2.4MB, 69 pages), our modelling illustrates for the investment strategies shown that the chance of the overall stack paying full member benefits is over 99%.
5.52 We believe our approach strikes the right balance between security and affordability, while meeting the underlying principles behind our approach. We also believe that the high level of certainty under our approach is consistent with the aims of the DWP’s consultation.
5.53 We have also modelled the probability of a superfund falling below our wind-up trigger – set at 105% of S179 – in order to assess the level of risk to the PPF. The ratio of full scheme benefits to the PPF level of compensation is a very scheme-specific consideration. However, we have modelled different starting funding ratios of our technical provisions basis relative to S179 - with and without including the capital buffer - and tested the probability of hitting our wind-up trigger. The results of this can be seen in slides 42 to 45 of the evidence document (PDF, 2.4MB, 69 pages) and show that when the capital buffer is included, unless the ratio of S179 benefits is close to full benefits, the probability of hitting our PPF trigger within the next 20 years is very low across all the investment strategies we modelled.
5.54 Under strategy C with a 15% buffer, the cumulative probability is less than 1% over 20 years unless the ratio of full benefits to S179 is less than 120%. The analysis also shows that we could expect that the PPF liability value will have fallen by 40-50% over that time reducing the quantum of risk to the PPF.
5.55 While the risks to the PPF are mitigated through the provision of the wind-up trigger, the probability of wind-up will be a key consideration for superfunds, as it results in the loss of all investor capital. Superfunds are unlikely to provide additional capital in situations where the risk of wind-up of a scheme or section is material. Only in exceptional circumstances, such as where the PPF wind up trigger would be enacted before our low risk trigger or the scheme experienced significantly heavier mortality than implied by s179, would we consider agreeing a PPF wind up trigger below 105%. This would need to be agreed with TPR in consultation with the PPF.
Conclusions
5.56 The results of the modelling and analysis we have performed have provided us with significant comfort that the risk to members receiving less than full benefits and the risk of a claim on the PPF is very low.
5.57 This approach also helps to mitigate the key issues highlighted by respondents that concerned us regarding our original proposals. Specifically, it is based on a clear and measurable market, provides for a robust and risk-based approach to supervision with clear triggers, and aligns with the policy intent.
Value extraction
We said
6.1 Superfunds represent a new type of business model, managed on a ‘for-profit’ basis seeking to extract ‘surplus value’ at future points in time in return for the input of investors’ capital. For ‘bridge to buy-out’ models, this will be at the point when liabilities are bought out. ‘Run-off’ models would seek to extract ‘surplus value’ once the combined funding levels exceed certain thresholds.
6.2 The ‘management for profit’ motive creates a number of risks and incentives for superfunds that are not present in traditional DB schemes – including the risk that short-term investment gains are prioritised over longer term outcomes. The DWP’s consultation highlighted a number of options here for controlling these risks. These included:
a) Creation of an additional ‘profit trigger’ set at a higher level than the authorisation basis for transfers in before profit could be taken.
b) Prevention of any profit being taken until all member benefits are bought out in full.
6.3 We proposed for our interim regime that no value should be extracted from the superfund until the earlier of:
a) legislation being brought in defining the rules for profit extraction
b) benefits being secured in full with an insurance company
6.4 We considered this would help ensure that the incentives of those running superfunds would be better aligned with the interests of the members. Together with our capital requirements and other expectations, this would limit the potential for excessive risk being taken in the hope of increasing profit during this interim period.
You said
6.5 The consultation responses were generally supportive of the principle and understood that a cautious approach was required during this interim period before a long-term regulatory regime is developed.
6.6 Some responses suggested we modify our proposals. These suggestions included that a ‘long stop’ to the proposals should be included on the basis that it was not reasonable to expect investors’ capital to be tied up indefinitely should they not buy-out members’ benefits. More specifically, they indicated that it should be made clear that our proposals were temporary, and they would have a defined end point.
6.7 One respondent argued for surplus value to be allowed to be extracted at the point at which the superfund was funded above the initial capital requirements, and that this could be coupled with a ‘comply or explain’ basis or the use of vesting periods before capital is released.
