First Actuarial on CDC (submission to W&P Select).

First

First Actuarial

First Actuarial is a partnership of consulting actuaries and administrators that offers the full range of pension services to both trustees and employers. Our clients’ pension schemes range in size from £0.5 million to nearly £2 billion in assets. However, the majority of our clients’ pension schemes have assets of under £50 million.

Amongst our actuaries we have actuaries who have a long standing interest in sustainable, quality pension provision. For example, we have an actuary who submitted a paper proposing a CDC scheme to the Stakeholder pension consultation in 1998.


 

Approaches to Collective Defined Contribution (“CDC”) schemes

Before we turn to the WPC’s questions, we make some introductory comments.

The term “CDC” has been used to describe a number of existing and imagined schemes set up in different legal frameworks and with very varied levels of benefit protection. It is important not to assume that the term “CDC” describes a fixed arrangement. The flexibility that is possible within a CDC arrangement is an advantage, not a drawback. It is often suggested that Dutch CDC schemes are a good model to follow. Whilst there are advantages to the Dutch approach, we would not wish to see it being fully replicated here as we believe it has, in its operation at least, been seen as a quasi-DB scheme which has led it into some difficulties more recently.

The key features of the type of CDC scheme we would like to see possible in the UK are:

  1. Contributions are defined in advance for both the employer and employee. If investment returns, experienced mortality, expenses or other items of experience affect costs, this is reflected in changes in member benefits. Risks are shared between members and there is no liability on an employer or current active members to contribute any more
  2. Assets in the scheme are held collectively for all scheme members and there are no individual pots of money. This means the scheme can plan cash flow collectively, using contributions and asset income to pay pensioners. This avoids the costs of investing money for some whilst at the same time divesting it for others. It also means the scheme has a longer investment time horizon, increasing the range of possible investments.
  3. The need to manage the scheme collectively means the scheme must have a funding plan setting out how it will manage the assets to generate investment returns and how it will allocate resources as the experience of the scheme plays out. The funding plan will be the key mechanism that determines how resources will be allocated between members including between generations of members.
  4. In order to understand their benefits, members will need clear communications. Member expectations might be framed around a certain target benefit or they could be set as a multiple of a varying pension “unit”. High quality communications will be key to ensuring members do not misunderstand the expected level or variability of their pension.

From the current position in the UK, it is possible to suggest two obvious ways in which a scheme meeting these principles might be designed. These are also perhaps polar opposites of potential approaches.


Employer sponsored CDC

  • A CDC scheme which is open only to the employees of the sponsoring employer.
  • The CDC scheme is not competing with other CDC schemes to attract membership from the general public.
  • There is a target benefit, which might include the same accrual rate of pension for years at a time, for all members regardless of their age.
  • The demonstration of inter-generational fairness is by moderately even benefit outcomes over time.
  • Like DB schemes, the target benefit accruing for older members is more costly than the target benefit for younger members. This is tolerable where there is an employer contribution, the value of the target benefit accruing exceeds the members’ contribution and the more even benefit outcome per person justifies an uneven value per person.
  • Quite possibly the employer has a Defined Benefit scheme which has been closed to new entrants and has become very expensive. The new CDC scheme is a sustainable alternative to the closed DB scheme.
  • To summarise, this is CDC as an alternative to employer sponsored Defined Benefit

CDC scheme open to the general public

  • A CDC scheme which is open to all. There is not necessarily an employer contribution.
  • Cross-subsidies in the purchase of target benefits need to be more limited. Each contribution purchases a target benefit of broadly equal value. For example, the conversion of contributions to benefits could be on an age-related basis reflecting market conditions on the date of payment. The conversion terms could also be gender related.
  • The demonstration of fairness is by crediting a target benefit of equal value to the contribution.
  • A best estimate actuarial basis is required for converting contributions into target benefits.
  • We can subdivide this category into two:
    • CDC schemes which focus on receiving transfer values at retirement, to provide an income-for-life in retirement. i.e. CDC as a decumulation vehicle. Underwriting might be necessary and/or desirable. This is CDC as an alternative to annuity purchase (more specifically, non-profit annuity purchase).
    • CDC schemes which collect regular contributions during working life. This is CDC as an alternative to individual savings accounts.

Of course, a CDC scheme can receive both regular contributions and transfers in, it is not one or the other. However, we make the distinction because “CDC as an alternative to an annuity” is one of the models which some commentators tend to focus on.

The spectrum of kinds of pension scheme

FA1    

 

UK pension provision tends to be polarised between a wholly defined benefit pension scheme at one extreme, and an individual defined contribution (money purchase) scheme at the other extreme. Defined benefit schemes are in severe decline: active membership of open DB schemes in the private sector is less than 1/10 of what it used to be. The common replacement, individual defined contribution (money purchase) schemes, place severe and difficult to manage risks on their members.

