Written submission to the Work and Pensions Select Committee inquiry on collective defined contribution pension schemes by Professor David Blake, Director of The Pensions Institute at Cass Business School, part of City, University of London.
Please note that this response is based primarily on findings of Professor Blake’s Independent Review of Retirement Income and related consultation, in which further relevant information can be found.[1]
Should it be useful to the Committee, Professor Blake would be pleased to contribute further to the inquiry through the provision of oral evidence – expanding on this initial written submission.
BACKGROUND INFORMATION
Dr David Blake is Professor of Pension Economics at Cass Business School, City University, Director of the Pensions Institute and Chairman of Square Mile Consultants, a training and research consultancy.
He is also: Co-Founder with JPMorgan and Towers Watson of the LifeMetrics Indices; Senior Research Associate, Financial Markets Group, London School of Economics; Senior Consultant, UBS Pensions Research Centre, London School of Economics; and Research Associate, Centre for Risk & Insurance Studies, University of Nottingham Business School. Formerly Director of the Securities Industry Programme at City University Business School; Research Fellow at both the London Business School and the London School of Economics; and Professor of Financial Economics at Birkbeck College, University of London.
David Blake was a student at the London School of Economics in the 1970s and early 1980s, gaining his PhD on UK pension fund investment behaviour in 1986. In 1996, he established the Pensions Institute (www.pensions-institute.org) as the only academic centre in the UK exclusively devoted to pensions studies.
David chaired the Independent Review of Retirement Income which published its Report and response to the consultation in March 2016.
SUBMISSION
Would CDC deliver tangible benefits to savers compared with other models?
- Supporters of collective defined contribution (CDC) pension schemes claim that they can produce higher and more stable incomes than individual defined contribution (IDC) pension schemes. The main benefits claimed for CDC are risk sharing and lower operating costs. It is claimed that as a result of these benefits, CDC pensions can be 30% or more higher than in pure DC schemes.
- The evidence that I have examined indicates that CDC schemes could generate smoother pensions across different cohorts of members than IDC schemes. This evidence comes from both theoretical models of intergenerational risk sharing using an overlapping generations framework and stochastic simulation models using CDC designs that are typical of those in use in the Netherlands, such as career average revalued pensions with conditional indexation.
- However, theoretical models suggest that CDC schemes are only likely to be sustainable in the long run if (a) everyone joins (i.e., participation is mandatory) and (b) everyone remains in the scheme for life. However, these two conditions potentially break down in the UK context.
- A number of criticisms have been made of CDC, in particular, that the higher and/or less volatile potential pension comes at the expense of some severe restrictions on choice flexibility and that the schemes are complex to manage:
a) Transfers between schemes are, of course, possible, but this is in theory much easier with IDC schemes – where every member has their own account – than with CDC schemes – where members will simply know their target pension and could be subject to a market value adjustment (MVA) if they transfer. Transfers are much more complicated in practice than in theory, at least in the UK. In a CDC context, it would be much more efficient if the ‘scheme followed the member’ when the member changed jobs and hence transfers could be avoided. This, in turn, requires that there are only a few large CDC schemes in existence, all fully exploiting scale economies. A worker joins one when they first start work and stays with that scheme for life. This is only likely to be feasible if the CDC schemes are organised, not on a company basis, but on an industry-wide or national basis.
b) CDC schemes appear to work as intended only if people stay in for life and draw an income from the scheme, rather than take the accumulated pension fund out at retirement. The 2014 Budget introduced greater ‘pension freedoms’ from April 2015 which would allow DC savers to take their pension fund in cash from age 55 (subject to income tax after the tax-free cash is withdrawn). By keeping their assets in the scheme, some would claim that members would be ‘losing’ their pension freedoms. If sufficient savers exercised these new freedoms, it would make CDC schemes unfeasible. The CDC schemes in the Netherlands, for example, do not permit this flexibility. CDC schemes have little flexibility over the age of retirement. The CDC schemes in the Netherlands have a fixed retirement age and the investment strategy in the accumulation phase is designed with this retirement age in mind.