6.8 Almost all respondents wanted more clarity on the definition of ‘value’ and how it would apply to ongoing administration and other management fees that would be part of normal business operating activities. Several respondents were keen to ensure that transaction fees and the withdrawal of legitimate business fees and expenses, once appropriately benchmarked and disclosed, would not be limited by this principle.
6.9 Some respondents highlighted that value extraction could be achieved not only through extracting surplus value (ie above a particular funding level/ capital adequacy level threshold) but also through excessive charges in some areas, and through ‘creative’ ways of extracting fees and charges against a wide range of services. Some felt we needed to be comfortable that we could oversee the range of costs the superfund would be exposed to, that we would have access to sufficient information to do so, and that we would have the ability to address any concerns.
6.10 It was also suggested that de facto profit-taking could take place within non-segregated superfunds, which could use any surplus capital built up from earlier transfers into the superfund to serve as the investor capital for future transfers in. While this would not be profit extraction per se, it was highlighted that this would be equivalent to a release of profit built up from earlier transactions, which was then reinvested to provide capital to fund for future transactions.
6.11 Some also called for ongoing (quarterly) disclosure of any forms of indirect return and more generally for public disclosure of returns.
Our response
Surplus value
6.12 There was broad consensus around the merits of restricting the extraction of surplus value from the superfund during this interim period. As superfunds are designed to operate over the longer term and we expect their investors to have an understanding of the longer-term nature of pension liabilities, it is reasonable to expect some caution during this interim period.
6.13 Additionally, while we do not want to create barriers to different models entering the market, the DWP’s position on whether and when to allow profit-taking is not yet settled. If we were to set triggers to enable profit extraction before benefits were bought out, we risk being at odds with future legislation. Superfunds operating during this time extracting profits on an ongoing basis may find they are not well-placed to meet the eventual requirements.
6.14 While we understand the concerns raised by some stakeholders, we believe it appropriate to continue with limiting the extraction of surplus value. However, we understand investors will want greater certainty. We are therefore adjusting our approach to one where no surplus value should be extracted for three years. However, in the absence of legislation at that point, we will commit to reviewing our policy position on this issue and publishing it within this three-year period. Our review will be informed by our experience of the operation of the regime during this period and by developments in government policy and legislation regarding superfunds.
6.15 We note the suggestion that co-mingled superfunds could look to use surplus capital built up from previous transactions to fund the investor capital part of future transactions. We believe this would go against our policy intention in this area. Therefore, we will also require that all transfers into a superfund should be able to meet the capital adequacy test on a ‘standalone’ basis. We would expect fresh capital to be provided at a level which, together with the ceding scheme assets and any ceding employer contributions, would satisfy our capital requirements if that scheme is to be considered in isolation.
Fees and charges
6.16 We agree with respondents that value extraction through fees and charges is also a key area, and that our policies regarding value extraction should include our position on these. Limiting profit extraction from surplus assets built up within the superfund could encourage some superfunds to increase fees and charges in other areas. It could also encourage more ‘creative’ ways of extracting profit to be developed.
6.17 As with all DB schemes, we would expect any fees and charges to be justified by the nature of the services being provided and the value for members. Ultimately, we want to safeguard against ill-defined and inappropriate levels of fees, charges and expenses being levied while not limiting legitimate fees, expenses and charges. We would also expect trustees to ensure they are getting value for the services they commission.
6.18 The range of fees and charges that DB pension schemes pay varies significantly depending on a range of issues, including the schemes’ governance arrangements and the complexity of the investment and risk management arrangements. The most significant component costs can generally be easily benchmarked. We will be mindful that our additional expectations of superfunds during this interim period may lead them to incur some fees and charges which could be higher than for normal DB schemes of equivalent size and complexity. We will consider this within the context of the superfunds’ expectation to achieve economies of scale and delivery, aiming to provide more efficient and effective management of the schemes to deliver those economies.
