For their own reasons, usually venal, objectors to the implementation of collective defined contribution have managed to surround the concept with a fog of obfuscation and confusion. This has only been possible because of one of the great strengths of CDC, its flexibility and breadth of application. CDC can accommodate many different designs and structures, but there are also many that it cannot, or should not. Bogus claims abound – inter-generational theft, compulsion is necessary, freedom and choice will be lost, these are Ponzi schemes, and many more.
If we are to bring some clarity, the Department of Work and Pensions description of CDC[i] is a good place to start:
“Although it is possible to design collective pension schemes which also share risks with the employer, a Collective Defined Contribution scheme is a scheme in which all the risks are shared amongst the members, and the employer has no ongoing liability and therefore no possibility of assuming a liability should a deficit accrue in the pension scheme.”
This clarifies the critical point for employer sponsors, that of risk exposure. However, the reference to employer somewhat obscures the fact that a sponsor of these schemes need not be an employer; indeed, they may be non-workplace arrangements, and in turn, this means that they could accommodate the irregular working and income patterns of the “gig” economy, and the self-employed. These are defined contribution (or money purchase in UK terminology) schemes for which there is already a satisfactory regulatory regime in effect. The question of whether any further regulation is needed revolves around and reduces to their collective nature.
Reverting once more to the DWP description:
“There are many ways in which to design a Collective Defined Contribution scheme, …”,
which has already been noted, but the description then goes awry with:
“… but essentially these schemes allow the gains from years in which investments perform well to be allocated to years when investments fare less well.”
The returns earned on scheme assets in any year are the same for all members; there is only one return on scheme assets. The idea of reallocating gains from year to year is rooted in the former practices of with-profits policies and bonus payments. Properly designed CDC should not contain such reallocations, discretionary or otherwise. Members’ claims on scheme assets, their interest in the scheme, can and should vary over time, as well as with the risk-pooling and risk-sharing arrangements of the scheme and its experience. This is an arrangement among members. It determines the net asset or capital value of their interest in the fund. Transfers out of the scheme may be effected at any time without detriment to the accrued interest or net asset value of other members of the scheme.
There are numerous ways in which members may agree that their interests could or should be defined. It is important that this definition should be agreed at or before any contribution is made. It may take the form of no risk pooling or sharing, which is the situation with a traditional DC fund. Such an arrangement offers no incentive for members to join or to remain in the scheme; it stands or fails purely on the basis of the (common) investment performance of the fund over the period in which a member’s funds are invested. It happens that there are many better ways for members to define their interests; the usual award basis (accrual) of traditional DB is particularly attractive. The uniformity of award for a period of service, irrespective of the age of a member is a powerful risk-sharing mechanism, a form of mutual insurance.
Returning once more to the DWP description:
“This means that while collectively savers have a more stable experience, individual savers could have been better or indeed worse off in a traditional Defined Contribution scheme.”
This is somewhat misleading. Two funds with similar asset allocations will produce similar results over the same periods. Saving into a common investment fund is the acceptance of the return and risk profile of that fund. Greater stability of outcome than may be achieved by an individual investing personally arises from the ability of a collective fund to be more highly diversified than that individual’s might be. Higher returns from a CDC fund than a standard collective arrangement stem from the longer horizon over which that fund may be invested in risky assets. A collective fund does not usually de-risk as retirement approaches; it does not know the circumstances of its members. Members actually need to sell this vehicle progressively as retirement approaches.
A CDC fund is not inherently more stable than a tradition DC mutual fund. It is the effects of risk-sharing and risk-pooling on a member’s interest in that fund which generate greater stability for the member’s net asset value. There is a structurally induced smoothness in the member interest in CDC funds. Another way to view this is to consider the member’s interest as a form of shadow liability. In the case of traditional DC funds, the member experiences the volatility of the fund during the accumulation phase, and in particular the point in time outcome at retirement, and is then faced by the volatility of either annuity purchase or drawdown in the decumulation. In CDC the post retirement decumulation is included within the member’s interest.
The accrual rate of the member’s interest before and after retirement is the same. So, there is a fund with a higher return arising from the longer-term allocation to risky assets and a member’s interest in that which is structurally smooth – indeed over the lifetime of a CDC member, if there are no risk management interventions in interests arising from the sharing and pooling arrangements, the member faces the volatility of the fund over the individual’s full lifetime. The shift is from point in time volatility of a fund to a long-term moving average, driven by the decumulation process.
