This is a review of: The Legal Differences between CIDC and CDC by Prof. dr. Hans van Meerten and Elmar Schmidt, February 2018. It takes the form of a line by line commentary (in red) on the verbatim text of the paper, which is followed by some concluding remarks as to the paper’s utility in the current discussions on CDC in the UK.

The author is Con Keating , well known to this blog. The Ape is very much his animal. His remarks are in red.

Abstract

Both Collective Defined Contribution (CDC) and Collective Individual Defined Contribution (CIDC) schemes place any risks on pension scheme members instead of an external risk-bearer. This is a gross oversimplification. There is nothing to prevent any CDC scheme from buying insurance against any of the risks it faces.

In CDC schemes, assets are pooled collectively, allowing for risks to be shared between pension scheme members. This is incorrect. The pooling of assets in any collective arrangement is the acceptance, by the parties to it, of a common risk and reward profile. There is no sharing of asset risk among these members.

In Individual DC schemes (IDC), the scheme members bear such risks individually. But CDC’s collective nature leaves little room for individual risk management This latter sentence is untrue. The members of a CDC scheme have available to them a similar set of risk management opportunities as individual DC members.

and the pension assets are allocated to scheme members via rules that are often complex and ambiguous. This may be true in some cases but it does not have to be the case. CIDC schemes strive to retain the desirable aspects of CDC and IDC schemes, while improving on some of the drawbacks. The veracity of this statement depends critically on the definition of CIDC.

The drawbacks of a CDC scheme are mitigated by the introduction of

1) individually quantifiable pension pots through individual accounts, Without definition of an individual account, the content of this statement is difficult at ascertain.

2) individual risk management This is nonsense, The important aspect of CDC is that it should treat members equitably.

and 3) a simplified scheme. Simplicity may be desirable in some cases but it is neither necessary nor sufficient for a sustainable CDC scheme.

The drawbacks of an IDC scheme are mitigated by

1) mandatory participation, This is a complete nonsense. Compulsion is neither necessary nor desirable in CDC.

2) collective management of assets,

and 3) sharing of risks. It therefore seems that CIDC schemes have a number of important advantages over CDC schemes. This assertion is a complete non-sequitur.

CIDC scheme members should be clearly informed of their legal position vis-à-vis their employer and pension provider, There is no need for any employer to participate. CDC schemes may operate as non-workplace pensions. The provider of course is the collective; for example, it may be a member mutual trust in which a member’s interest is clearly defined. and the contract should clearly define the risks. This can be achieved within CDC. Scheme members appear to benefit from individual risk management and individually identifiable pension pots, while employers and/or pension providers seem relieved from risks and enjoy the security of fixed pension contributions. Employers, when present, pay contributions for service rendered; that is all. The content of the first half of this sentence, because of its ambiguity, is impossible to judge.

The possibility to take out a lump sum seems contrary to the collective sharing of risks in both CIDC and CDC schemes. Lump sums are not a problem in CDC. Indeed, one of the unique properties of the form of CDC I have advocated is that the lump sum equates the monetary amount of the asset share basis of DC with the income share basis of DB. In general, transfers into and out of CDC schemes may be undertaken at any time at the net asset value of a member’s equitable interest.

Introduction

In the simplest archetypes, pension schemes can be divided into Defined Benefit (DB) and Defined Contribution (DC) schemes. They can be told apart easiest by remembering who – in principle – bears the risk: in a DC scheme, the scheme members themselves bear the risk; the employer commits merely to paying a fixed sum as a contribution. This risk classification is incomplete – DC is simply a tax-advantaged savings scheme while DB promises an income for life in retirement. This distinction finds expression elsewhere; for example, the tax-free lump sum of DB is 25% of the pension income expected while for DC, it is 25% of the assets. Comparison of the two is inevitably one of apples to pears,

Pension benefits are determined by what has been paid into the scheme and the investment yield, minus costs. A pure DC scheme – also known as Individual Defined Contribution (IDC) – features no risk sharing, apart from statutorily required elements of solidarity, and the sharing of investment risks that appears inherent to the collective management of the investments. The misconception of risk-sharing in collective schemes is repeated.  In a DB scheme, the risk bearer is typically the employer or a pension provider. This is incomplete. The ultimate risk-bearer is the scheme member. It arises from situations in which the sponsor employer is insolvency and the scheme in deficit relative to the cost of replacing the benefit promised. This is compounded by priority rules among classes of scheme member.