6.19 Our approach to the level of fees and charges paid by superfunds will be of a principled nature rather than through fixed limits or charge caps. We will expect superfunds to disclose their charging structures and trustees should be in a position to demonstrate they are getting value for money for the services they receive, for example through benchmarking and by completing competitive open market tender exercises for services.
6.20 We will ask superfunds to submit full details of their policies to us in relation to fees, costs and charges. We will expect this submission to include details of the governance processes around the appointment and payment of service providers and how fees will be set and allocated between the superfund and the capital buffer assets. We will expect superfunds to justify to us the extent to which expected fees, costs and charges have been capitalised and included in the technical provisions. In addition, we will expect the submission to include details of benchmark analysis, completed in line with market levels of fees and charges for pension schemes of similar scale and complexity.
6.21 We expect fees, costs and charges to be monitored on an ongoing basis and the superfund trustees to regularly demonstrate to us that they are getting value for money for the services they commission. Any material changes to policies in this area would need to be reported to us. We would expect to see in place legally binding agreements that give us comfort that superfunds will adhere to this approach.
6.22 We believe this approach, coupled with full levels of disclosure to the trustees of the superfund and ceding schemes, is appropriate to our aim of helping to ensure that fees, costs and charges are transparent and reasonable.
Investment principles
Introduction
7.1 Alongside our proposals regarding capital adequacy and value extraction, clear requirements are required during this interim period regarding how the superfund invests. As well as the general points, this is also to reflect the following:
a) It is essential that the assets in the capital buffer are invested in a manner appropriate for the scheme because the capital buffer may be called upon (and its assets may need to transfer to the scheme) to provide support for the scheme.
b) On transferring to a superfund, the link with the ceding employer’s covenant is severed. Any non-cash investments that are offered by the ceding employer to enable the transfer to the superfund to complete must be of appropriate quality and value, as they provide support for members’ benefits.
7.2 We proposed some investment principles to help ensure:
- that all the assets held were invested appropriately for the liabilities
- concentration risks were limited
- the assets held were ‘transferable’
- investments accepted as part of any scheme transfer were consistent with investments that the trustees would ordinarily hold
7.3 Our final investment principles are not exhaustive. However, we believe they are the minimum that should be applied, and they are reasonable to apply during the interim period until a long-term regulatory framework is developed. We also recognise that, in the longer term, a more refined approach may be appropriate.
7.4 We would expect to see in place agreements where superfunds can demonstrate how they comply with our principles. The scheme will remain subject to the applicable legal requirements that apply to its trustees.
Consultation Principle 1: Capital buffer assets should be invested as though the investment regulations 2005 apply.
We said
7.5 Under current legislation, the Occupational Pension Scheme (Investment) Regulations 2005 (the investment regulations) apply to the assets held in the scheme but not the assets in the capital buffer.
7.6 The assets held in the capital buffer replace the ceding employers’ covenant and are relied upon by the trustees to provide support to the scheme until members’ benefits are secured with an insurer or paid in full. Given the intrinsic link between the scheme and the capital buffer, and the use (and potential use) of the capital buffer, we believe it is appropriate for the capital buffer to be invested appropriately for a pension scheme and in accordance with the investment regulations.
7.7 We therefore believe it is important that the protections that apply under the investment regulations to assets should also be put in place for the assets in the capital buffer. This was also proposed by the DWP in its consultation paper.
You said
7.8 Respondents acknowledged the need for the investments within the capital buffer to be subject to some controls and oversight from us. They were generally supportive of a principles-based approach as opposed to a prescriptive approach.
7.9 However, there was some variation between the responses received. Some agreed that the capital buffer assets should be invested as though the investment regulations applied, to ensure they could be readily transferred to the pension scheme. Another suggested that, to provide members with greater certainty, a more prescriptive approach involving a requirement for close cashflow matching of the schemes’ assets with the liabilities should be taken for the interim period. Another highlighted that, while the application of the principles was potentially appealing, the regulations as drafted applied to trustees and did not easily read across to the capital buffer.
