Is CDC “Flat World” – or “Flat Earth”?

The World Is Flat: A Brief History of the Twenty-first Century is an international best-selling book by Thomas L. Friedman that analyzes globalization

The Flat Earth model is an archaic conception of Earth’s shape as a plane or disk. Many ancient cultures subscribed to a flat Earth cosmography – (both Wikipedia)

The distinction between Flat World and Flat Earth suggests how easily a simple idea can have binary engagement. CDC is much the same – I see  CDC as progressive and a means of binding society together, John Ralfe sees CDC (and me) as bogus.  Trump has proved that the insistence of the globalisation lobby can be checked and while I’d stop short of likening John to Donald, I suspect that both of them have an important role in keeping the world honest.

The point of this article is to explain that while it is fun to ridicule CDC, it is impossible to ignore it, like globalisation , collectivisation is not going away!

The recent conversion of St Guy of Opperman to CDC – suggests that the £2bn uplift in the Royal Mail share price, resulting from the CDC settlement – is a market indicator – even he could not ignore.

Flat earth

The Price of Freedom

No one has costed “freedom and choice”. The Treasury tried in their various impact assessments. If they had gone to consultation over the changes in the tax laws on the payment of DC pots, we (as Friends of CDC) would have pointed out

  1. That paying a wage for life is difficult and requires scale
  2. That managing a wage for life with any certainty – as an individual – requires an annuity

It would not have been enough for us just to have said this, Kenny Tindall asked on Twitter yesterday what proof there was that CDC provided more certain pension outcomes and I’ll start by showing how expensive it is to manage a typical drawdown

Breakdown of costs

The model above is only one of a number knocking about, but it fairly represents the kind of costs you can expect to pay for an efficient drawdown product.

Paying 0.5% of your fund for financial advice – is akin to paying somebody to look after your car, you could do it yourself, but most people will pay a mechanic.

Investment Management costs are what you pay upfront to people to manage your money, typically in a fund. They’re business as usual costs – the price of fuel. They do of course come down where a fund grows in size as is evidenced from this article on Prufund .

Platform fees are payable to those who manage the technology and in car maintenance terms , they represent routine maintenance charges – including MOTs!

Finally there are the occasional costs of running a cost which include the cost of parts when things break and the cost of labour to fit the new parts. As always in these things, these costs are underestimated as unknown and could be called the hidden costs of running a car or fund

The price of freedom, in this model is 1.65% of the amount invested, a considerable amount compared with the “rule of thumb” 4%pa drawdown.

Put simply , for most people, the price of freedom is a high price; it is too high to make drawdown sustainable for most people. Drawdown is not a mass market solution because even for 1.65% pa- you are unlikely to get sufficient value to meet your need for a satisfactory income that lasts as long as you do.

So simply in terms of defining pensions freedoms in terms of an alternative, there is no mass market alternative to the discredited annuity.

The three arguments for more certain collective outcomes

A collective approach can do three things

  1. It can bring down the cost of pension management by disintermediating advisory fund management , platform and hidden (transactional) costs .
  2. It can improve value by investing in more appropriate assets (for the purpose of paying a wage for life)
  3. It can manage the problem we have of not knowing when we are going to die.

These are the three reasons I give Kenny Tindall for why CDC gives more certain outcomes. But I am aware that so far- I have only dealt with  “1”.

Collectives add value

Collectives aren’t just about reducing the impact of costs and charges, they add value in their own way.

It’s long been known that pension schemes can take certain kinds of risks , other schemes cannot – and be rewarded for those risks. Taken together those risks give rise to an “illiquidity premium”, a measurable amount of out-performance resulting from a scheme investing in long-term illiquid assets which give favourable long-term returns,

However, the illiquidity premium has lately been squandered – because pension schemes of all types have chose to de-risk rather than take a long-term view. Exponents of de-risking point out that there are other risks that their “de-risking” avoid that make missing out on the illiquidity premium worthwhile. That may be the case if DC plans are forced to de-risk prior to buying an annuity and if DB plans are put on a similar “glide path to buy-out. But handing over your assets to an insurance company for the payment of an annuity was never an essential end-game! I find myself in agreement with Ed Truell who writes in the FT

 “As a co-founder of Pension Insurance Corp, I know better than most the strictures of an insurance regime that is ill-suited to the long-term nature of pension asset and liability management.”

He added that insurance was an expensive way to secure pension liabilities when employers ought to be devoting their attention elsewhere.

“British companies need to be freed up from their legacy liabilities to invest in their businesses, restore productivity and continue to boost employment and growth,” .

The original point of running a pension scheme was to keep it going for future members.

CDC schemes have the opportunity to restate that original aim and avoid de-risking altogether, they can take a long-term view and invest in assets that give them an illiquidity premium. They do not have to sell these assets to pay pensions (as has to happen in the individual DC model). They can rely on future contributions and income from existing assets to meet the target pensions.

Certainty of income from longevity pooling

The third and most obvious advantage of a CDC scheme is that it can insure longevity risk from within its pool of members.  As Abraham puts it in his recent book,


Instead of having to rely on super-outperformance or an outsourced insurer (either traditional insurance or capital market solutions), a CDC scheme relies on pooling the mortality experience of all members, those who die soonest within the scheme subsidise the pensions of those who live long.

This process is abhorred by libertarians who consider any form of cross-subsidy, the spawn of the devil. But it should be pointed out that this kind of social insurance is the basis of all mutual movements, there is a solidarity between people based on our fundamental insecurity about our capacity to survive.