c) Members of a CDC scheme have no identifiable pension pot, so the valuation of each member’s claim in a CDC scheme is as challenging as it is in a DB scheme. Members who transfer out of a CDC scheme when they change jobs might experience a reduced transfer value via a MVA if the scheme has an implicit deficit. The large CDC schemes in the Netherlands are industry-wide schemes and most people when they change employers move to different employers in the same industry and so stay in the same pension scheme. This suggests that CDC schemes should ideally be established on an industry-wide basis or that we move away from workplace pension schemes sponsored by an employer to a small number of very large nation-wide multi-employer pension schemes with employees choosing to join one of these when they first enter the labour market and then stay with it for the remainder of their career.
d) If the risk sharing in a CDC scheme is not fair between generations, it could turn into a Ponzi (or pyramid investment) scheme, with older members taking out more than their fair share at the expense of younger members. Ponzi schemes come to a sudden stop when new investors refuse to join. There is, of course, the opposite problem that the first generation in the scheme receive less than their fair share due to the trustees being overly cautious. Trustees therefore need to be aware of – and put mechanisms in place – to avoid both possibilities. CDC schemes also face the demographic risk that the working population is too small to pay the pensions of a large and growing retired population. Supporters of CDC schemes need to answer the question why younger workers would join a scheme that was in deficit (which would happen if older workers were regularly drawing a pension based on the targeted performance of the investment fund which was higher than the realised performance). CDC would be especially problematic in declining industries with fewer and fewer new members joining each year.
e) Related to this is the criticism that CDC schemes cannot work without an ‘estate’ or initial reserve that can be used for smoothing returns. Supporters of CDC schemes might argue that, with good governance, it is not necessary to have an estate. Alternatively, it might be possible to start a CDC scheme without an estate, but to require an estate to be built up by the first group of members. In other words, this group takes out less than is justified by the fund’s investment performance in order to build a smoothing fund. This would help to establish the scheme’s credibility.
f) The risk-sharing rules lack transparency. This is especially true in CDC schemes that operate on a similar basis to with-profit schemes. While it is claimed that there is greater transparency in CDC schemes that operate on a unit-linked basis, nevertheless, risk sharing usually involves actuarial discretion. It could be argued that discretion is the enemy of transparency. Some, however, have argued that some degree of opacity is necessary for such schemes to work at all.
g) While under current Government proposals, CDC schemes fix employer contributions, future Governments or EU directives might change this. This could happen if the target pension was turned into a guaranteed pension which resulted in a deficit being created in the scheme which the employer was required to fill. Supporters of CDC schemes argue that the way to protect against this is to have a clause in the scheme rules which automatically triggers a switch in the CDC scheme to an IDC scheme should this happen.
h) Some CDC schemes in Denmark have introduced a zero-rate minimum guarantee (i.e., the saver gets at least the accumulated value of their savings back) or a guaranteed minimum pay-out in nominal terms (equivalent to the purchase of a deferred annuity). This begins to introduce a defined benefit element to a defined contribution pension scheme (i.e., makes the scheme a hybrid scheme). From a regulatory point of view, hybrid schemes are very complicated to run in the UK, especially if such guarantees require levies to be paid to the Pension Protection Fund (PPF).
j) A question mark has been raised over whether the proposed 0.75% charge cap would apply to CDC schemes.
How would a continental-style collective approach work alongside individual freedom and choice?
- The best time to have introduced CDC was in 2009 when the Government of the day first looked at the possibility of introducing it, but turned it down. This might have helped stem the flow of private-sector DB schemes switching straight to IDC. That flow has since become a flood and the end of private-sector DB is now unstoppable in the UK. So CDC had a past.
- It might also have a future if employees use their new freedoms unwisely and deplete their pension pots to an extent that they cannot afford to retire: recall that pension schemes in the private sector were initially set up by enlightened employers to manage the exit of their employees from the workforce when they were no longer capable of productive work.
- But I do not believe that CDC has a present: the new pension freedoms are completely incompatible with CDC’s requirement that members stay for life and draw a pension in retirement, rather than use the pension pot as a bank account.
- However, recognising the benefits of risk pooling, I believe that collective individual defined contribution (CIDC) schemes might be the only form of collective scheme that is feasible in the short term following the introduction of ‘freedom and choice’. Because they maintain individual accounts, they are better able to deal with sudden cash withdrawals than CDC schemes, yet are still able to exploit economies of scale to the full which lowers costs, e.g., through automatic enrolment and the pooling of investment and longevity risks.