7.10 One area of concern highlighted within the provisions of the investment regulations as a concern was the requirement to prepare a Statement of Investment Principles (SIP). It was acknowledged that the capital buffer assets needed to be invested in line with a clearly articulated investment strategy. However, it was not clear how the detailed requirements of the SIP would apply, the potential for conflicts of duty to arise between the requirement for the capital buffer assets to be invested in the best interests of members (and the requirements on the directors of the company that hold the capital buffer funds), and the provisions that arise in relation to borrowing, guarantees and employer-related investments.
Our response
7.11 Our approach to controlling the risks that the capital buffer assets present was developed from the existing investment regulations. We aimed to retain the flexibility within these regulations by adopting a high-level and principles-based approach, which mitigated some specific risks. They would also help to ensure the assets in the capital buffer fund were invested in a way that was focused on the key aim of providing security for members’ benefits.
7.12 We do not believe a detailed approach, including involving specific requirements for close cashflow matching of liabilities, is necessary. This is mainly due to our investment requirements being designed to work alongside and complement our capital adequacy expectations, which provide for a risk sensitive assessment of the assets held in the capital buffer.
7.13 We acknowledge that the investment regulations do not directly fit with the various ways that capital buffer funds may be structured within different superfunds. However, we believe that applying the high-level principles in the investment regulations will help to ensure risks in the capital buffer investment arrangements are mitigated during this interim period.
7.14 Having considered the consultation responses relating to this principle, including the concerns raised, we have decided to include this investment principle in our final expectations. However, we recognise the concerns raised about the direct transposition of the investment regulations to the capital buffer. As such, we have reflected those concerns by modifying the principle to reflect our expectation that the investments held in the capital buffer should be invested in line with the principles and provisions underlying the investment regulations. Our revised principle is: the investments held in the capital buffer should be invested in line with the principles and provisions underlying The Occupational Pension Schemes (Investment) Regulations 2005 (investment regulations).
7.15 Where the specific provisions did not fit specifically due to the structure of the capital buffer, we will expect superfunds to ‘comply or explain’ why they do not.
Consultation Principle 2: Assets held to cover the scheme’s buy-out liabilities must be invested in a manner appropriate to the nature and duration of the expected future retirement benefits payable under the scheme. The superfund should adhere to the principle underlying regulation 4 (4) of the Occupational Pension Schemes (Investment) Regulations 2005 but where liabilities are based on the buy-out level.
We said
7.16 Our proposals for capital adequacy focused on the superfund holding enough assets between the capital buffer and the scheme to cover the price of buy-out. This would enable members’ benefits to be paid if the scheme had to transfer to an insurer. To this extent, and because of the nature of the scheme’s reliance on the capital buffer, we proposed that the capital buffer should also comply with this regulation (in addition to the scheme).
7.17 This would mean the capital buffer assets would be invested in a manner similar to the scheme’s. Under this regulation, the scheme’s assets (and the capital buffer assets) are invested in a manner appropriate to the nature and duration of the expected future retirement benefits payable. We also proposed that this important protection under the investment regulations to how a scheme’s assets are invested should also be put in place for the assets in the capital buffer.
You said
7.18 This proposal received a very limited response. The main concern expressed was that the proposal would not be compatible with superfund models that are not pursuing buy-out and whose long-term investment strategy is not designed to match short-term fluctuations in buy-out pricing. However, it was acknowledged by one respondent that as a short-term measure, until the superfund achieved sufficient scale, this proposal could be appropriate.
Our response
7.19 Our original capital adequacy proposals focused on the price of buy-out and the ability of the scheme to be able to transfer to an insurer, if the superfund failed to meet the requirements. As set out in section 5, our proposals have moved to focus on a funding objective based on a high probability of being 100% funded after five years and a high chance of members’ benefits being paid in full if the scheme had to run on without access to any additional capital.
7.20 In view of our revised capital adequacy proposals, we believe there are sufficient protections in place to help ensure the quality of the investments in the capital buffer is appropriate. We are, therefore, removing this principle.
Consultation Principle 3: Recognisable maximum allocations. Apart from investments in gilts, a recognisable maximum allocation to any one issuer (or group of undertakings) within each asset class of 10% (as proportion of asset class) where entity holds AAA credit rating (or equivalent) and 3% if A or below credit rating (or equivalent).