Better solutions lie towards the middle of the spectrum. A defined benefit scheme can be made more manageable by lowering the amount of defined benefit accruing and replacing the benefit reduction with a discretionary benefit. One way to do this is make the annual pension increases before and after retirement discretionary. We include a brief section on revitalising collective defined benefit pensions at the end of this paper.

A defined contribution scheme need not operate by means of individual money purchase accounts. Investment of a DC scheme would be much less constrained, and have a much longer time horizon enabling a wider choice of investment, if the assets were pooled and invested collectively. Less constrained investment should mean, on average, better returns and better benefit outcomes for members. Money purchase accounts require close member engagement to be invested wisely before and after retirement. A collective scheme requires less member engagement and is more suitable for the majority who do not have the time, skill or inclination to closely monitor their own investments. We provide below a “traffic light” chart comparing collective DC with individual DC.

 FA2

We think it is clear that Collective DC schemes have substantial potential advantages over individual DC schemes, whether the post retirement income is provide by draw down or by an annuity. This response paper provides the detail behind our conclusions.


The Work and Pensions Committee’s questions

Benefits to savers and the wider economy: Would CDC deliver tangible benefits to savers compared with other models?

CDC compared with Defined Benefit

Private sector defined benefit schemes which are open to new entrants are all but gone. ONS data shows that the number of active members in all private sector occupational schemes fluctuated between 5.8m and 6.5m over the period 1975 to 1995. Some of those are money purchase scheme members, so the number of defined benefit scheme active members was slightly lower, but an estimate of over 5m actives in open defined benefit schemes throughout 1975 to 1995 is reasonable.

The number of active members in DB schemes which are open to new entrants is down to 0.5 million (ONS Occupational Pension schemes survey 2016). It is an extraordinary state of affairs that the interest of employers in defined benefit schemes is less than 1/10 of what it used to be, and is still falling.

We have here an intergenerational divide, between past generations of private sector workers who have defined benefit pensions, and current and future generations of private sector workers who do not.

We also have significant intra-generational divides:

  • Between public service workers (still accruing defined benefit pensions) and those in the private sector, and
  • Between those in the private sector on good DC benefits and those on auto-enrolment minimum levels.

Employer Sponsored CDC provides a viable option for employers to sponsor a scheme providing a target benefit which has some similarity to a defined benefit, subject to remembering that a target benefit is variable and may not deliver stable benefit outcomes over time as a defined benefit would.

There is no contribution risk to the employer in sponsoring CDC, in contrast to the major employer contribution risk in defined benefit schemes.

The sponsoring employer has a legislative risk in defined benefit schemes. This risk has been realised by legislation increasing benefits with retrospective effect. This risk has also been realised by legislation imposing funding standards. An employer’s sponsorship of a defined benefit scheme is a major financial commitment, and if an employer does not feel in control of its commitment, it is unlikely to be willing to continue to make the commitment.

The accounting standard for pensions has some major problems. It values defined benefits as if the assets are invested entirely in corporate bonds, which for many schemes is an expensive and unsuitable investment strategy. An employer with insufficient assets on its balance sheet may find a defined benefit accounting deficit renders it technically insolvent. A CDC scheme could avoid this accounting problem provided that the legislative framework is clear in not establishing constructive benefit obligations. More employers would then be able to sponsor CDC than are able to sponsor DB.

The accounting standard values discretionary benefit practices as a constructive obligation: in effect, a discretion is treated as a defined benefit, even though it is not. A creative defined benefit scheme design which incorporates a discretionary benefit as a means of risk management and funding control is conceivable and entirely workable actuarially speaking, but is ineffective for accounting purposes. The accounting standard treats a discretionary benefit award as a cost to the employer charged to its profit and loss account, whereas the truth of the matter may be the discretionary benefit is paid for from the investment returns achieved.

In a CDC scheme, target benefits are provided as can be afforded from a fixed contribution. The employer’s contribution rate is defined, and there is no defined liability (or constructive obligation) to put on an employer’s balance sheet. The legislation should be framed such that it is clear to the accounting profession that CDC schemes are to be accounted for as DC schemes.

These problems which have put employers off defined benefit sponsorship (increasing contributions, lack of control of funding, legislative risk, accounting problems, insufficient covenant to cope with employer sponsorship risk) are absent from CDC schemes. Making CDC available creates the scope for employers to sponsor a target benefit scheme without risk to themselves.

The inequities which currently exists between defined benefit provision and current and future individual DC provision would be materially reduced, were employer sponsored CDC schemes to become a material part of private sector pension provision.

There are many closed to new entrant Defined Benefit schemes, with new employees admitted to individual DC. The employer’s contribution rate to the replacement individual DC scheme is usually set at a much lower level than the employer had previously contributed to the DB scheme. This is often attributed to the cost of deficit contributions to the legacy DB scheme absorbing part of the employer’s budget for pension contributions. But with FTSE companies spending 14 times as much on dividends as on deficit contributions[1], this affordability argument is difficult to defend. Employers have used the move to DC as an opportunity to significantly reduce pension scheme contributions, creating a two tier workforce.