The employer commits to a certain level of benefits to be received by the employee upon retirement. In the Netherlands, the pension provider has a number of options to compensate for underfunding without increasing pension contributions, such as a benefit reduction or non-indexation. Non-indexation is problematic. It is profoundly inequitable among members. Even pensioners in payment have different expected periods of receipt of their pensions, corresponding to their life expectations, and with that bear different proportions of the loss.

In reality, few schemes conform to the description of either archetype, and can thus be placed on the spectrum that exists between them. On that spectrum, Collective Defined Contribution (CDC) and Collective Individual Defined Contribution (CIDC) schemes can be found. There really is no continuum between the two types of scheme. In the Netherlands as in other countries, a discussion questioning the sustainability of DB schemes and their pension promise is taking place, as well as on their complexity. See my response to the DWP’s DB Green Paper.

Defined Benefit schemes feature collective pension pots and the assets are managed collectively, making the identification of who owns what complicated. There is an embedded misconception here. The members of a DB scheme do not own the assets; they have an interest in the scheme. The analogy of the relation between a shareholder and the company’s assets is perhaps helpful. The link between benefits and entitlements in these schemes is not easy for a scheme member to understand, and these collective pots allow little room for tailor-made solutions to accommodate wishes and risk-appetites of various age-groups. As stated earlier, the members of a scheme do not own the assets. Their wishes, risk-appetites and desires are irrelevant as their benefits and entitlements are clear; they are to a particularly formulated income in retirement. Freedom and choice has to an extent muddied the waters here, but that is merely transfer of the asset equivalent of their interest to the extent that it is secured by asset coverage. Scheme members have limited control rights usually implemented through voting on selected matters.

CDC

As a way to limit the risks stemming from DB schemes for employers, Collective Defined Contribution plans (CDC) were created. These are hybrid schemes: a combination of elements of DB and DC. The premium is fixed for a number of years – a typical DC feature – and the level of pensions is typically defined in advance – a typical DB feature – with the explicit caveat that the level of pensions is guaranteed only to the extent that premiums paid are sufficient to reach that level of benefits. The risk that the premium is insufficient to achieve the pension benefits is borne by the scheme members collectively. There is confusion here between premium (contribution) and assets.

In the Netherlands, CDC schemes of two types exist. In the first, an assessment is made annually of the annuity that can be purchased on the basis of the contributions paid. Most CDC schemes in the Netherlands, however, are based on a career average salary like traditional DB schemes, but with the clear message that no guarantee as to the level of benefits is given. The employer seems not burdened by any legal or economic risk regarding the level of pension benefits achieved. In the Netherlands, CDC plans can also qualify as defined benefit schemes if there is a sufficient amount of certainty that the agreed upon level of pensions will be attained. Members of CDC schemes should be clearly informed of the manner in which the amount of the contributions have been determined and how likely it is that these premiums will achieve the indicated level of benefits.

In a Collective Defined Contribution scheme, the assets are pooled collectively. Such collective pooling allows for risk sharing between members “both within the same generation of members (i.e., intra-generational risk pooling) and risk sharing between different generations (i.e., inter-generational risk sharing).” As noted earlier there is no risk-sharing in collective investment of this type and it is notable that the paper continues not with an asset illustration but with mortality, where risk of different lifespans may be efficiently shared. For instance, if one scheme member lives longer than expected, the increased cost of that scheme member’s good fortune can be financed through another’s misfortune of living shorter than expected.