We said
7.21 Regulation 4(7) of the investment regulations requires a scheme’s assets to be properly diversified and avoid excessive reliance on any asset, issuer or group of undertakings. This is to avoid accumulations of risk in the portfolio.
7.22 We were concerned that members’ benefits could be put at risk if concentrated investment positions held in assets or issuers failed. To control this risk, we proposed introducing certain limits in relation to the maximum allocation to any one issuer (or group of undertakings) within each asset class that we would recognise as part of our capital adequacy assessment. This approach provided a simpler solution than developing detailed capital charges for concentration risk and avoided additional complexity in the calculation of additional capital requirements.
7.23 As superfunds achieve scale, the diversification of asset risks may happen more naturally, but it is important that there are controls in place to manage these risks from the outset.
7.24 We suggested some limits which are similar to those used in Investment Management Agreements. If investments held in assets or funds breached the individual limits, we proposed to only recognise those allocations up to the level of the limits when we assessed how the superfund would meet our capital adequacy requirements.
You said
7.25 Although there was active support for a well-diversified investment approach being adopted, several concerns were raised. These included the potential for the provisions to discourage investment in some pooled investments. There were concerns that, in some risk management products, the process was overly prescriptive and we should be able to rely on superfunds establishing their own risk and concentration limits. There was also some concern that ceding schemes’ asset holdings could be misaligned with the requirements. Therefore, some portfolio adjustments could need to be made in the short term by the trustees, which might involve additional costs to either the ceding scheme or superfund and loss of value in the event of a ‘forced’ sale of assets.
7.26 Some respondents wanted us to exclude investment in certain assets (for example, those related to the superfund), and proposed introducing capital charges based on concentration risk.
Our response
7.27 We believe our proposal provides a simpler solution than developing detailed capital charges for concentration risk. It avoids the need for complex calculations to establish the requirement for additional capital where concentrated positions are taken.
7.28 We acknowledge respondents’ comments about the exclusion of certain employer-related investments and take the view that the requirement for the capital buffer to adhere to the investment regulations (as detailed above) will address some of the concerns.
7.29 It was not our intention to dis-incentivise investment in certain risk transfer products (for example, buy-in policies) and other appropriate third-party pooled investments. Therefore, we will make it clear that certain exclusions or concessions (for example, in relation to buy-in policies), will apply.
7.30 We have retained the principle of the limits and refined the detail in line with additional advice received from Mercer. We obtained this advice because we were concerned that if superfunds held overly concentrated positions, in the absence of an additional capital charge for concentration risk, those positions could have an impact on modelled and actual outcomes for members. The advice from Mercer indicated the following:
a) Within their modelling, they had assumed that superfunds invested in a diversified portfolio within each asset class.
b) In practice, this would require concentration risk to be appropriately managed within each asset class and at an overall level in respect of: security level, issuer level, industry sector.
c) The overall principles should be as follows:
- The superfund does not hold an excessive proportion of any individual security or an issuer’s debt and/or security.
- The superfund does not hold an excessive proportion of its assets in any one security, issuer or industry.
d) A prescriptive approach to limiting concentration risk by industry or country was not necessary but these concentrations should be monitored through the supervision regime.
7.31 In the light of the consultation responses and the advice we received, we have made the following changes:
a) We have introduced a new principle (“Additional principle 2”) to address concerns over the time available to transition ceding scheme assets so they align with our investment principles following the transfer in.
b) We will make it clear in our guidance that certain investments (such as buy-in policies) will be excluded.
c) We will amend the limits as outlined in our revised principle below.
d) In addition, we will require trustees to provide full disclosure of all investment service agreements and the risk controls (concentration limits, bandwidths, etc) within them as part of the initial authorisation process and as part of ongoing supervision.
7.32 Our revised principle is:
Maximum allocations will be limited as follows:
a) 5% limit of the total issuance of a security and 2.5% limit of an issuer’s total debt and equity issuance (including subsidiaries and associated companies, excluding government bonds rated AA or higher) across all sections. Measured at an overall level across all mandates with look-through.
b) 1% to a single security, 2% to a single issuer (excluding government bonds rated AA or higher) within each section. Measured for each section across all mandates with look-through.