CDC could provide a vehicle to unify pension provision for all employees, eliminating the present DB/individual DC divide for future service. CDC could be a more acceptable replacement for DB for future service, for those employees still in a DB scheme. CDC could be better than individual DC for newer employees.


CDC compared with individual defined contribution

Saving for retirement using an investment account has fundamental problems.

CDC vs Individual DC with annuity purchase

Buying a non-profit annuity is an expensive way to provide an income for life. The insurance company is constrained by insurance company regulation to invest cautiously. A certain investment return is certain to be low, giving low investment support for annuity payments. The insurer must also provide capital support, and a return must be provided to the providers of capital. Real yields on bonds are currently negative, which means that a person needs to outlive his or her life expectancy to appear to get pay-back for the cost of the annuity. Unsurprisingly, members tend not to see this as attractive.

Converting an investment pot to an annuity on retirement incurs large amounts of investment risk, as pre-retirement investments are sold and reinvested into an annuity. Movement between bond and non-bond markets can and does result in very large variations of annuity outcomes.

For an individual, annuity purchase curtails the investment time horizon at the point of retirement. A shortened investment time horizon reduces the scope for rewarding investment and increases the risk from making potentially rewarding investment.

In addition, to spread the disinvestment and reinvestment risk of selling investments and buying an annuity, investments can be sold over a period of time prior to retirement (referred to as “lifestyling”). This further curtails the time horizon for rewarding investment, reducing the average pension outcome.

Bond yields can be low for a generation, making for severe inter-generational differences of individual DC scheme outcomes for the same contribution input.

A collective DC scheme does not need to invest on a per person basis. It does not need to buy an annuity to provide an income for life. Rather, it has an actuarial plan for paying an income for life to its pensioners from its assets and contributions. The large at retirement risk of annuity purchase is avoided. The volatility of benefit outcomes caused by annuity purchase is avoided. The investment time horizon is not curtailed at retirement, but extends over the lifetime of the members, increasing the scope for more rewarding investment, to the potential benefit of the members.

This potential benefit to members of CDC is not guaranteed. For CDC to produce a better average pension than an annuity requires the CDC scheme to achieve a better return in retirement than the return guaranteed by an annuity, net of expenses. This is not certain to happen, but it is very likely.

CDC vs Individual DC with draw down in retirement

Since the introduction of the Freedom and Choice reforms, people with a pension savings account are no longer required to buy an annuity, but may instead withdraw money from the account however they like.

Draw down can avoid the at retirement disinvestment and reinvestment risk of annuity purchase, and it can avoid the curtailment of the investment time horizon caused by annuity purchase. Potentially more rewarding investment can be continued into retirement, if that is what the member chooses. But it must be borne in mind that it is potentially dangerous for an individual to continue to invest aggressively into retirement without a guaranteed income for life.

However, individual draw down leaves the member with the problem of uninsured longevity risk. Someone retiring at 65 ought to plan on living until 105 (say, if they are to have a low probability of outliving their savings), and spend their pension savings accordingly. Should the member then die at a typical age of 85, half their pension savings are left unspent at death.

To put this another way, if 100 65-year olds want to hold a pot to draw down a real income of £10,000 pa in retirement, then saving as individuals (and assuming a zero real rate of return for simplicity) they’d need £10,000 x 40 x 100 = £40,000,000 to safely plan for drawing an income until age 105. But some of those 100 pensioners won’t live for 40 years – we expect them to live on average for say 20 years. If they save collectively, they only need to save £10,000 x 20 x 100 = £20,000,000. If they do save £40,000,000, they can collectively have an income 100% higher. But if they only save £200,000 as an individual, they risk running out of money at age 85 when they might still have another 20 years of retirement ahead of them. This pooling of longevity risk that is a reason for the higher income available from a CDC arrangement, relative to draw down.

As an aside, while people find it difficult to comprehend the immense value of this pooling of risk in pensions drawdown, it is better understood in life assurance. Let’s say I want to provide £500,000 for my family if I die next year. I can’t afford to put away £500,000 but I only have a 0.1% chance of dying in the next 12 months – so I can just save £500. But if I put my £500 in an individual account, my family only get £500 on my death. It is only if I pool my £500 with other people’s by taking out a life insurance policy that those of us who are unfortunate enough to die can provide our families with £500,000. The principles are just the same.

Draw down is only really suitable for those who are well enough off both to live comfortably and to outlive all their assets (both pension assets and other assets) regardless of how long they live. Proportionately few people are in this fortunate position.

Conclusions

CDC can provide a good choice for the many who are less well off, for whom sufficient income for life is important. Compared to annuity purchase, CDC has fewer constraints on its investment strategy which can be exploited to probably, but not certainly, provide a better pension income than annuity purchase.