- There are key features of CIDC schemes that are specific to each individual member and which make the scheme easy to understand:
a) The CIDC scheme maintains individual accounts for all members in the accumulation phase, so it is easy to value each individual’s pension pot
b) The contribution rate is set to be actuarially fair to each member, implying that there is a direct relationship between the contributions that an individual pays into the scheme and the pension they eventually receive. This contrasts with CDC schemes in which contributions are averaged on a collective basis to meet a target average salary pension
c) Each individual has their own de-risking investment strategy in the lead up to retirement.
- For this reason, I recommend that the Government looks at the feasibility of establishing collective individual defined contribution schemes – for both the accumulation and decumulation phases. Such schemes would be compatible not only with the defined ambition agenda, they would also be compatible with the new pension flexibilities following the 2014 Budget, while, at the same time, exploiting economies of scale to the full and allowing a high degree of risk pooling.
- If CDC (or one of its ‘DC-heavy’ alternatives) is introduced in the UK, then it is useful to take into account lessons from other countries. Of particular relevance are the Netherlands and New Brunswick schemes. CDC works in the Netherlands because the Dutch are willing to cooperate to make the system work. Large-scale industry-wide schemes are built on employer and union agreements. Employers and unions meet as ‘social partners’ in works councils in a spirit of ‘social solidarity’. This type of collaboration is far less common in the UK, given its history of labour relations.
- Scale and cost are important issues to deal with. The Dutch CDC schemes were not set up from scratch but were converted from DB schemes in which the benefits were not guaranteed, but instead were conditional on the performance of scheme investments and could be cut if necessary. By contrast, the accrued benefits in UK DB schemes are guaranteed and cannot be changed in solvent companies, unless members agree. This means that UK DB schemes cannot be converted into CDC schemes. Companies with DB schemes would have to set up new CDC schemes which would be a costly exercise. Further, the companies would have to be large ones with a large number of potential members in order to generate scale.
- Another important lesson from the Netherlands is that the CDC schemes are run by not-for profit organisations that are largely trusted by all generations of scheme members. This trust is very important when risks are shared across generations. It is likely that a for-profit organisation would rapidly lose trust if it were awarding dividend payments to its shareholders at the same time as cutting pensions in payment which will inevitably happen at some stage in a CDC scheme – even if, as supporters of CDC schemes claim, this will happen rarely. This, in turn, would seem to suggest that a CDC scheme should operate in a trust-based framework where the trustees are professionally qualified and independent of the sponsor.
- The New Brunswick SRPPs[2] provides an interesting case study on how a collective scheme might be introduced:
a) Pension plans must be subject to robust risk management, and be checked annually, including stress tests, to ensure that the plan complies with that task.
b) A pension plan must provide benefit security. This means: risk management targets are focused on delivering a high degree of pension security for members and retirees; and the plan must be governed by an independent trustee(s) who can force employers and employees to increase (or decrease) contributions when appropriate, subject to realistic and manageable limits.
c) A pension plan should be able to demonstrate that it will be sustainable over the long term.
d) A pension plan must be affordable, which means that contributions must be stable and affordable for both employer and employees.
e) The plan must be equitably designed – no single age cohort should unduly subsidise another, and no one should be able to ‘game’ the system.
f) The plan must be transparent. The pension goals and risks must be clearly stated upfront; who shares the risks and rewards and by how much must be clear and preestablished.
g) Benefit changes as a result of conversion will apply only in the future; everyone keeps the amount that has already been credited.
h) There should be no sudden shocks to members and retirees’ retirement plans.
i) All groups of employees should be treated consistently, including part-time employees.
j) At inception, the actuarial assumptions must be closely related to market benchmarks, such as International Accounting Standards.
Does this risk creating extra complexity and confusion? Would savers understand and trust the income ‘ambition’ offered by CDC?
- The sad reality is that vast numbers of people in this country do not understand their pension scheme or how it works. Many people do not even know they are members of a scheme, even amongst those who have recently been auto-enrolled. So with all the recent changes in pensions legislation, the introduction of CDC is bound to create extra complexity and confusion (except amongst those who pay no attention to their pension and will be blissfully ignorant). We would also have another term ‘ambition’ with very little obvious meaning to contend with.
- We are just approaching the end of the first phase of auto-enrolment with very modest contributions of 2%. We are a long way from seeing whether raising contribution rates to 8% will be effective in a soft compulsion framework. Even this will be inadequate and contributions need to be raised to around 15% in order to have a decent pension in retirement. We have yet to see the full effects of ‘freedom and choice’, but some early signs are there in the form of scamming and distortions to the housing market as pension lump sums are flooded into the buy-to-let market.