In contrast, in “pure” DC schemes, or IDC schemes, the scheme members bear the risks largely individually. In collective schemes without sponsor guarantees like CIDC and most CDC schemes, surpluses or deficits are not transferred back to the sponsoring undertaking or made up by it, but are rather shared by the scheme members collectively between young, old and future generations “by adjusting either contributions, or benefit levels or both, which leads to inter-generational transfers.” It is important that contributions are fairly priced, neither repairing past deficits, nor distributing surpluses Ex ante, the contributions are set in such a manner that a newly entering generation funds its own retirement, but ex post, it may turn out that a given generation is a net payer or a net receiver. What is omitted here is that the aggregate risk faced by a member is far lower than in individual DC. It is clear, then, that in comparison to a DB scheme, any risks are transferred to the scheme members. But the risks in DB are not borne by members equally. The risk for a high paid/high pension member in CDC may be less than that in DB, the loss relative to the PPF cap. This need not be a problem in itself, although the scheme members should be duly informed of their new economic and legal position. In the UK the legal position would be the same if they are both trust based. The economics could though differ.

Although CDC schemes allow risk sharing, their collective nature appears to have as a consequence that little room is left for tailoring such risk management to the needs of individuals. In my vision of CDC, a member could transfer at any time. A criticism leveled at such schemes is that the pension assets are allocated to scheme members via rules that are “typically incomplete [read: unclear] and often modified”, and the collective nature of CDC schemes makes determining an individual’s pension assets and risk-sharing arrangements ambiguous. The criticism may be levelled but it is not warranted.  The rules according to which these risks are shared can be complex and arbitrary, but they do not need to be so.

In addition, for a CDC scheme, the premiums are not determined individually but collectively.15 In the Netherlands, a system of so-called average premiums is used for DB and most CDC schemes, whereby all pension participants – regardless of their age – contribute the same percentage of their salary and receive a set percentage of accrual in return. This system of average accrual has been under discussion recently in the context of talks on pension reform,16 as it is said to be unfair to young employees whose contributions still have years of investment returns ahead of them, while those of older employees do not – yet under the current system, they are valued the same. This analysis is incomplete. At very low rates of expected return, the old in fact support the young, and in any event, the young may expect to grow old. “Valued” here is ambiguous; in scheme valuations a common discount rate is applied to all liabilities making the stated present values reflect the age of the member.

This system finds its origins in the post-war era, when it was necessary to enable older employees to build up a decent pension in a relatively short period of time.18 The Netherlands Bureau for Economic Policy Analysis (CPB) notes that this system is “not ideal or even problematic”, as it leads to a structural redistribution from younger to older scheme members, and from scheme members with low life expectancy to those with a high life expectancy. The first part of this sentence is not necessarily true and it fails to recognise that the young will grow old. This system hinders portability and labor mobility. As CDC may permit transfers at any point in time at the net asset value, neither of the assertions holds true.

CIDC

CIDC schemes strive to retain the desirable aspects of CDC and IDC schemes, while improving on some of the drawbacks. A CIDC scheme features a fixed premium and collective asset management (as in a CDC scheme), but with individual pension accounts. The individual account appears to allow for clear definition and individualization of the sum in the scheme member’s pension pot and the risks involved. In a system without average contributions (back-loading), the accrual of pension rights would be actuarially fair: the benefits correspond directly to the contributions and their investment returns. In this arrangement we have the same one to one mapping of contributions and investment returns as with individual DC – there is no risk sharing or pooling. The sole potential benefit would arise from economies of scale and scope in investment management.

