We can’t reopen closed railways or pensions – but we can build anew

When I wrote this blog on Boxing Day about “coping with falling markets” – I did not explicitly make the link with CDC. The chart at the bottom of John’s tweet talks about the cost of closing open collective pensions and this is what John is picking up on.

Over the Christmas period I walked along several sections of the Somerset and Dorset Light Railway, some of the track closed down by Dr Beeching in the 1960s. It was a line that carried families down from Manchester to Bournemouth on the Pines Express.Pines Express

Alan Pickering told me of travelling on to Weymouth and so to Jersey by pubic transport. Today we look to drive or fly, we have closed the rail option for good. Houses now are built where the lines were, stations converted or destroyed. It cost a lot to close the railways but the cost of getting them back as they were is too high to be considered.

I am nostalgic for the days when you could buy back your DC benefits for what was called a “scheme pension”, you either bought added years or you just swapped your DC pot for a pension , the rate of exchange being determined by trustees with the help of actuaries.

The cost of doing this is as prohibitive as reopening the lines Beeching shut. Some things are gone and no amount of nostalgia can bring them back.

But how does CDC help a saver in a falling market?

The reason why a trustee will not grant a guaranteed scheme pension for a cash input ( a transfer in) is because the grant of the guarantee is made at the expense of the scheme sponsor who will pick up the cost of the guarantee if things go wrong. This isn’t what employers are for- they provide jobs – they do not act as pseudo insurers, there are limits to the liabilities they will take on and universally employers have stopped paying scheme pensions on transfers in.

However, the same need not be said of a DC scheme, which can take transfers in without increasing the liability to the employer. In individual DC arrangements , a member currently has the choice of individual buy-out – swapping the pot for an annuity – or individual draw-down- where the individual is on the hook for managing the “nastiest problem in finance”, an income for life.

This is where CDC could help. CDC could pay scheme pensions to people transferring in DC pots. The scheme pensions would not be guaranteed by anyone, not by the scheme or a sponsor or by the member, the CDC scheme pension is prone to fall as well as rise – though by judicious management – the CDC trustees can protect members from the most heinous risks of drawdown and the scant annuities offered by insurers.

In direct answer to John’s question, CDC can continue to provide scheme pensions at times of falling market on the mutual principles on which it is set up. The mechanism for paying scheme pensions is typically the allocation of cash in to pay pensions out. Cash comes into a collective pensions from dividends, bond coupons, rents and new contributions. Cash flows out of CDC plans through the payment of cash sums (commutation) , the payment of transfer values and the payment of CDC scheme pensions.

Professor Leech is making the same point in his comment to the blog John’s reposted

The answer is that asset prices are characterised by excess volatility. Market prices – determined by the irrational exuberance of the stock market rather than economic fundamentals – are many times more volatile than the economic fundamentals such as dividends. An open pension scheme can ride out (short term) market volatility because it is the economic fundamentals in terms of investment income flows that matter.

This fundamental principle of collectivism, is what makes an open collective pension scheme so attractive. As with Dr Beeching and his railways, the problems for open collective pensions is when they become closed collective pension schemes.

life cycle open

The only time that assets would need to be realised from a CDC arrangement, was when there was insufficient coverage from cash-in to meet payments out. This is what’s known as a run on the fund.

We will not see CDC schemes taking transfers in any time soon.

The Friends of CDC are a patient lot. Some of us (Derek Benstead in particular) have been voices crying in the wilderness the best part of 20 years already.

The CDC consultation – on which many of us are working – does not allow for transfers in or the setting up of schemes specifically to pay scheme pensions from DC pots. Both John and Andrew are right.

It may be that savers like me have to wait a decade to have scheme pensions paid from our DC pots. We may never get there!

But if we do not going on pointing out that at times like this (the S&P 500 was up 5% yesterday and has fallen many times that in the first part of December), people are being ruined by drawdown. CDC pensions , paid from a CDC fund at a rate determined by trustees on advice from actuaries are a half-way house between the perils of individual drawdown and the perils of a sponsor taking pension guarantees onto a balance sheet.

