Rt Hon Frank Field MP
Work and Pensions Committee
c/o Clerk of the Works and Pensions Committee
House of Commons
7 Millbank, London, SW1P 3JA
Dear Frank and colleagues
Michael Johnson’s Commentary on Collective Defined Contribution (CDC)
Michel Johnson chose to write to you, after the evidence phase for your enquiry into CDC schemes had all but completed, with some strong views. He describes his views as dispassionate, although they read as anything but. As one of the “small cabal of pensions consultants” (charged by Michael with seeking to replace our diminishing DB income streams) I am writing to offer you some observations on Michael’s commentary, and to correct some of the errors in his paper.
In truth there is little new that he has advanced, and the vast majority of his comments were covered in my earlier analysis of the full set of responses to your committee. I am writing in a personal capacity, but for good order, I should declare that in a professional capacity Aon (my employer) is an adviser to Royal Mail.
I know and have worked with Michael on pension policy issues over a number of years and, in general, respect his views. However, I struggle to understand the motivations for his attack on CDC schemes. He sets out his own vision of how a CDC style scheme could be delivered, using a combination of with-profits arrangements and his own version of drawdown, with annuity purchase at an advanced age (which he calls auto-protection). His advocacy of with-profits is surprising given the number of pages he devotes to pointing out the failures of with-profits funds.
His use of with-profits would lead to what is referred to as CIDC plans (where the “I” stands for individual) – a route which he later dismisses: “unsurprisingly, CIDC has gained little traction.” His preferred use of auto-protection uses a form of collective risk sharing to address longevity (in the context of drawdown of a DC member’s pot) – through the use of an insured longevity pool. For reasons discussed elsewhere (eg in the Aon submission to your enquiry) I believe that insurance is an inefficient and expensive mechanism to deal with the longevity risk and that CDC solutions offer a superior approach to dealing with longevity risk in the context of the freedom and choice agenda.
These CDC solutions go well beyond the potential for schemes like that proposed by Royal Mail, and would be of relevance to the whole generation of future DC savers who will need to convert their savings pots into sustainable retirement incomes. Michael is doing a grave disservice to these individuals by seeking to deny them what would become a core component of their future retirement planning.
One central error in Michael’s commentary is his description of what the proposed Royal Mail plan will be. In his letter he accuses people of “muddled reporting”: “Elsewhere, DB language has been adopted in a CDC context, such as the totally misleading “DB-style benefits”. Indeed Royal Mail itself has comingled DB and CDC terminology, one manager referring to combining a CDC scheme with a DB-style lump sum at retirement.”
The pensions agreement reached between Royal Mail and the Communication Workers Union has two components:
• a CDC pension, subject to the necessary legislative and regulatory changes, together with
• a DB lump sum scheme, which provides a minimum lump sum at normal retirement age.
The lump sum is a defined benefit, underwritten by Royal Mail. Like all defined benefit obligations this will be reported in the accounts of Royal Mail using current accounting standards (eg FRS17 or IAS19). The proposed pension solution in its totality is therefore a hybrid arrangement incorporating a DB lump sum and a CDC pension income. The Royal Mail manager’s description is spot on.
Michael’s letter sets out 10 key risks which he sees for either Royal Mail or the legislators (the emphasis seems to vary between the risks). My observations on each of these are set out below.
Risk #1: a leap into the unknown. Royal Mail has committed to deliver a pension scheme that is untried and untested in the UK.
Michael asks why did Royal Mail need to volunteer itself to be the UK’s CDC guinea pig? I will leave Royal Mail to answer for themselves, but the agreement between them and CWU seems to me an admirable compromise that dealt with the inability of Royal Mail to shoulder conventional DB pension risk, yet a shared desire between the parties to deliver pension based outcomes for their members, rather than individual DC solutions. The CDC plan proposed delivers the prospect of a higher, stable income for life when compared to DC, and obviates members from taking complex pension decisions in relation to the accumulation and decumulation of their DC savings, and. New it may be, but the intention is surely to be applauded and supported, rather than derided?
Risk #2: lack of client ownership of the CDC agenda: is Royal Mail, as client, the driver or the passenger in what may be an undeliverable adventure? The case for CDC may not be being driven primarily by its performance merits. This is not a sound basis for the formation of innovative pensions policy.
Again, it is not for me – nor indeed Michael – to speak on behalf of Royal Mail. But his suggestions that “successful product innovation is customer-driven” and the case for CDC “is not being driven primarily by its performance merits” deserve comment. I suspect if we asked customers of pensions what they would want from product innovation, many of them would end up describing a conventional DB plan – something that delivers a stable, predictable income for life that does not require them to make complex personal decisions, and lasts as long as they live. That is exactly what a CDC style plan is seeking to do – so far from Michael’s assertions about self–serving consultants, CDC should really be seen as the industry looking to innovate to meet its customers’ demands.