We all know that there is a general underestimation of how long we live and this appears to be particularly the case among people who live longest. Ironically, those who stand to gain most from guaranteed pensions are those who are most wealthy. These wealthy people are most likely to leave a longevity pool – making things that much the easier for those with genuine short-life expectancy!

This is why I think CDC schemes should offer the door to all members who want to transfer out, including those who are receiving pensions, for every pensioner who lives on the grounds of ill-health, there will be twice that many leaving for reasons or wealth (typically inheritable wealth).

CDC will I am sure, benefit from it being spurned by those wealthy enough to live forever.

Flat world or flat earth?

I hope in this article, I have put up a reasonable argument for considering CDC as a more certain source of income in retirement than either drawdown or annuity purchase. I’m aware that this article is not splattered with numbers, I’ll leave that for my actuarial colleagues.

I hope too, that I am seen to be talking sense, and that through reading this far, you have inched closer to the “CDC is “flat world” – rather than “flat earth” thinking”.


Flat earth thinking

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.

3 Responses to Is CDC “Flat World” – or “Flat Earth”?

  1. Con Keating says:

    I willoffer another argument in response to the question: what proof is there that CDC offers greater certainty. The TUC published a paper on IDC outcomes -they exhibited 17.7% variability (standard deviation) and this was after a quite conservative 11% variability of pot outcomes in the accumulation phase. They did not attribute many of the costs outlined by Henry above. Now 17.7% is the sort of volatility we would expect of an all equity portfolio – the market. So if we were to run an all equity CDC scheme, then members would be no worse off than using IDC, and of course the risk-sharing rules would reduce the volatility of outcomes for members to levels below that of the asset portfolio.
    It is of course trivial to deliver portfolio outcomes in collective arrangements which are less volatile than markets. This is greatly facilitated by the fact that with iDC, the portfolio management objective is to maximise asset values at all times and is a matter of inter-member conflict, while the CDC objective is to maximise returns on average

  2. Con says: ‘It is of course trivial to deliver portfolio outcomes in collective arrangements which are less volatile than markets. This is greatly facilitated by the fact that with IDC, the portfolio management objective is to maximise asset values at all times and is a matter of inter-member conflict, while the CDC objective is to maximise returns on average.’

    I hesitate to challenge Con but I cannot agree. The objective – and the technical challenges associated with it – is exactly the same: to smooth draw from from a volatile portfolio. And as modellers we do not think this is trivial.

    If Con’s point is that the individual DC portfolio is not structured optimally to maximise sustainable draw, that’s because he’s assumed the DC portfolio behind the drawdown is collective but standardised at the asset allocation level. This may indeed point up a weakness in the adequacy of today’s DC for decumulation purposes – i.e. a case of Pension Freedoms arriving before the product solutions they called for. Even if true, this does not need to be the case. It is perfectly possible to combine scale economies from collective passive instruments with customised allocations to those building blocks using the cost advantages of technology. It is (for instance) possible to construct the individual DC portfolio optimally to reflect both idiosyncratic risk preferences and the personal cash-flow dates (for example using a two-asset or LDI-style approach combining a common collective equity portfolio for later liabilities and risk-free assets to match early liabilities, acting like a temporary annuity).

    If we assume a fit-for-purpose, dedicated DC portfolio, then the question posed by Con becomes whether everyone can safely draw at a higher level from the CDC scheme than everyone could if drawing from their individual ‘LDI pots’. On the flat earth I apparently inhabit, I can find no reason to assume they could. Both are subject to identical errors in estimating the return path of the portfolio, as a function of the distribution of real equity returns over different holding periods. My flat earth may be bounded, so I’m misjudging what happens beyond the horizon, but at least the CDC problem is usually defined to lie well inside the visible horizons, on the basis of smoothing short and medium-term returns.

    In an article here,, we gave some estimates of the holding-period variances in outcomes from Fowler Drew’s drawdown model that includes a mean-reversion assumption implicit in smoothing. The very large holding-period differences are important as they force you to think about the issue of fairness not only between cohorts of retirees but also between pre- and post-retirement members. Incidentally, the data outputs from the model are all based on that two-asset portfolio so they eliminate completely the massive uncertainties for inflation and nominal interest rates that proved so damaging for with-profits funds.

    The errors are the same in each model but the consequences are different. The individual member drawing at a rate with a 50% chance of sustaining income in real terms faces a large cut in spending if returns prove to be lower. To avoid large cuts they therefore typically draw at a lower rate, with a higher chance of sustainability but also, then, a high chance of a surplus emerging if actual returns are higher than the prudent level assumed. That option on higher spending has a value even though the timing of the payoff may be somewhat suboptimal (although this can be addressed to some extent by profiling the spending – more early when it is most highly valued). But if the CDC plan paid at the same rate as the individual’s 50% draw, it has to accept a high chance of needing to cut pensions in payment significantly. And I’m not talking small course corrections like the Dutch schemes did. If instead the CDC paid only an income that had a high chance of being sustained without large cuts, it creates the same option of higher income later but this payoff will not necessarily benefit the same members. That reduces and may even destroy its utility.

    If CDC schemes, like individuals, have to limit the size of cuts but, unlike individuals, also have to limit the chance of surplus accruing to the wrong people, they can only do so by holding a portfolio that trades off lower volatility against the median return. In fact that’s the trade off all the products with smoothing inherent in their structure have made: for instance Prufunds with-profits and MetLife retirement funds with guarantees (now closed). They also have to adopt a pro-cyclical policy (cutting or limiting losses) even though the best long-term outcomes accrue to a contrarian policy.

    I admire the ambition behind defined ambition pensions and would love them to work. If my scepticism about the underlying maths is misplaced, please show me why.

Leave a Reply