Yet another way to view this is that the member is a CDC scheme faces only the volatility of the amount he or she is drawing at a point in time – and the risk-sharing arrangements of the scheme may mitigate even that.
There is further confusion evident in the continuation of the DWP’s description:
“In addition to this stability in the accumulation phase, most would agree that a Collective Defined Contribution scheme would also attempt to provide an income in retirement, again through spreading risks (inflation, investment and longevity) across members.
Having already addressed the greater stability question earlier, it remains only to note that traditional DC funds face greater variation than CDC in outflows arising from the size of member withdrawals – at retirement, it is the entire amount of a member’s fund which is withdrawn, whereas with CDC, it is annual pensions which are drawn.
“…As with the accumulation stage, some members could generate a better income in retirement through other options available to them as any payout from a collective scheme will represent the smoothing of the best and the worst of outcomes.”
This is problematic. Some members could generate better outcomes through other options – there are winners and losers in the lottery of financial markets. However, it is clear that the average under CDC will be materially higher than under individual DC. For the reasons expounded earlier, it is also likely to be less volatile. However, there is no smoothing of outcomes. For any period, there is only a single outcome. There is no smoothing of the type indicated.
It is perhaps as well to understand that the investment objective function differs between a collective fund bought by an individual and the collective fund of a CDC scheme. In the individual case the objective is to maximise returns and asset values at all times; an objective which is intrinsically short-term in nature. The CDC fund by contrast has the objective of equalling or beating the implicit target rate (elsewhere referred to as the contractual accrual rate). With the risk-sharing inherent in the scheme design, it is sufficient to do this on average. For this reason, we may expect to see considerable divergence in asset allocations between these two forms of collective fund.
In addition to ex ante arrangements which share risk among members, further risk sharing among members is optimal if we wish to reduce the possibility of cuts to pensions in payment. It is important that these arrangements maintain the equitable structure of member interests arising from the ex-ante arrangements. There are some mechanisms which will not do this, including that in use in the Netherlands, cutting indexation. In most cases this operates only upon the pensioner community, with active and deferred unaffected; it also places the greatest cost on those pensioners with the longest remaining life expectancies. There is a further potential complication with reinstatement of such a cut.
Returning to some of the bogus claims: provided equitable balance is maintained between members, the scheme should be sustainable and new members would even have a positive incentive to join the scheme when it is in “deficit”. There are no intergenerational subsidies or transfers, any and all support among members is recognised and compensated. Compulsion is completely unnecessary and would erode many of the benefits of a co-operative mutual arrangement such as CDC. This arrangement is beneficial to all members, with a value which should attract and retain members.
Some have asserted that the flexibilities of Freedom and Choice will be lost. This is simply untrue – CDC schemes can accommodate, at any time, transfers out at the net asset value of a member’s interest without harm. CDC funds may well apply new contribution monies to the payment of pensions, but this does not make them Ponzi schemes, rather it means that they are being efficiently run in the sense of minimising transaction costs in the fund.
While an income in retirement is the ambition of a CDC scheme, it could also offer all of the flexibilities of drawdown – the retirement income does not have to be some form of life annuity. Indeed, annuity type structures may be considered as specific cases of drawdown.
Finally, it should be recognised that if any of the risks faced by the scheme, or indeed cuts to be imposed, become unacceptable to members, the scheme may hedge them commercially or dissolve the scheme returning the net asset value of members’ interests to them.
Many have argued that there should be risk management buffers maintained by CDC schemes – along the lines we observe in Holland. This is nonsensical. The schemes assets are the common property of the members and all are available for the purposes of the members under scheme rules. There is absolutely nothing to be gained from introducing an arbitrary partition of the scheme and member assets. Holding such buffers would greatly complicate the equitable distribution of the scheme’s assets to members. In fact, it is the risk sharing and pooling rules which substitute for such buffers. It is these rules which expressly mitigate against untoward variations in the price of the asset portfolio, and deliver smooth and predictable outcomes for pensioners. It is these rules which make possible the average objective for the asset portfolio performance.
The continuing challenge for readers is distinguishing the bogus, usually objecting, claims from the real benefits and flexibilities of CDC.
[i] DWP evidence submission to the Parliamentary Work and Pensions Committee’s inquiry on CDC – January 2018