Compared to draw down, CDC provides an income for life which draw down does not, enabling a higher rate of income from CDC than from draw down, if the rate of draw down reflects the possibility of a very long life.

CDC is potentially more suitable for the majority, whose retirement savings need to be spent on themselves, rather than left partly unspent and passed on to family after death.

CDC and individual control of investment

Some people wish to manage their own investments. Many individual DC schemes provide a very large number of pooled funds for members to choose from. Typical experience of DC schemes that only a small minority of members make an investment decision of their own, the large majority letting their contributions go into the scheme’s default investment strategy. Self Invested Personal Pensions exist for those who wish to use retirement saving to make particular investments of their own choosing lying outside the scope of pooled funds.

Similarly, some people take an individual approach to saving. In a money purchase scheme, the member’s contributions and investment returns are ring fenced for the member alone and are not spent on anybody else. The pension outcome might be good, if investment conditions were beneficial. Or the pension outcome might be bad, if investment conditions were poor.

For those who want to manage their own investments, or who want to ring fence their own assets from everybody else, an individual Defined Contribution account is a natural choice.

Otherwise, for the large majority who do not wish to choose their own investments, CDC would add a good additional choice to individual DC. We have argued above that CDC is potentially better than both annuity purchase and draw down for the majority of people.

Efficiency

For those who already have the knowledge to arrange their own investments, it may be efficient to allow them to do so. The DC pension system does permit self directed investment, whether through the wide pooled fund choices in individual DC schemes or through Self-Invested Personal Pensions.

For those who do not already have the knowledge to arrange their own investments, it is not efficient to insist that everyone learns to do so. It is not reasonable to expect that everyone can devote the time and energy to attaining this knowledge. It could be more efficient for one informed person to take an investment decision on behalf of 10,000 than for 10,000 uninformed people to make their own decisions.

Collective defined benefit provision is much less common than it used to be. There needs to be a viable collective pension option for use in future. It is unlikely to be defined benefit, it should be CDC.

How would a continental-style collective approach work alongside individual freedom and choice?

We think the words “continental style” are unnecessary in this question. If it is a reference to Dutch CDC plans, then we would say that the Dutch model is neither the only model for the UK to follow, nor the preferable model.

Presumably the word “collective” is shorthand for “collective defined contribution”. Defined benefit schemes are also collective schemes, but they are not the focus of this inquiry. We have included a section on reinvigorating collective defined benefit schemes towards the end of our response.

We also prefer not to use the expression “freedom and choice” as we believe this is misleading for members (and repeats the mistake made in the late 1980s when members were encouraged to break free from the restrictions of their good defined benefit schemes – leading to a huge mis-selling review). It is very difficult for members to assert that they do not want freedom and choice – although for many the restrictions of a reliable income for life are more valuable.


The question we are answering is:

How would a collective defined contribution approach work alongside existing options which allow members to:

  • Use an individual defined contribution account to purchase a lifetime annuity.
  • Or purchase other forms of annuity (short term annuity, flexible annuity).
  • Withdraw their tax free cash allowance.
  • Withdraw further cash, subject to income tax.
  • Transfer from a defined benefit pension and use the transfer value in some combination of the above three ways, subject to receiving independent financial advice where the transfer value is over £30,000.

Some of these choices have material limitations:

  • A non-profit annuity provides an income for life, but the underlying investment is very cautious, making the annual income look poor relative to the purchase price.
  • Draw down enables the DC account to remain potentially more rewardingly invested during retirement, but then there is no income-for-life guarantee. The income drawn should be low in case a person lives a long time. Leaving money behind after an average lifespan is also inefficient for the less well off majority, who need to spend their savings on their own retirement, rather than leave them to be inherited.

The addition of CDC adds a useful additional choice to the above menu.

  • A CDC scheme should not have the limitations on investment which an annuity provider has. A CDC scheme can combine the provision of an income for life with a potentially more attractive investment policy.
  • Draw down provides the possibility of more rewarding investment, but cannot provide an income guaranteed to last a lifetime. A CDC scheme can combine a potentially more attractive investment policy with an income for life, albeit the rate of CDC income is variable.

Adding CDC to the existing choices provides a major improvement to the current choices. Aside from tax free cash, the three major options for taxed income become:

  • An annuity providing a certain annuity for life, although the rate of income will be comparatively low.
  • A CDC scheme, invested more rewardingly than an annuity and providing a less certain, but probably better, income for life.
  • Draw down, providing taxed cash on demand, possibly invested more rewardingly, but not an income guaranteed to last a lifetime.

It seems to us that this is a much better suite of options than the current two options of annuity or draw down.


Does this risk creating extra complexity and confusion? Would savers understand and trust the income ‘ambition’ offered by CDC?

There are difficulties with both of the existing pension solutions legislated for in the UK which mean few savers properly understand them and many actively distrust them.

