Con Keating -“CDC is no hardy perennial”



A critic of CDC began a recent blog with: “Collective Defined Contribution is a hardy perennial”, which, to my mind, is only appropriate for a scheme of arrangement capable of delivering pension benefits more efficiently and equitably than any other – including occupational DB. An entire mythology of “problems” with CDC has developed.

I suspect this is another debate, rather than discussion, where the entrenched position is not subject to change or adaptation, but I write in the hope that others may be more amenable to discussion and reasoned argument, and be able to see the many advantages CDC brings. I take extracts from that recent blog, italicised below, as a straw man to illustrate the waywardness of the flawed negative analysis. The blog was described to me as having been circulating on Twitter (to which I do not subscribe) and I do not know the author.

Like the three bears’ porridge, chairs and beds, CDC has a number of conditions that have to be ‘just right’ if they are to appeal to Goldilocks, or in this metaphor, UK plc. If just one of those aspects is missing, CDC simply cannot work.” This is simply untrue at many levels. CDC does not have to appeal to UK plc or indeed any external sponsor. The employer does no more than make a contribution into a scheme – something it is already doing with DC and in a few, increasingly rare, DB arrangements. There is no further recourse. Let’s now deal with the so-called problematic elements.

“Compulsion in enrolment. CDC relies on inter-generational risk sharing.” This is untrue. It does utilise intra-labour force-generational risk-sharing, and that means that there are circumstances in which young subsidise old and others where old subsidise young. It is valuable to all. An attraction which will hold across generations, as they arise.

“Or, more simply, it relies on young people to take the financial strain for older people if assets do not perform as expected.” This is again untrue. Any financial strain from asset underperformance may be shared equitably across all members.

“Risk sharing cannot happen if younger people opt out.” This is true not just of younger members but of any and all, but it involves moving to an arrangement which is riskier for that member. Given the attractions of CDC, opting out is an act of self-harm.

“Scale. CDC is a hungry beast and needs feeding with people and money. If you want to share risk, you need lots of people to share it between.” The lower bound for the value of the risk-sharing properties of CDC to be material is around 400 members; it is not a hungry beast. Larger scheme populations are superior, in a probabilistic sense, but they are not in any sense necessary. It takes just two to share. One of the attractions of CDC is that it may easily attract members from many differing employers. Diversity in membership is attractive.

“It would be nice to know whether ‘along the lines of the Dutch system’ includes permitting non-consensual conversion of already-accrued DB pensions. If it does, that should be expressed openly, so we at least understand exactly what we are debating. If it does not, UK CDC could only realistically emerge from the conversion of existing DC, and lots of it.” There is no plan for any non-consensual transfers. Indeed, schemes may prefer to start from a tabula rasa. The idea that schemes could only emerge from the ashes of DB or DC is simply fallacious. If they chose to do so, schemes could accept (even new) members and their ‘pots’ at any point in time. In this they might constitute an attractive decumulation option to individual DC savers.

“Compulsion for employers. CDC requires a constant stream of people to be able to risk share. While today’s 60 year old may want to risk share with a 25 year old, today’s 25 year old needs to be confident that there will be some 25 year olds in the plan in 35 years’ time. That is going to be very hard to assure if employers are able to walk away from the system. Few employers will want to make such a commitment.” The question of employers walking away shows a fundamental lack of comprehension. No employer has any obligation beyond making the current year’s contribution. There do not need to be any 25-year-olds entering or in the scheme in 35-years-time. Risk-sharing takes place among those members present.

“Compulsion in retirement. If risk sharing is to work, then disgruntled pensioners wouldn’t be able to transfer their benefits, because that might permit them to ‘select against’ the other members and generations. They would have to see their CDC pension through. Non-transferability was one reason that annuities were unpopular.” This is again untrue. Members would be able to transfer out at any point in time. The transfer value would be their equitable share of the scheme’s assets at that time. In transferring, they are leaving an arrangement which has value to them and also exercising an option which is most valuable when unexercised.

“Indexed pensions. CDC’s safety valve is in escalating pensions. Once risk has been shared between the generations, if there is still not enough money in the pot, pensions in payment would have to be cut. It’s pretty much unthinkable that this would mean actual reductions to pensions in payment, so CDC advocates usually propose that instead, pensions simply would not increase in payment.” CDC’s safety valve does not in escalating pensions – CDC may offer either, or both, fixed or escalating pensions. CDC’s safety valve lies in adjusting the entitlements of all members equitably. This applies to both gains and losses. As this may require cutting pensions in payment, there are further inter-temporal measures that may be considered and applied subject to the agreement of members – a further dimension to the most basic risk-sharing properties of the member mutual.

“There is just one problem here; it requires the member to want an escalating pension in the first place. If they wanted a level pension, they could probably do better by simply buying an annuity, because some expectation of escalation will have been priced into CDC.” This is just nonsense based upon an earlier misconception.

“The trouble is, most people do not want escalating pensions. Take a look at figure 6 on page 26 of this FCA paper. It shows us that from 2008 to 2012, every year just 7% of DC retirees bought annuities that escalated in payment. That means 93% went for level annuities. ‘Like an escalating annuity, but it might not go up’ does not strike me as a winning product. That first 100,000 members is beginning to look even further away.” The author of this blog is piling ossa on pelion. There is no doubt that people did select fixed rather than escalating annuities, which was a product of the terms then offered. However, surveys of what people actually want show the following preferences:con graph

That FCA data predates freedom and choice, but that is likely to have simply exacerbated the gap between what CDC could theoretically offer, and what real people actually want. Indexed pensions are on offer from DB plans, but we are seeing unprecedented numbers of people voting with their transfer forms and moving to DC.” The level of CETVs is a product of the nonsensical valuation standards about to DB liabilities and the unwillingness of trustees to commission insufficiency reports and apply the findings rigorously. This of course suits many employer sponsors as it discharges liabilities more cost effectively than bulk buy-out.