Could seriously underfunded DB pension schemes be resolved by changing their pension contract to CDC, along Dutch lines?
- I would have thought this was most unlikely. Dutch CDC schemes have high funding ratios and strict mechanisms such as conditional indexation to help reduce deficits that emerge. Companies in the UK want to offload their DB pension liabilities completely, not transfer them into a new type of strange arrangement that no one really understands, especially if there is a chance that it leads to some new unanticipated liability down the road. Further, the deficit could not be transferred to the new CDC scheme, given the requirement to have a high funding ratio. So what happens to the deficit? Will the DB members agree to take a benefit cut (equivalent to a market value adjustment) when they move over to the CDC scheme? Another problem is that most large companies have already closed down the DB schemes and CDC needs scale to work effectively.
How would this be regulated and how would the loss of DB pension promises to scheme members be addressed?
- There is a new type of arrangement called a regulated apportionment arrangement (RAA) permitted by the Pensions Regulator.[3] This allows schemes to restructure their benefits to avoid the insolvency of the sponsoring company. The restructured benefits have to be better than PPF benefits. This requires the agreement of the members. Such agreement has frequently been achieved, especially in a smaller company, since the older members are prepared to accept a cut in benefits, to keep the company (which often employs younger family members) afloat. A similar approach, overseen by the Pensions Regulator, could be used with DB to CDC transfers.
How would CDCs be regulated?
- The regulation of CDC schemes would be no different in principle from the regulation of DB schemes. There would have to be regular scheme valuations (probably annually rather than triennially), but the adjustment, in the event of a sufficiently low funding ratio, would be to benefits rather than to the sponsor contribution as in the case of DB.
Is there appetite among employers and the UK pension industry to deliver CDC?
- I see little current appetite from employers. It will take many years for the full effects of auto-enrolment and ‘freedom and choice’ to work through and for someone to say that ‘something must be done’ about the old age poverty that results and to propose a better risk sharing solution, such as CDC or CIDC. It would, however, be possible for large companies with IDC schemes to switch to CDC at reasonably low cost should they wish to do so. However, it would be even cheaper for such companies to convert their IDC schemes to CIDC schemes. The pension industry is keener, especially the actuarial consultancy industry, since it needs to find a revenue stream that replaces that coming from the dying DB industry.
- However, in due course, with appropriate governance and management, it might be possible to introduce some form of collective scheme in the UK. A large single employer might find it attractive to do this. But, it is more likely that a multi-employer trust-based scheme (like NEST) would find it easier to do this if there were sufficient appetite amongst scheme members.
- Supporters of CDC schemes in the UK include the Pensions and Lifetime Savings Association (PLSA) and the Trades Union Congress (TUC). So in principle, there is support for some form of collective DC scheme amongst some of the UK’s ‘social partners’. Further, it is claimed that there is a group of employers who might be interested in setting up CDC schemes and that is those employers planning to close down their defined benefits scheme and wish to offer their employees something more reliable than individual DC schemes (even if they cannot change accrued benefits).
- Some employers might be attracted to CDC in preference to IDC if they could convert their defined benefit (DB) pension schemes into CDC schemes. This would allow DB promises to be converted into non-guaranteed targets in the CDC scheme. This would require retroactive changes to accrued DB benefit entitlements. While this is permissible in Holland, for example, the Government has so far refused to allow this in the UK. The overarching Government objective is to make pure DC stronger rather than make pure DB weaker.
- CDC ought to be attractive to public sector employers with DB schemes which have precisely the same problems as their private sector equivalents.
Would CDC funds have a clearer view towards investing for the long term?
- The real benefit of long term investors such as pension funds comes from their ability to earn a liquidity premium from investing in illiquid assets that also benefit the economy by helping to increase the productive capital stock. The classic asset fitting this role is infrastructure. Canadian and Australian pension funds have led the way, but UK funds are catching up at last, after a long history of short-termist investment behaviour. CDC schemes should in principle follow this investment approach too. Certainly Dutch CDC schemes do, but they have the advantage that members are in for life and draw a pension rather than a lump sum. ‘Freedom and choice’ means that schemes have to hold more liquid lower yielding assets to meet cash withdrawals.
-------------------------------------------------------------------------------------------------------------
Professor David Blake
Director, Pensions Institute
Cass Business School, City, University of London
2