The drawbacks to the IDC and CDC schemes are addressed in a CIDC scheme in the following ways. The drawbacks of a CDC scheme are mitigated by the introduction of 1) individually quantifiable pension pots through individual ‘semi-legal’ (see below) This appears to be absent accounts, 2) individual risk management through tailored investments, enabled by individual accounts This runs entirely contrary to the collective-wide risk pooling and sharing. and 3) a simplified (and therefore more understandable) scheme. I wonder how many will understand the implications of or ways to hedge the risks they face in CIDC. The drawbacks of an IDC scheme are mitigated by 1) mandatory participation, This is an appalling idea. Opting out of a pension scheme has been a right in the UK at least since the early 1990s – compulsion has numerous other deleterious effects on a collective scheme. Under compulsion, trust will not develop easily. The absence of exit (or non-joining) will lower the governance discipline and increase the likelihood of management abuse.2) collective management of assets, leading to scale economies and 3) sharing of risks, such as investment and certain longevity risks. Longevity risk is shared by redistributing leftover funds from scheme members who pass away early into a “collective pool”, to the benefit of surviving scheme members, just like in a CDC scheme, but in a CIDC scheme only risks are shared, that can be shared, such as micro-longevity (A gets to be older than B). It appears that annuitisation is also compulsory in this model. Furthermore, a CIDC scheme is fully funded per definition. This can only hold true if the entire proceeds of the accumulation phase are utilised in annuitisation by all.

The collective nature of DB and CDC schemes leaves fairly limited room for individual freedom of choice for scheme members: This is simply not true. The member has the freedom to transfer at any time, to arrangements which permit the desired actions.

“CDC plans pursue the same uniform investment policy for all participants, even though older participants would typically make a more conservative trade-off between risks and return than younger participants.” Embedded in this is the idea that members cannot tolerate market volatility at older ages – one of the principal attractions of CDC is that the collective may insulate older members from this while not foregoing investment return. In CIDC as described in this paper this advantage would be lost.

Contrariwise, younger scheme members could benefit from a more aggressive investment strategy. In a system in which more and more risk is shifted to the scheme member, a dearth of options to adapt the scheme to personal preferences could be problematic. I cannot understand this convolution.

The individual accounts in CIDC schemes make individual tailoring of investments possible. This also makes it possible to share only those risks that are appropriate to share, such as micro-longevity risks that occur within a certain collective rather than macro-longevity risk. It now appears that members of CIDC schemes may opt into or out of some aspects of risk pooling or risk sharing while remaining within the scheme. This introduces different classes of member;it also begs the question as to when this decision may be taken.

Even though IDC schemes can afford scheme members with more options for personal decisions, they do not insulate such members from the potentially far-reaching consequences of the many decisions they must potentially make in such schemes. Default options can provide solace in such situations, but the absence of risk-sharing in such schemes can place scheme members at risk of adverse developments in, for instance, financial markets or life expectancy. Future needs are difficult to anticipate and the future consequences of decisions made in the present are difficult to foresee: individuals are left to invest in the “unknown and unknowable”.

Conclusion

In addition, there appears to be a lower risk of inequitable transfers between generations. In fact, with CDC this risk can be entirely eliminated.

Since 2015, UK pension scheme members have the option of withdrawing the funds in their pension pot as a lump sum for DC schemes. The result of such a withdrawal is that those choosing the lump sum withdraw themselves from a pool in which risks can be shared. This conflates several aspects-the longstanding tax-free lump sum may be taken and the balance placed into a drawdown arrangement, which includes the possibility to take all, after payment of income taxes. The tax-free lump sum may be taken with the balance being used to purchase a life annuity of one form or another.

Other options – such as a fixed-term annuity or a drawdown arrangement – appear to allow for (limited) risk-sharing, but a lifelong, fixed-rate annuity does not appear possible without some form of external risk bearer. In fact, the member may structure their drawdown profile such that it resembles a life annuity, though of course the funds may be extinguished before the member dies. In the case of a flexible annuity or a fixed-term annuity, the risk can stay with the scheme members within their risk pool and an external risk bearer does not appear strictly necessary. In that case, the amount of the annuity can fluctuate based on the investment returns and life expectancy. Indeed, with risk-pooling and sharing this fluctuation may be minimised within CDC.

