Pension Schemes Bill

Written evidence submitted by John Ralfe (PS 09)

I would like to submit my comments on Part 3 of the Pensions Schemes Bill 2014/15, new "Collective Defined Contribution" pensions and would be very pleased to provide oral evidence to the Committee. CDC plans are intended to provide a higher and more stable pension versus a Defined Contribution pot combined with buying an annuity, by smoothing investment returns through "inter-generational" risk sharing.

I am an independent pension consultant working with a range of private sector employers, including FTSE350, as well as smaller unquoted companies and partnerships. One of my FTSE100 clients has introduced a form of "Defined Ambition" pension for its staff.

I have written widely on many aspects of pensions, in the Financial Times and elsewhere, as well as being interviewed regularly on the Today Programme. I have taken part recently in debates on CDC, including those organised by the NAPF and the CSFI.

My summary points are:

1 The detailed mechanics of how CDC could work in practice have not been clearly defined by its advocates and it seems to mean "all-things-to-all-people". This lack of clarity is reflected in the Bill, with much left to regulations.

2 Although the Netherlands is often used as an example, the structure and history of its CDC pensions are often misunderstood. I believe that the fundamental differences in Dutch CDC mean it cannot be adapted as a model for the UK.

3 Robust and transparent regulation is crucial for any new financial product to gain the confidence of consumers. Whilst Dutch CDC plans are highly regulated, the UK seems to be envisaging "DIY" regulation by individual trustees, with no specific funding rules.

4 "Inter-generational smoothing", when properly analysed, looks very like a Ponzi scheme, requiring a continuous stream of new members to continue, with the first generation of members getting a "free ride".

5 None of the employers I speak to are planning to offer CDC.

6 Although some of the practical issues of CDC mechanics are capable of resolution, CDC still has a fundamental problem. Its underlying premise is that equity risk declines with time, so a CDC plan can hold equities for longer than any individual and obtain the expected equity risk premium. Since this premise is incorrect, even if the details are resolved, CDC is flawed.


Definition of "CDC"


Unlike Defined Benefit or Defined Contribution pensions, which have clear definitions and clear mechanics, the commercial details of how "CDC" pensions could work in the UK are still unclear. As CDC has been discussed for several years, including a formal consultation process by the DWP in 2009, rejecting it, this lack of detail is extremely surprising. Setting up a new legal framework for CDC pensions, without very clear details of how it will work, risks wasting public time and money.


I suggest that the Committee ask CDC advocates to provide a straightforward two-page explanation addressed to potential members, who are not pension experts.


Dutch CDC plans


The Netherlands has often been cited as a CDC model which can be adapted to the UK, including by the pensions Minister. However, we should be clear that Dutch CDC plans were not originally set up as this, but as what we would recognise as Defined Benefit plans, with a company sponsor liable to make good any shortfalls between pension assets and liabilities.

In recent years, however, the Dutch Courts have ruled that the company sponsor had no such liability, and "DB" plans, transmogrified into what were christened "Collective Defined Contribution". A Dutch actuary, Falco Valkenburg, [1] chairperson of the Pension Committee of the Actuarial Association of Europe has said, "CDC was set up to look like Defined Benefit to members and Defined Contribution on company balance sheets".


As well as the obvious underlying social and political differences between the UK and the Netherlands, pension membership is compulsory for employees in the Netherlands, ensuring a continuous stream of new CDC members. Since pension scheme membership is voluntary in the UK, there would be no such continuous stream of new members, which is crucial for "risk sharing".


Many younger Dutch employees are concerned that their current pension contributions are being used to used to pay current pensioners, rather than to pay their pensions when they retire, and with little certainty that their pensions will ever be paid. There are also concerns that there will be a mass exodus, if pension membership becomes voluntary.

In addition, a number of influential Dutch pension experts have criticised the shortcomings of CDC arrangements and are recommending a move to DC, including this recent excellent article by Professors Lans Bovenburg and Raymond Gradus. [1]


Regulation of UK CDC plans


The Dutch pension system is, by UK standards, highly regulated and all CDC plans are obliged to use market discount rates fixed by the DNB to measure the value of their liabilities. Longevity assumptions are also set down by the DNB. There are strict deadlines for making up any shortfalls against the value of assets, by increasing contributions, not giving inflation increases or reducing pensions paid.