And in relation to the performance merits of CDC – as part of the Defined Ambition agenda from Sir Steve Webb, they had clearly been sufficiently accepted on their own merits that the government of the time was prepared to commit to legislation to permit them. That legislation may have been squeezed out by subsequent pension events, but the fact that it still remains on the statute books suggests its basic merits are still accepted.
Risk #3: legal risk. Given CDC schemes’ lack of legal definition, there remains the potential (depending upon any scheme’s final design) for some balance sheet exposure, perhaps arising through “legal creep”.
Michael’s letter serves to muddy the waters about possible legal, regulatory, taxation and accounting ambiguities. Why would that be surprising, for a new type of pensions venture? The development of the CDC legislation and associated regulation will clarify these matters, but there are some aspects that we can state with certainty even now. The risk of “legal creep” – or what we refer to as re-categorisation risk – towards DB obligations is self-evidently the most fundamental of risks associated with CDC plans. Does Michael believe that a company would work towards a CDC solution, only to see it potentially re-defined as a DB obligation? There will need to be a clear categorisation framework for CDC schemes – in my view the emphasis would be on the DC part of the CDC, so there is no question of re-categorisation risk, but this is ultimately a decision for DWP. Michael states that “CDC schemes reside somewhere on the DC to DB risk allocation spectrum”. This is incorrect. They are quite firmly DC schemes in terms of their regulatory, taxation and governance structures.
Interestingly in this section of his note, Michael states that a CDC scheme could be classified as a DB scheme “because of its potential to develop a target funding deficit”. But he answers his own question just a few paragraphs later when he states that “CDC schemes cannot become underfunded.”
Risk #4: a regulatory No Man’s Land. There is no regulatory framework in place for CDC schemes, heightening the risk of stakeholder misunderstandings.
This is exactly what is currently being worked through to decide how CDC schemes fit into the regulatory landscape. Whichever route, Part 2 of Pension Schemes Act 2015 clearly sets out the issues which will need to be addressed in this process. The key features to my mind are:
• Setting targets for benefits, and effective communication of those targets to members
• Regular actuarial valuations to assess progress against the targets
• A robust benefit adjustment process, to ensure that targets remain consistent with the assets available
• A consistent policy for investment, financing and benefit adjustments
• Public disclosure of these processes in action
• Freedom for members to leave the scheme with a fair share of assets
This regulatory framework will minimise the risk of stakeholder mis-understandings that Michael is concerned with.
Risk #5: reputational risk. Employers would invariably be attaching their reputations to the wellbeing of any CDC scheme that they may sponsor.
This risk is stated in the context of benefit cuts being unacceptable, and having to be ‘bailed out’ by the employer. Yet Royal Mail and the CWU have entered into this benefit design on the explicit understanding that the potential for benefit cuts is a central part of any CDC scheme. The financial structure – the combination of investment policy, benefit adjustment policy and valuation method – will seek to minimise this, but the potential will still exist, and is fully acknowledged by the key stakeholders.
Interestingly in this section Michael repeats familiar accusations that CDC schemes need “a continuous flow of new entrants to replace the deceased” and that a shrinking workforce will pose “complex risk management challenges, summarised as a tontine system.” These accusations have – incorrectly – been levelled against CDC schemes repeatedly, without acknowledgement of the fact that this can be addressed in the design of a CDC scheme. He could, for example, have read the detailed research in the Aon paper on CDC schemes called Stability and Fairness, and accessible here http://www.aon.com/unitedkingdom/attachments/retirement-investment/defined-contribution/Collective-DC-Stability-and-Fairness.pdf
Our analysis comprehensively dispels the myths that CDC scheme require a constant flow of new entrants, or cannot “work” in the context of declining memberships. We expect that both of these analyses will be replicated and vindicated in the context of the proposed Royal Mail scheme, and its context.
Risk #6: irreversible inter-generational injustice through excessive liquidation of communal assets (“over-distribution”) to pay today’s pensioners at the expense of current and future employees.
Michael lists the factors and assumptions to be taken into account in determining any changes to target benefits under a CDC scheme – demographic, financial assumptions and asset returns. But he then jumps to the conclusion that the trustees will inevitably if faced with a tough decision look to change the assumptions “until a more accommodating answer is achieved”. It is precisely to protect against this type of behaviour that we have recommended that all assumptions should be on a publically accessible website. Michael himself would be able to see if scheme trustees are are adequately facing up to tough decisions – how much more transparent could a system be? There is no such public pressure in defined benefit schemes for example.