Perhaps the most important thing to recognise is that the individual member’s need to annuitize at retirement is intrinsically the same as the need to buy out pensions on sponsor insolvency. In one case with timing known and the other unknown.

However, opting for a lump-sum would mean not only an end to risk-sharing in the payout phase – in which the sharing of longevity risk seems most important – but it seems to be contrary to the idea of a CIDC scheme (and indeed a CDC scheme). Allowing members to withdraw the net asset value of their equitable interest at any time, by transfer or as tax paid cash, is perfectly sensible. It will tend to engender trust and confidence in the scheme and its collective benefits. Such schemes are meant to share such risks: if not, the first ‘C’ in the CIDC/CDC name will become meaningless. This is incorrect. It would still only reduce the collective risk sharing to that associated with the award process and the definition of a member’s equitable interest, if all members were to do this. The UK legislator could bar the possibility for taking out a lump sum for CIDC or CDC schemes. They could but this would lower income tax receipts.

In the Netherlands, the qualification of CDC and CIDC schemes is not always clear.

The Dutch Central Bank (DNB) requires those CDC schemes which are based on an average career salary to comply with the same funding requirements as a ‘conventional’ average salary DB scheme, which – in principle – guarantees a level of pension benefits. DNB states in its guidelines that the maximum premium for CDC schemes should not be fixed for a longer period than five years. This is in contrast to a DC scheme, for which the DNB allows a fixed premium for an indefinite period. No explicit guidelines appear to exist for CIDC schemes, but it seems that DNB’s approached could be extended to these schemes as well. These rules have little merit for a CDC arrangement and should not be imported to the UK. This section is discussed further in concluding remarks

We would like to conclude with a caveat. In the Netherlands, the discussion on the revision of the pension system to a certain extent revolves around creating a system that allocates “clear and individual property rights” to pension scheme members, and CIDC schemes are favored by some because they are said to be able to provide such rights. In the UK members have well-defined property rights in trust based DB and DC. However, such terminology is confusing as the assets in the scheme do not belong to the scheme member as such. In a DC and CIDC scheme, it is possible value an individual’s pension pot, but that is not the same as saying that the participant owns that sum. It appears that their Individual Accounts of CIDC are in some ways similar to the equitable interest of a member in the scheme, though not as fully or completely developed. Their use in the management of an equitable balance among members appears to have passed unconsidered.

Concluding Remarks

If this paper is to serve any purpose in the UK context, it is to illustrate why we should not pursue here a Dutch approach to CDC. The paper exhibits a number of important misunderstandings; most notably over the risk properties of collective investments. The key proposal in the paper is the introduction of individual accounts into CDC. It becomes clear only by inference that these are simply replications of the individual accounts familiar from the world of DC savings. The paper does not develop these in any meaningful way (see earlier comment)

The paper proposes the introduction of compulsion for membership and schemes from which exit is extremely difficult or impossible. This is certainly not necessary and runs against the grain of UK pensions thought. The ability to opt out as well as the exercise of access rights under freedom and choice are valuable to individuals. They also tend to engender trust and confidence in the scheme and system.

The Dutch Central Bank (DNB) requires those CDC schemes which are based on an average career salary to comply with the same funding requirements as a ‘conventional’ average salary DB scheme, which – in principle – guarantees a level of pension benefits. This framework of regulation fails to recognise that CDC differs from traditional DB and imposes upon it that regulatory regime. The demise of traditional DB was in large part caused by its regulation. In other words, it would be totally counter-productive.

DNB states in its guidelines that the maximum premium for CDC schemes should not be fixed for a longer period than five years. While there are a number of possible motivations for such a rule, it appears that the most likely is that the authors of the rule saw (new award) premiums (contributions) as a potential source of deficit repair. This leads directly to problems of intergenerational inequity. This is in contrast to a DC scheme, for which the DNB allows a fixed premium for an indefinite period.

The single most important take-away would be not to ape the Dutch model. That is fraught with problems which can be avoided.