The existing collectively provided option, defined benefit, is in serious decline. More flexible versions of defined benefit do not work for many employers due to inappropriate reporting required by the accounting standards. Many employers do not trust defined benefit, they feel they have been burned by liability increasing legislation and they fear the possibility of future adverse legislation.

Similarly many members misunderstand and distrust defined benefit. They frequently believe that changes made to future accrual in schemes have reduced their accrued benefits (when in general, they have not), they believe ending up in the PPF is a terrible outcome (often it is not), they think taking tax free cash is financially advantageous (it might not be, it depends on the rate of exchange between pension and cash) and they see paying the member contributions as poor value for money (almost universally it is not).

The common replacement, individual DC pensions, requires a high level of understanding and engagement from members. Most members do not have the knowledge or the inclination to engage with their individual DC pension scheme. This is evidenced by the high proportion of members using the default investment strategy. Individual DC accounts are not pensions, that is, income for life, until such time as an annuity is purchased, if it is.

Collective schemes, whether defined benefit or defined contribution, require some trust: both in the CDC concept, which does not guarantee its pensions, and in the trustees managing the scheme. Pension products already exist for those who do not wish to pool their savings with others and are able and willing to manage their own investments. These are individual defined contribution schemes (i.e. money purchase), including Self Invested Personal Pensions. The fact that some do not wish to pool is not a good reason to refuse to introduce a new kind of collective scheme based on pooling. Those who do not wish to pool are provided for with individual DC pensions. Those who do wish to pool, and/or who do not have the knowledge or inclination to manage their own investments, need to be provided with a collective solution. Defined benefit is mostly gone in the private sector. Collective defined contribution could be the way forward for collective schemes.

Trust is more likely to be forthcoming in a scheme whose management includes member representation in the form of member nominated trustees. New CDC schemes should also be a way of democratising pensions.

Consumer understanding is not a condition which needs to be met to introduce a product. Do consumers understand how a car works, or a television, or a mobile phone? Most do not, although some do. The production of these goods is regulated to ensure they are adequately safe to buy. It is neither necessary nor appropriate to attempt to ensure that everyone understands every kind of pension.

A pension scheme which delivers good outcomes with low consumer engagement is good. The weakness of individual DC is it can deliver poor outcomes if the markets are against you at the time of retirement, and high consumer engagement is not necessarily a fix for this problem. We think CDC will produce less volatile outcomes than individual DC, and probably better outcomes. It should be a better product for the unengaged than individual DC.

For decades, vinyl records were the principal means of enjoying recorded music. Over the years, new ways of storing sound recordings were developed: reel-to-reel tape, cassette tape, compact discs, MP3 players, internet streaming. At no stage was it reasonable for Government to say, “We already have enough ways to play back music, we are not going to permit any new ways, it would be too complex and consumers would be confused.”

The pensions industry in the UK is going backwards. Collective defined benefit schemes used to be more rewardingly invested and provided reliable pensions for millions of people, at an affordable cost to their employers. (The main unreliability of DB, benefit outcomes after employer insolvency, was fixed by the introduction of the Pension Protection Fund.) Now, a combination of legislative changes which have increased liabilities or reduced the scope for flexible defined benefit design, and increasingly cautious investment and funding strategies which have raised the funding target and raised the proportion of the target paid for by employer contributions, have resulted in very few DB schemes remaining open to new members.

In place of DB, individual DC is now the principal means of retirement saving in the private sector. The administration of individual DC pots is time consuming. Arranging investments individually rather than collectively is inefficient, the investment time horizon is shortened and material at retirement investment risk is introduced. A constrained investment strategy is likely to be less rewarding, resulting in lower average pensions. The at retirement investment risk makes for a highly uncertain retirement income outcome. We are moving from collective DB, which used to provide cost efficient reliable pensions, to individual DC, which provides expensive to administer, unreliable and probably lower pensions. This is a retrograde step.

To move pension provision forward, we need a pension vehicle which is collective, not individual, to avoid the at retirement investment risk and inefficiencies of individually planned investment. We need schemes which employers can reasonably decide to sponsor, because the risk to the employer is removed. We need schemes which provide more reliable pension outcomes than individual DC. CDC is the kind of scheme which will move the industry forward.


Converting DB schemes to CDC:

Could seriously underfunded DB pension schemes be resolved by changing their pension contract to CDC, along Dutch lines?

How would this be regulated and how would the loss of DB pension promises to scheme members be addressed?

The principal purpose of introducing CDC is to provide a better option for future pension provision. It is not about dealing with a legacy of the past.

The issue of dealing with seriously underfunded defined benefit pension schemes[2] is a separate matter which should be removed from the considerations of whether CDC should be introduced for future service. Should the Work and Pensions Committee decide to run a separate enquiry into defined benefit schemes, we would be pleased to give input to it, but the CDC inquiry is not the time or place.