Michael throws into this section of the paper the words Madoff and Ponzi scheme – with no justification or evidence to support his assertions that CDC scheme share these characteristics. Our modelling (as referenced above) is extensive modelling of CDC schemes in both the start up and run off phases and is concrete evidence – rather than assertion – that CDC schemes can still work in these circumstances. There is no inherent or inbuilt feature of a CDC scheme that makes it clear that there will be what he calls over-distribution: “excessive liquidation of communal assets to pay pensioners at the expenses of current and future employees”. The adjustments to target benefits can be set to be “generationally blind” treating active and retired members in an equitable fashion. And all carried out in a fashion and with public disclosure so that anyone can see if the stated policies are being applied in practice.
Risk #7: unconfirmed tax and accounting frameworks. Any potential CDC scheme sponsor would be ill-advised to launch such a scheme without absolute clarity as to its tax treatment, both for itself and also the membership. The accounting treatment is similarly unspecified.
Inevitably a new type of scheme will require clarification from HMRC and the accounting profession. No employer would embark on a CDC plan if they were not convinced that the accounting treatment would be defined contribution based, with no balance sheet disclosures beyond the contributions paid.
Risk #8: incompatibility with pensions freedoms. CDC’s inflexibility in accommodating the individual can only be overcome at the price of additional cost and complexity.
Members are not ’locked in’ to a CDC scheme, and can take a transfer value of their fair share of the assets at any time prior to retirement. Presumably Michael would argue that conventional DB schemes are equally inflexible at accommodating individuals – but the market today seems to suggest otherwise.
In this section of his note, Michael launches both an attack on with-profits as a concept, as well as his advocacy of them as a way to deliver CDC style benefits. I will leave you to judge which of these cases he makes most strongly, but I would note Michael’s comment which goes to the heart of why I think with profits would be a less efficient means to provide these benefits on page 11; “Unlike with-profits funds, CDC schemes do not provide guarantees and therefore ostensibly (?) do not require any form of (insurer’s) reserve fund.” CDC schemes do not need to hold back reserves – a process that will accentuate the inter-generational inequity that Michael is expresses concerns about. Michael’s advocacy of the use of annuities to pool longevity risk similarly takes us into the need for huge, inefficient capital reserves (as well as insurers’ profits) – an unnecessary diversion since a CDC scheme can pool longevity risk every bit as effectively in the pensions space, rather then insurance space.
Risk #9: modelling risk. CDC proponents point to superior returns relative to DC pots, but these are modelled, not founded upon empirical evidence, underpinned by assumptions (a “mature”, stable and fully funded scheme from inception) that are wholly inappropriate to Royal Mail’s particular circumstance.
I am not sure how the risks can be anything else other than modelled – empirical evidence would presumably mean actual operation of a CDC system identical to that proposed for 50 years or more. But self-evidently there is no such system to point to – references to the Dutch and Canadian systems are fundamentally wrong since those systems converted existing obligations into CDC style benefits – that is NOT being proposed for the UK.
Apart from noting the wide range of institutions – eg RSA, GAD, PPI as well as Aon – who have come to the same broad conclusions about the efficiency of CDC schemes over individual DC schemes (30-40% uplifts) we repeat that modelling has been carried out not just for what Michael calls mature schemes, but for schemes in their start up and wind down phases.
Risk #10: employer / employee miscommunication. It is unclear to what extent there is a gap between the CDC pension scheme that Royal Mail is offering and what its workforce thinks it voted for. Is there a risk of inadvertent mis-selling?
Of course there is the risk of mis-selling anything to do with pensions and personal finance more generally – DB pensions, DC pensions, transfers whatever. It is in order to protect members against this risk that we would expect efficient communication to become a mandatory part of the legislative process for classifying a scheme as a CDC scheme. Royal Mail and the CWU have committed to promoting the scheme to their members – not just at the start, but constantly and consistently throughout the life of the scheme.
CDC Politics and perspectives
Once he has listed the risks above Michael moves into a discussion of a number of issues which are not exactly central to the question of the merits or otherwise of CDC scheme, and whether legalisation should support their introduction. Most of his comments are assertions or speculations about the intentions or views of other parties – it is not apparent to me how Michael has arrived at his views of what these parties feel.
For example, the DWP and broader Government stance on CDC schemes, and in relation to international experiences, he repeats erroneous “lessons learnt” from the Netherlands and New Brunswick, territories that have introduced a fundamentally different version of CDC schemes. Michael also raises the issue of whether UK public sector schemes should move to a CDC structure. As a tax payer I may support this view – but I seem to remember a government minister talking about no changes to public sector schemes for the next 25 years?]
He concludes with his view that a combination of with-profits investments, drawdown and annuitisation at an advanced age will do a similar job. I would turn this on its head. If CDC schemes can do the same as his proposed combination of with-profits and annuities, but do so more efficiently (ie bigger benefits for members because there is no need to provide expensive insurance guarantees) with less need for member decision making – why not let CDC flourish?