7.33 We believe these revised proposals provide some additional flexibility to superfunds, avoid unintended consequences and help to ensure risks are mitigated during the interim period until a long-term regulatory framework is developed.
Consultation Principle 4: Limits on illiquid assets
7.34 Recognisable illiquid assets held need to fall within the fair value hierarchy set out in Financial Reporting Standards (FRS) and the allocation within each category needs to be within the following ranges:
- Level 1: 60% - 100%
- Level 2: 0% - 40%
- Level 3: 0% - 10%
We said
7.35 The capital buffer and scheme assets need to be sufficiently liquid so that, for example, any scheme transfer to an insurer would not be unduly impacted.
7.36 To manage ‘mark to market’/‘mark to model’ valuation risk and illiquidity risk, we proposed that the level of illiquid assets that could be held over both the capital buffer and scheme should be limited by controlling the maximum allocation that trustees could hold within different liquidity categories. We proposed that assets should be categorised under the three-level, fair value measurement hierarchy set out under Financial Reporting Standard 102 (FRS 102) and for each level we set out ranges within which asset allocations would be recognised. The FRS is a well know and used standard, so assets should be easily categorised by superfunds in this way.
7.37 If investments held exceeded the upper ranges set for the individual levels, we indicated that we would only recognise those investments up to the level of the upper range limit which had been breached.
You said
7.38 This proposal attracted a range of views. One respondent suggested that to have the best chance of being able to transfer assets to an insurer, Level 3 assets should be prohibited, and that more flexibility should be allowed within the ranges for Level 1 and Level 2. Other responses suggested that insurers and occupational pension schemes typically held higher allocations within the Level 3 asset category than the upper limit proposed and that well-managed portfolios of illiquid assets could be beneficial for a buy-out process, as insurers would not need to allow for the costs of transition.
7.39 Some acknowledged that the fair value hierarchy was a useful starting point and that reporting against the hierarchy on a periodic basis would be useful. However, there was some concern that the allocation ranges proposed would unhelpfully skew portfolios towards more liquid assets.
Our response
7.40 Our original capital adequacy proposals focused on the price of buy-out and the ability of the scheme to be able to transfer to an insurer if the superfund failed to meet our requirements. As set out in section 5, this has changed. Therefore, the requirement to be able to transfer to an insurer in the short term is now of lower significance.
7.41 We had sought to reduce the level of investment in illiquid assets as the realisable asset valuation for some illiquid assets can be hard to assess and may involve significant ‘mark to market’ and ‘mark to model’ risk. In addition, in times of market stress or uncertainty (when some developing superfunds may struggle more), the realisable asset valuations can be even more uncertain. This can lead to an overestimation of value and overstatement of the level of security supporting members’ benefits. We also believe it can be advantageous to trustees to have access to a sufficient proportion of liquid assets if a scheme needed to transfer to an insurer following the failure or withdrawal of a superfund from the market.
7.42 We have retained the principle of the limits and refined the detail in line with additional information we obtained from the DB Pension Scheme Leverage and Liquidity Survey (PDF, 1,178KB , 53 pages) that we completed in Q4 2019 and in light of advice from Mercer. We obtained this advice because we wanted to make sure that the revised limits we were considering (which allowed for higher allocations to the more illiquid, Level 2 and Level 3 asset categories) would be reasonable to apply. They advised that some additional flexibility could be allowed in relation to the Level 1 and Level 2 limits that we proposed. In light of the advice we received, and the consultation responses submitted, we have revised the range limits we proposed to provide greater flexibility to allow for the capital buffer and scheme to invest more in Level 2 and Level 3 category assets (and correspondingly less in Level 1 assets).
7.43 We acknowledge that the fair value hierarchy is a simple measure and recognise that, in the longer term, a more refined approach coupled with additional supervisory tools could mean a more bespoke approach is appropriate. However, for the reasons set out above, we believe these proposals are reasonable to adopt during the interim period until a long-term legislative framework is developed.