We do not envisage running UK CDC schemes on Dutch lines. The reasons for this are given towards the end of this paper.

Member consent is not mentioned in these questions. If CDC is introduced, a member would be able to transfer between a defined benefit scheme, an individual DC scheme and a CDC scheme, in any direction, if they decided to do so.

It will of course be possible for employers struggling with their defined benefits schemes to construct transfer offers which may or may not be attractive to members, where the member’s consent is required to transfer. The requirement for the member’s consent is right and proper.

It will be open to The Pensions Regulator and to employers to agree on Regulated Apportionment Arrangements in which the offer of a transfer to a CDC scheme is part of the agreed package of terms for removing responsibility for a DB scheme from its employer.

One wonders whether, had CDC been available, a CDC scheme would have been a good option to put to members of the British Steel Pension Scheme. Were CDC schemes to be introduced, it would be possible for those who have already transferred out of BSPS into individual DC contracts to transfer on to a CDC scheme, if they wished.

CDC schemes are worth introducing in their own right. It is not helpful to confuse the issue by simultaneously talking about the undermining of defined benefits.


Regulation, governance and industry issues:

How would CDCs be regulated?

The actuarial valuation of a CDC scheme is mathematically similar to the actuarial valuation of a defined benefit scheme, it is the question being asked which is different. For a defined benefit scheme, the question is, “What contributions are needed (in addition to the assets) to pay for these benefits?” In a CDC scheme, the question is reversed, “What benefits can we afford to pay from these contributions and assets?” But the mathematics of answering these two questions is essentially the same.

Given The Pensions Regulator already has a statutory role reviewing the actuarial valuations of defined benefit schemes, it seems natural to give The Pensions Regulator a role of reviewing the actuarial plans of CDC schemes.

However, we do not think this goes far enough. In a defined benefit scheme, a member can look at the defined benefits as the definition of their rights. In an individual DC scheme, the member can look at their individual investment account as the definition of their rights. In a CDC scheme, the members’ rights are not clearly defined, they are the result of the actuarial plan. Consequently we think that full transparency of the management plan of a CDC scheme is important.

We propose that the principal documents setting out the strategy of the CDC scheme, the actuarial valuation, the statement of funding principles and the statement of investment principles (or whatever the equivalent documents for a CDC scheme are called), are published on a web site. This web site is to be available to all, not only to the members of the CDC scheme.

Of course, the primary valuation monitoring role lies with The Pensions Regulator. But publication of the strategy documents will make it possible for researchers and critics to evaluate the plans of CDC schemes and publish any commentary and criticisms which they wish to make. Some people are concerned about the risk of inter-generational inequity in the running of CDC. They should be given the material they need to evaluate whether their concerns are valid. If valid criticism is raised, then CDC trustees can heed it.

We think that the actuarial policy should set a basis for calculating transfer values in which the sum of transfer values for all members is maintained in close proximity to the value of the assets.


 

Is there appetite among employers and the UK pension industry to deliver CDC?

There are many are employers with a defined benefit scheme which is closed to new entrants and an open individual DC scheme. Two employees doing similar jobs may have very different pension packages as part of their remuneration.

Employers have been understandably reluctant to withdraw defined benefit accrual from their existing employees. It is likely that those employers who have not terminated defined benefit accrual already are those who are most reluctant to do so.

The introduction of CDC could make it easier for these employers to reform their pension provision. A replacement CDC scheme targeting, but not guaranteeing of course, the benefits of the closed defined benefit scheme could be a more attractive replacement scheme proposition than individual DC.

Because a CDC scheme does not give contribution risk to the employer, the employer could bring those employees presently in an individual DC scheme into a CDC scheme, unifying pension provision across the work force.

We have explained why we think CDC is a good solution. If the solution is good, employers and providers will adopt it. Where one large well known employer leads, others will follow. We expect that pension consultants will be keen to put forward the new CDC solution for their clients’ consideration.

Royal Mail has expressed its clear determination to adopt CDC if it is made available. This would be a unifying scheme covering an expected 140,000 employees formerly in defined benefit and individual DC schemes. Relative to the 500,000 workers in active membership of defined benefit schemes, this single new scheme would be a major positive development. Other employers will follow.

We expect you will find some positive responses to your consultation from some master trust DC providers. It would be good to add CDC accumulation and decumulation options to the many private sector workers being auto-enrolled into DC master trusts.


Would CDC funds have a clearer view towards investing for the long term?

In an individual DC scheme, the investment time horizon is limited to the member’s time of retirement, if an annuity is purchased on retirement. There are material disinvestment and reinvestment risks. Individual DC investment is limited to readily tradable investments.