We may not see transfers in , any day soon; but logic suggests that the we will see them within the next 20 years. Anyone who is following the debate about default decumulation options from DC, will understand that the alternatives aren’t much more palatable than what’s on offer today.

Keep on pushing

Despite the obstacles to achieving CDC legislation, I am hopeful that in 2019 we will see the writing of the rules that in the next decade will allow the Royal Mail to run a CDC scheme. It is a start – and a decent start- but it is not the end.

There is no end – that is the message about pensions. We do not have to close collective schemes and when we do so – we cannot reopen them. We will keep on pushing to keep those schemes open – which are open, and open new schemes to replace those that are closed.

In the meantime we will keep on pushing to make sure that DC schemes run effeciently and they they are as well funded as can be.  We do not just need need dramatic reform, we need better practice with what we have got.

A recognition that there is necessary risk in pensions

We cannot afford to run pensions on the yields we get from gilts – we can’t now – we never could. Providing pensions is not risk-free.

We need to find the correct balance between risk and reward. right now we are offering people only the binary choice of annuity and drawdown, we are not offering a balanced option.

For people to take a balanced pension, they must accept some of the risk, and market risk is part of it – but they do not have to suffer the extremes of annuity penury or the pounds cost ravaging of drawdown gone wrong.


(but you’re on your own!)

About henry tapper

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

6 Responses to We can’t reopen closed railways or pensions – but we can build anew

  1. “CDC pensions , paid from a CDC fund at a rate determined by trustees on advice from actuaries are a half-way house between the perils of individual drawdown and the perils of a sponsor taking pension guarantees onto a balance sheet.”

    Worth thinking about what this means. Historically, DB trustees, advised by actuaries, have determined the contributions required to fund liabilities. They did so taking into account the expected returns on the invested capital, which are uncertain. So the cost of funding the liabilities largely depends on the required certainty. DB has been killed off by an increase in required certainty, under accounting pressures but also as a natural consequence of putting a mutual protection fund in place.

    DC, as a replacement for DB, offers much wider choice about the certainty required, notably after pension freedoms. Hedging the real income requirements is as expensive for an individual as for a scheme, but leaving inflation exposed (with a nominal annuity) does not meet everyone’s preferences for sustainability of living standards; drawdown, by contract, maximises the scope for matching both required certainty (or tolerance of real cuts) and preferred time profiles, with affordability. The problem with drawdown is not theoretical – it is the same problem as the contribution problem for DB schemes with an altered sign – but practical: who is to advise the joint drawdown rate and investment approach? The subtext here is ‘can they do it any better than actuaries?’

    Along comes CDC. But it is only another variation of the same problem of required certainty. If members value maximising spending subject to sustainability, and attach greater certainty than 50% to the sustainability (both period to period and for life), then the actuaries advising the draw rate will not be able to recommend paying a real income level which only has a 50% chance of being sustained year to year. But whereas in an individual arrangement the member requiring more certainty than 50% creates his or her own upward-only option by drawing initially at a lower rate, in a CDC arrangement the same cannot apply because the option will not benefit the same members. The scheme must base income on the mean expected return. Returns above or below the mean, which themselves trend, and adjustments to the means themselves over long periods, which we can term ‘smoothing errors’, will then lead to either real cuts larger than the member tolerates, or (to minimise the cuts) the investment strategy will have to trade more of the long-term real risk premium for lower nominal return volatility. Either of these are likely to be suboptimal relative to individual DC where the same constraints need not apply.

    For this logic to hold it is necessary only that the advice underpinning IDC should be equivalent to the advice available to CDC – or in other words the same ability to define utility and the same modelling capability ensuring consistency between each of resources, risk, time and outcomes. Though I do not wish to overstate the level of such skills in the actuarial profession, I do concede that the skill level in DC drawdown advice (amongst both IFAs and wealth managers) is by comparison still on a steep part of the curve. But that is not to say that it will not develop rapidly, nor spread rapidly with the aid of online interaction with decision models, particularly stochastic models that support decisions involving required certainty.