Indexed pensions bring to life the central problem of CDC. CDC pensions do not need to be indexed. CDC could theoretically offer a better pension, but everyone has to fit in with the way it works. One size has to fit all. It does not.

The way CDC works might well be what society and government (and some sections of the pensions industry) think people should want and should have. We think that it is what most people want- it is what they say they want. But, for better or worse, with its reliance on compulsion and its almost complete absence of flexibility, the way CDC works does not remotely match what people’s actions tell us they want. CDC does not rely upon compulsion. The actions we have observed are in a world without CDC – a world with many problems which condition member behaviour. CDC is actually more flexible and efficient than occupational DB. Actions, such as those cited, are responses to particular circumstances. People do not all want the same thing. This is one of the few things in this blog which is true, but few well-informed people would wish to make inferior choices.

I am not saying these tenets cannot be achieved. But I am saying they cannot be achieved without forcing a lot of people against their will.” There is absolutely no need whatsoever for compulsion.

It should also be understood that CDC has eliminated the risk of sponsor insolvency. The member mutual organisation means that pensions promises are promises made to themselves, which renders the concept of scheme insolvency economically vacuous. The risk-sharing arrangements are beneficial to all; the key is equitable participation for all within this arrangement. The technology to facilitate this, cost-effectively, exists.

About henry tapper

Founder of the Pension PlayPen,, partner of Stella, father of Olly . I am the Pension Plowman
This entry was posted in CDC, pensions and tagged , , , , . Bookmark the permalink.

2 Responses to Con Keating -“CDC is no hardy perennial”

  1. Hello. It is my linkedin article that you’re critiquing, so I guess I should respond. There is much here that I’m afraid I cannot agree with. I struggle to understand why you are willing to overlook what 93% of people actually did in favour of how they responded to surveys of theoretical questions seemingly asked without context (I’d welcome the source). But I’ll start with scale.

    Your article is quite clear that you don’t see CDC as relying on the conversion of existing DB or DC – that clarity is welcome. But the issue of scale is not just one of risk sharing; there are also considerable costs to share and without conversion, how will the costs be met?

    Who is going to pay for lawyers to draft the trust deed and rules? Who is going to pay for trustees, for auditors, for investment managers and consultants? Who is going to pay for the likely Mastertrust Assurance Framework? For communications? For the professional who will have to advise both on appropriate contribution rates and on how returns should to be divided up between the various generations? For administration and record keeping?

    This will be (for the UK) a unique form of pension saving, so precious little of these services will be standard or off the shelf. Even in the Netherlands, very few CDC plans have been set up from scratch. This is new ground being trodden, and that is rarely cheap.

    The model you describe means the plan has no ‘sponsors’, and so I must assume that all these costs will need to be met by members. It will take considerably more than 400 members to make bearing these costs a worthwhile exercise for anyone but the advisers.

    And so I ask again. Where are the first, say, 100,000 members and £5 billion in assets going to come from? If you believe those numbers are too high, what scale do you believe a CDC plan would need to be self-sustaining?

  2. Con Keating says:

    I do not overlook the behaviour of people. Rather I consider how I or a reasonable person would have behaved in the circumstances then prevailing.
    The survey cited was conducted by IPSOS REID – it was well designed. It was conducted in August 2016 and published in March or April of this year. It is titled, correctly in my opinion: “The pensions Canadians want.”
    I think you make far too much of the likely costs. In the past five years I have set up both a Credit Union and a Friendly Society – in neither case did the costs of establishment exceed £200,000. A CDC scheme is less complex and would likely cost far less. But this conversation has added an item to the “to do list” – create a model CDC legal template. That is less than 10% of the initial year’s contributions in my 400-person scheme.
    I have looked closely at running costs for a CDC – with today’s technology, which would also offer a communication channel for members, the administration and record keeping is largely automated, and the 400-person scheme would require an establishment of no more two people. Auditors would be necessary. investment managers would be paid as they currently are, from the returns of the assets managed. It is far from obvious that trustees, consultants, advisors and other so-called professionals are necessary. It is clear that your thinking is very much in an existing DB framework – for example, division of returns. Assets are allocated equitably among members on the basis of their entitlements – and those entitlements are determined by the contributions made and the benefits “promised”.
    In a mature or steady state, a 400-person scheme would have assets of £300 million or so; it could sustainably support all-in costs of 0.5%, and be cost effective for members. The gains from the institutional organisation, the risk-sharing and risk-pooling, are far greater than that.
    You say that “This will be (for the UK) a unique form of pension saving, so precious little of these services will be standard or off the shelf. Even in the Netherlands, very few CDC plans have been set up from scratch. This is new ground being trodden, and that is rarely cheap.” I agree that this is new ground. One of the beauties of today’s technologies is precisely that they may be used to customise the services offered to their users at marginal costs which are trivial.
    If you want to see an example of a highly successful form of CDC, (though it differs in some regards from that described in the blog and this comment), have a look at the Canadian CAAT. The backlog of firms wanting to offer that to their employees is years long.

Leave a Reply