A collective scheme (whether DB or DC) can plan its investment strategy for the scheme as a whole. Income can be used to make benefit payments, reducing the inefficiency of simultaneous investment of contributions and disinvestment to pay benefits. There is no need to buy an annuity at each member’s retirement, avoiding the disinvestment and reinvestment risks inherent in an individual DC scheme. A collective scheme does not need to be able to disinvest its assets all at once (unlike an individual DC pot at retirement), therefore some of its assets can be less liquid.

An open collective scheme in which new members join about as fast as other members leave or die has an infinite investment time horizon. It is highly capable of investing for the long term. Only a shrinking collective scheme has a finite investment time horizon.

CDC schemes have a much clearer view towards investing for the long term than individual DC.


Other issues

We conclude with a section on some issues which we think are relevant and worthy of explicit comment.

Cost efficiency and benefit outcomes from different kinds of scheme

A common debating point is whether or not a collective DC scheme can reasonably be expected to provide a larger pension than an individual DC scheme.

A traditional claim in favour of defined benefit schemes is that they are a cost efficient way to provide pensions.

The primary reason for different cost efficiencies between different kinds of scheme is the investment strategy adopted. The return earned on the investments increases the amount of pension which can be provided from the contributions. Whether or not the investment strategy is constrained, and whether or not the investment strategy is potentially rewarding, are important issues.

Individual DC is much the most common form of retirement saving in the private sector, so we will use the typical investment strategy of an individual DC scheme as the benchmark against which to compare other kinds of scheme.

After retirement, we assume the provision of an income-for-life by purchase of a non-profit annuity. In effect, this is investment in bonds. Before retirement, there is a free choice of investment, but the investment time horizon is constrained by the switch from pre-retirement investment to annuity purchase. We will characterise an individual DC scheme as having “growth” investment before retirement and bond investment after retirement, although this is a simplification.

A Defined Benefit scheme may have a better long run cost/benefit ratio than individual DC if it has more growth investment and less bond investment than this individual DC investment premise of growth before retirement, bonds after retirement.

A Defined Benefit scheme may become unattractively expensive if it has more bonds than this premise of growth before retirement, bonds after retirement. It is the case that as defined benefit schemes have been more cautiously invested, employers have found them unattractively expensive to sponsor. Increasing numbers of defined benefit schemes have been closed to new entrants or closed to benefit accrual entirely.

Similarly, if Collective DC is to expect to provide a bigger average benefit than Individual DC from the same contribution, then it needs to earn a better investment return than Individual DC, better than “growth before retirement, bonds after retirement”.

A collective scheme (both DB and CDC) has no need to change its investments as each individual member retires, it can plan its cash flows collectively. An open collective scheme, if it is not shrinking, can expect to have sufficient contribution and asset income to pay benefits without being a forced seller of investments. There is not a cash flow need to invest in bonds. The investment time horizon is long, increasing the scope for more rewarding investment than where there is a time horizon for selling investments to buy an annuity.

These comments still apply to a CDC decumulation scheme, in which the main source of income is transfers in at retirement, although the income from transfers in at retirement could be patchy and unreliable, so care is needed.

The investment advantages of an open, collective scheme are considerable, and should result in a better average pension outcome, most of the time, than individual DC. This of course is not guaranteed, but it is probable.


Inter-generational risk

A common worry about CDC schemes is that they may be inter-generationally unfair. The more specific worry is that excessive benefits might be paid to the current generation of pensioners at the expense of the generations of pensioners to follow.

First, we note that the unfairnesses of not proceeding with CDC are very great. We have the divides between:

  • Older generations of private sector workers who may have some defined benefit provision and current and future generations of private sector workers who do not.
  • Within each generation, public service workers typically have a defined benefit scheme available to them, but most private sector workers do not.
  • Different generations of retirees from individual DC schemes may have extremely different income in retirement outcomes.

Introducing CDC would reduce (but certainly not eliminate) this unfairness. To the extent that CDC produces benefit outcomes which are less volatile than individual DC (and, for employer sponsored CDC schemes, there could be a benefit target which has a DB type of structure, which could further serve to stabilise CDC outcomes), the above unfairnesses are all reduced.

  • A CDC scheme could be a less severe step away from DB than a move to individual DC.
  • CDC for private sector workers could reduce the public / private sector differences of pension provision.
  • CDC is expected to provide smaller differences of benefit outcome between and within generations of members than individual DC.

Considering single employer sponsored CDC schemes, we think the deliberate overpayment of benefits to pensioners is an unlikely scenario. On the contrary, it is rather more likely that early generations of pensioners are paid too little. This would be the outcome of a deliberately prudent actuarial policy for paying benefits. Money would be wrapped up in the actuarial margins and not spent on members.

We think there is merit in the prudent communication of benefit expectations to members. But this is different from a prudent plan for the payment of benefits to members. A prudent benefit payment plan is not necessary for the purpose of prudent communication.

A prudent benefit payment plan might be acceptable in an employer sponsored CDC scheme where the member is receiving benefits worth more than the member’s own contribution. It is less acceptable in a CDC scheme open to the general public.