    It is therefore misleading to see CDC as somehow escaping the logic that applies to each off DB and DC. Modelling of any planned CDC solution should reveal whether the changes introduced to ‘make it work’ in fact alter the sustainability approach (testing tolerance of cuts) or the risk level (lowering the mean expected mean lifetime income) or whether some sleight of hand is in fact made by recourse to money illusion (failing to model inflation, as well as market, uncertainty – or pretending that a spending cut caused by inflation is somehow more tolerable than one caused by bad markets). I’m still waiting to see the modelling that has been done for Royal Mail (which as far as I know has not yet publicised) so my position is still theoretical.

  2. henry tapper says:

    Stuart, the modelling done by WTW and Aon has been triple checked by them and First Actuarial. I have seen it and it is only modelling – by definition a best guess. The questions for 140,000+ postal workers is how they got access to financial advice and how they paid for it. Frankly – there was no good answer to those questions.

  3. I understand that and it’s a pity, Henry. It would help if Royal Mail released any technical papers they are planning to base their scheme on, including the rules, the assumptions and stochastic tests of the rules under those assumptions, for review by others. I understand very well the limitations of modelling but it’s critical that there be testable rules where a product is based on a theoretical supposition. Particularly in this case, because risk sharing has been tried before and failed, not just because rules were broken but even where they were adhered to because they relied on poor assumptions. Kevin says he can’t release Aon’s modelling because RM won’t allow it.

    You make a separate point about access to advice. This is interesting because one of they key differences between CDC and DC with drawdown is (I thought) the need for advice. The latter gives a lot of choice to the member, including the all-important draw rate and the dependent risk approach. These choices require the involvement of an an adviser (preferably the investment manager itself) in developing collaboratively and iteratively the right definition of the individual utility that the retirement plan must seek to maximise, including any constraints, time preferences and valued optionality. The whole point of CDC is that it proposes to trade off this rich customisation, and its advice requirement, for a collective definition of utility and for a single set of constraints that operate for all. If CDC itself generates a need for advice, that affects the trade off significantly. It would be interesting to know what advice you think is necessary, at what points it arises and whether this is regulated personal advice requiring a recommendation.

    The point of the last question is that we are very much interested in the concept of informing personal selection without making a recommendation. This is currently incompatible with EU regulation which, because of the cost implications, is a key obstacle to supporting personal responsibility economically for all.

  4. henry tapper says:

    Stuart – read your post from the point of view of an ordinary person, one that is not steeped in pension lingo. Phrases like “rich customisation” mean nothing at all to most of us. That’s why 94% of us who don’t take advice.

    CDC generates no need for advice, inside the EU or otherwise.

    As regards RM not wanting to share modelling on social media – are you surprised?

  5. That’s not really very convincing, Henry. It’s not personal. RM or WTW (if that’s who they are relying on) may choose not to subject their modelling to peer review (this is after all in part a commercial contest) and nobody can force them to. Peer review is widely seen as the best way to ensure the technical integrity of new products or structures that depend on modelling, as CDC does. As for my ‘shocking negativity’, I have spent the greater part of my career in a quantitative environment (surely like many of your readers) where you try to test ideas to destruction and extending that principle to CDC is perfectly reasonable. ‘Stubborn agnosticism’ might be a fairer description. As one of few firms with a long track record of model-driven self-smoothing or DC drawdown, you might expect us to have an idea where the weaknesses are likely to lie. But if we can be shown there is a reasonably robust investment model to deliver the undoubted theoretical advantages of CDC, we will willingly acknowledge that.

    I’m sorry I mistook your comment ‘The questions for 140,000+ postal workers is how they got access to financial advice’ as implying a need for advice. It sounded like an interesting observation but obviously not for the reasons I thought.

Leave a Reply