The UK has no such consistent and tough regulatory framework, even for DB pensions, with the Pensions Act 2004 setting down "scheme specific funding standards". The current Bill seems to envisage a similar funding standard for CDC plans, with individual trustees deciding on the valuation of liabilities, how to measure shortfalls and how to repair shortfalls.


"Inter-generational smoothing"


Because CDC property rights, unlike DC, are undefined, we must be clear how "inter-generational smoothing", would work in practice. In particular, if, and when, any individual CDC member could receive in pension payments more than their underlying share of assets, including contributions and returns, at the expense of another member.


Each generation may be prepared to sign up to a cross subsidy from younger to older if it knew the next generation would, in turn, also sign up - each generation running the risk of paying an older generation, in exchange for the possibility of receiving a payment from a younger generation. Because occupational pensions are compulsory in the Netherlands, this issue does apply.

But the first generation in a CDC plan gets a "free-ride". It has the possibility of a subsidy from the second generation, without having faced the possibility of subsidising an earlier generation. Equally the last generation faces the risk of paying the penultimate generation, but it cannot receive a payment from a younger generation. End-to-end the first generation gains at the expense of the last.


CDC "intergenerational risk sharing" only works with new generations of members, each one able to subsidise the previous generation, in what is a Ponzi scheme. If this is properly understood would any employees voluntarily join a CDC plan, fearing they could be the first and last generation of members?


It seems that poor equity returns could, for a period, lead to older members already drawing a pension getting more than their underlying share of pooled assets, at the expense of younger members contributing. This cross subsidy from young to old may encourage younger people to simply leave the CDC or not join in the first place – threatening any possible "intergenerational risk sharing".


The CDC mechanism, must also allow pension drawdown, announced in the Budget, and allow unused pension pots to be passed on to heirs, as recently announced. Neither of these features are available in the Dutch CDC, which pays a pension only.


Attitude of employers


UK private sector employers have spent the last decade trying to get out of the DB pension business, with its huge investment and longevity risk, totally unrelated to their core business. They have closed pension schemes to new members, reduced new pension promises and closed altogether to existing members.

DB has been replaced by DC, with the legal commitment stopping with the monthly pension contribution.

Against this background, all the employers I talk to are extremely reluctant to move away from DC and have said they will not be setting up CDC plans.


This reluctance is partly because of the risk that regulatory requirements may tighten at some point in the future, but is mainly concerns about reputational risk that if members are unhappy with the CDC plan at any point in the future, they will look to the employer for redress.


Some employers have also said that increasing labour mobility makes setting up individual company pension schemes questionable. I am not aware of any cross company CDC plans being considered and, indeed, I understand that the TUC has said it will not sponsor a CDC plan.

Insurance companies could set up a CDC, plan but since they already have standard "With Profits" arrangements it is not clear how a CDC plan would differ in practice.


Can CDC deliver higher expected pensions?

6a Advocates of UK CDC have claimed that it can provide a more stable pension 30 per cent to 50 per cent higher than the equivalent defined contribution pension, which is then used to buy an annuity. These figures were developed before the Budget changes and assumes the saver buys an annuity at retirement with their pension pot. Presumably, since people can remain invested in equities the CDC "premium" is lower, although I have not seen any comment on this.

A CDC plan, with economies of scale, could certainly lower transaction costs, but so can existing DC arrangements, including NEST.

6b CDC appears to be about "risk sharing" in a collective fund, versus individual accounts. Advocates say that because it has a longer time horizon than any single individual, a CDC plan can take more investment risk – a higher proportion of equities, a lower proportion of inflation-protected bonds – to generate higher investment returns and a bigger pension. This is the familiar argument that the risk that equities will earn less than inflation-protected bonds decreases with time, so long-term pension savers should hold more equities.

6c But the proper measure of long-term equity risk is not the volatility of past equity returns; it is the cost of buying insurance against underperformance versus the risk-free return – a "put" option on a stock market index. If risk really does reduce over time, the cost of equity put options should fall the longer the option period. In reality, the cost increases the longer the option period, reflecting increasing not decreasing risk. The theoretical price and actual prices charged by banks are about 25 per cent for 10 years and 30 per cent for 20 years.

October 2014




Prepared 22nd October 2014