A single employer CDC scheme is not in competition with other CDC schemes, so there is no motive to over distribute benefits to early generations of members.

There is a risk of a CDC open to all competing on the basis of historic payouts, which risks encouraging excessive payment to current pensioners. But there is also a risk of competing on the rate of exchange of contributions for target benefits, which risks over-allocating target benefits to those in work at the pensioners’ expense. The risk of bias works both ways.

Regulation and transparency are key. Actuarial plans for the payment of benefits must be lodged with The Pensions Regulator, with TPR having a right of review. That is, there can be a similar regime of TPR oversight for CDC schemes as already exists for DB schemes.

We also propose that actuarial reports on CDC schemes are published on a single web site with public access. We expect that in due course studies of these reports will be produced by academics, researchers working for think tanks and actuarial consultancies. Critics of CDC schemes should be given the material needed to explore their criticisms. If work is published making valid criticisms, then trustees and managers of CDC schemes can learn from it.

The fiduciary governance framework should include significant representation of members and sponsoring employers (where there are employer sponsors).

The scope for inter-generational unfairness in CDC outcomes does exist, but we think the unfairness inherent in not introducing CDC is likely to be greater.


The Dutch pension model and Regulatory Own Funds schemes

The UK already has regulation for “Regulatory Own Funds” (ROF) schemes. We understand that ROF regulation is intended to be an implementation in the UK of what are often referred to as “Dutch CDC” schemes.

  • A UK ROF scheme provides a defined benefit. It is not a CDC scheme in which no benefits are defined.
  • But the contributions are fixed.
  • Therefore there is an irreconcilable conflict between the contribution rate and the defined benefits, were the defined benefits to become under funded. The UK ROF regulations provide no scope for the resolution of this conflict by either the reduction of the defined benefit or the increase of the defined contributions.
  • In the Netherlands, the actuarial funding of their schemes is driven by bond yields. This in turn encourages the investment in bonds and LDI contracts, which in today’s markets provide a low return. This does not result in an attractive cost / benefit balance.
  • Schemes in the Netherlands have been troubled by low bond yields. It has been difficult to continue to provide the promised benefits. It has been difficult to resolve upon modification to the promised benefits and to the contribution rates.

A Dutch CDC scheme is already possible in the UK, whether by utilising the UK’s Regulatory Own Funds legislation or by using Netherlands legislation. It has not been done, and it is not an attractive proposition to do so.

Making CDC schemes available in the UK is not about adopting the Dutch model. Criticisms of the Dutch model are not necessarily relevant to considerations of CDC for the UK.


Revitalising collective defined benefit

The topic of the inquiry is collective defined contribution schemes. Although it is not a topic for this inquiry, were the Work and Pension Committee minded to consider what could be done to reinvigorate collective defined benefit pension provision then we suggest investigating the following:

  • Whether to set The Pensions Regulator an explicit objective of increasing the active membership of private sector defined benefit pension schemes.

TPR’s approach to defined benefit regulation appears to positively discourage defined benefit accrual and to discourage affordable investment and funding strategies. We doubt that the implementation of the funding regime is working out as intended.


  • The accounting standard for pensions.

It is possible to conceive of a defined benefit scheme in which there is 1) a lowered defined benefit promise, 2) a fixed employer’s contribution, and 3) a material discretionary benefit which is used as the principal means of maintaining the balance between assets, contributions and benefits. The accounting standard would treat the discretionary benefit as if it were defined (a constructive obligation) and ignore the fact that the employer’s contribution is fixed. Discretionary increases would only be awarded if investment return paid for them, but the accounting standards charges the cost of an increase to the employer’s P&L as if the employer were paying for it and disregards the investment return which actually pays for it. The accounting treatment is neither true nor fair and is an obstacle to sustainable defined benefit scheme design.


  • The risk reserving requirements for banks and building societies.

The Prudential Regulation Authority bases its risk capital requirements on the accounting disclosures. The accounting standard hypothesises investment in corporate bonds which, if a scheme has mostly inflation related liabilities, as most do, is not a reasonable investment strategy. A bank’s pension scheme can invest in corporate bonds to minimise the bank’s risk reserving requirements, but to do so makes the scheme unaffordably expensive at current bond yields and unsuitably invested. Or a bank’s pension scheme can invest more rewardingly to reduce the expected cost of benefit provision, but to do so greatly increases the capital requirement. Banks and building societies are effectively prevented from sponsoring open defined benefit schemes by their regulatory environment.


 

[1] The Pensions Regulator: Annual Funding Statement May 2017

[2] Actually, rather than using the phrase “seriously underfunded defined benefit pension schemes” it would be better to use a phrase on the lines of “defined benefit schemes with employers who struggle to afford their contributions”. The principal location of the problem is in the financial strength of the employer, not in the financial strength of the pension scheme, although this may also be a factor.

About henry tapper

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

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s