7.44 Our revised principle is:
Recognisable illiquid assets held need to fall within the fair value hierarchy set out in FRS 102 and the allocation within each category need to be within the following ranges:
- Level 1: 40% - 100%
- Level 2: 0% - 50%
- Level 3: 0% - 25%
Consultation Principle 5: Recognisable in-house pooled funds could only be used for quoted assets that could be freely traded in an active market and that could be transferred in specie, without penalty.
We said
7.45 We recognise that in-house pooled funds have the potential to impose prohibitive fees on exit or may give rise to assets being realised for depressed prices on exit. This is something that could reduce the value of the realisable assets available to secure benefits for members following, for example, the failure or withdrawal of a superfund from the market. This was a principle we were applying to both the capital buffer and the scheme.
You said
7.46 The respondents generally agreed with this principle but suggested the need for caution in its application to ensure that certain funds (for example, such as life-wrapped and pooled liability driven investment funds) were not captured because it would not be possible to transfer life-wrapped pooled funds to a bulk annuity provider. While there was acknowledgement of the potential risks with inappropriate pooled funds, it was also suggested that there should be some flexibility to consider individual circumstances on a case-by-case basis.
Our response
7.47 We recognise that, while our thinking broadly aligns with the respondents and some controls on the potential use of in-house pooled funds are necessary, respondents are keen to maintain some flexibility. We accept this. To this end, we have added an ‘unless otherwise agreed’ caveat to the beginning of this principle. This provides the desired flexibility by allowing us the latitude to consider matters on a case-by-case basis.
7.48 In addition, as part of ongoing supervision, we will require the pension scheme trustees and capital buffer to evidence that any in-house pooled funds are consistent with investments that the trustees would otherwise ordinarily have chosen. This includes providing independent advice as to their suitability and the details of any exit restrictions or penalties.
Consultation Principle 6: Restrictions on ceding employer investments.
7.49 We do not propose setting any additional specific limits in this area. However, we want to emphasise the overarching principle that any superfund decision to accept such investments would need to be evidenced as being consistent with investments that the trustees of the superfund pension scheme would otherwise ordinarily have made.
7.50 Assets transferred by the ceding employer need to be consistent with that overarching principle. Any ongoing payment streams should only be considered where they would otherwise be justifiable as an investment made by the trustees in the ceding scheme’s employer.
We said
7.51 Although it may offer a ‘clean break’, a transfer to a superfund may also involve an ongoing payment stream from the employer of the ceding scheme to the superfund. This leaves the superfund (and so ceding scheme) with the default risk of that employer. We emphasised that any decision to accept an ongoing payment stream from a ceding employer should only be considered where that payment stream (which would effectively represent an investment) would have ordinarily been made by the trustees.
7.52 These ceding employers may also wish to use property or SPV structured investments as part of the transaction price they pay to the capital buffer. This also creates issues and risks regarding the valuation of such assets and the ability to realise them on transfer to an insurer. We did not seek to set any additional specific limits in respect of ceding employer investments. We emphasised the overarching principle that any decision to accept such investments in either the capital buffer or the scheme would need to be evidenced as being consistent with an investment that the trustees would have ordinarily chosen to have made.
You said
7.53 Respondents agreed with the principle.
Our response
7.54 We acknowledge respondents’ agreement with our proposal. We do not intend to make any changes to the principle.
7.55 We will require trustees to evidence that any ceding employer investments accepted are consistent with investments that the trustees would otherwise ordinarily have chosen and to provide independent advice as to their suitability. We will also require the trustees to provide independent advice on the fair value of any such investments, which allows fully for the credit risk embedded.
Additional principles
7.56 Following revision of our interim capital adequacy proposals, consideration of feedback received and revision of our investment principle proposals, we are extending the range of investment principles we propose to provide greater clarity and to set out our expectations.
Additional Principle 1: Requirement to manage scheme assets in line with asset scale.
7.57 In our October consultation, when setting out our high-level objectives behind our proposed investment principles, we indicated that “all of the assets held should be invested appropriately for the liabilities”. This objective was most directly linked with proposed principles 1 and 2. At the time, we were mindful that although superfunds aspire to achieve scale (their models are based on delivering economies of scale), some superfund models may fail or may not get sufficient market support to gain sufficient scale to be ‘viable’ in the longer term and may need to exit the market.
7.58 We have now withdrawn our proposed principle 2. However, we want to avoid superfunds from investing (or committing to invest) in funds not appropriate for their scale ahead of actual scale being achieved. If the scheme had to transfer to another provider or wanted to run on, following failure of the superfund or a decision to withdraw from the market, this could have significant consequences. It could affect the level and suitability of investments entered into and the level of members’ benefits that could be secured with another provider or provided if the scheme ran on. Therefore, we are adopting an aspect of proposed principle 2 and adapting it to deal with the scale issue.
Additional Principle 2: Assets ceding to the scheme or capital buffer must be subject to a transition plan.
7.59 A number of responses highlighted the need for some flexibility in the time allowed to transition the asset portfolios of the schemes ceding the superfund scheme. These assets may not immediately fit with either our principles or the trustees’ investment principles.
7.60 We want to avoid transferred assets having to be sold within a short period of time, as ‘forced sales’ could impair value and incur additional costs. We consider it is appropriate to include some flexibility around assets ceding into the superfund. We have, therefore, listened to these responses and are introducing a new principle.
7.61 Superfunds that take on portfolios of assets from ceding schemes will need to produce a clear transition plan for re-alignment of those assets with our principles and their investment strategy. We also want to avoid superfunds running elevated levels of risk for extended periods of time while ceding scheme assets are being transitioned and believe it is important not to leave this timeframe open ended. Therefore, unless otherwise agreed, we would expect the transition to be fully completed within 12 months of the transfer. If the scheme intends to run an elevated risk for an extended period of time, which is not supported by a credible transition plan, we will expect that risk to be supported by additional capital.
7.62 The same approach will apply to the capital buffer.
Additional Principle 3: Requirement to demonstrate resilience to certain risks and development of a detailed Integrated Risk Management plan.
7.63 Integrated Risk Management (‘IRM’) is a central component in our messaging to DB trustees. It is important that trustees monitor and measure the key risks they are exposed to. Our existing guidance for superfunds sets out our expectation that they should prepare and maintain an appropriately detailed IRM framework at an individual scheme level and an aggregate scheme level. Our existing guidance also sets out that the IRM framework should address the full range of risks the scheme and superfund is exposed to and should include consideration of the impacts of scale and the pace of change of scale.
7.64 This principle is not purely related to investment issues but has a wider application. Therefore, superfunds should look across the range of risks they are exposed to. However, we are including it here for completeness and want to bring your attention to two specific investment areas we expect to see covered. Our guidance covers this principle in more detail.
Liquidity management
7.65 In recent years, as pension schemes have been maturing and matching investment strategies have evolved, we have been highlighting the need for improved management of scheme liquidity (and the risks of not doing do). We believe that the active management of liquidity risk by superfunds is essential and we will expect superfunds to develop a liquidity risk management plan as part of their overall investment risk controls. We would expect the superfund’s ‘liquidity risk management plan’ to set out its policies and processes in relation to liquidity risk and their proposals for managing and monitoring those risks. We would also expect this plan to examine liquidity needs in benign circumstances and under market stress, and for it to allow fully for ‘mark-to-market’/’mark to model’ impacts on realisable investments.
Climate risk
7.66 Climate risk is an issue of increasing importance for pension scheme investments. It is also an issue where a number of regulatory requirements have recently been introduced. We are aware that some superfunds may invest heavily in illiquid and long-term assets. We have not included any formal recapitalisation requirements within our interim capital adequacy regime for superfunds. However, to limit the risk of the security of members’ benefits being undermined, we want to ensure that the risks relating to climate change are adequately allowed for in the development of superfund investment strategies. We will expect superfunds to develop a ‘climate risk management plan’, setting out its policies and processes in relation to the assessment of climate risk and its proposals for managing and monitoring those risks.