Our starting point must be the award process. Some combination of contribution, target benefits promised are chosen given the trustees’ expectation of long-term investment returns. The benefits awarded may be similar in design to those familiar from DB pensions or could simply be some elementary deferred annuity, a series of real or nominal payments in retirement. Many other benefits could feasibly be added, such as surviving spouses’ pensions or death in service benefits. Indeed, these schemes could offer further choice at the time of retirement, such as annuitisation and drawdown variants.
This should be a stand-alone process with no subsidy of prior deficits or enhancement of new award terms from the distribution of surpluses. The terms of award may or may not take account of the age of the member (and time to retirement) at the time of the contribution.
These terms define a member’s equitable interest in the fund. In the absence of other risk-sharing arrangements, pensions payable are restricted to the extent that the fund covers the total equitable interest. We should note that the accrued equitable interest may differ from member to member as a result of variation in the terms of awards over time.
The expected return on investments determines the rate of accrual of the benefits (pseudo) liability. In the absence of revisions to the parameters which determine the projected amounts of future pensions, this rate is fixed. (We have referred to this rate in other publications as the contractual accrual rate.) The overall rate for a scheme is the weighted average of all awards, over time and across members. It is a time consistent which means simply that the value of a liability derived by discounting projected benefits using this rate will be equal to the value arrived at by accrual.
As a safeguard against abusively generous new awards, it is possible to incorporate quarantine periods and rules during which these awards do not serve to dilute other members’ interests. As these are peripheral to the central theme of this note, mutual risk sharing, we shall not elaborate here. The converse case, entry of new members when the scheme is in deficit, is covered later under the description of risk sharing mechanisms or rules.
In the absence of any risk-sharing among members, alterations to pensions in payment could be quite frequent. For example, even if the long-term rates of return achieved by assets were the same as those envisaged by trustees, as the contractual accrual rate, there would still be (pseudo) deficits and surpluses arising from the animal spirits of financial markets. With pension payments limited to the asset coverage of the scheme – a scheme with a 20% (pseudo) deficit in any year could only pay 80% of that year’s targeted benefits. This variability could easily bring with it demands for lower variability of portfolio returns, and with that lower returns; a contrary investment strategy.
The first point to note is that these variations have no consequence if there are no pensions in payment; there is no need for any intervention. It is variation of pensions in payment which is undesirable.
The risk-sharing rules need to be fair, that is equitable among members. It should also be recognised that they are a smoothing mechanism, and they do not in themselves generate any additional revenues. They should however enable the near perpetual holding of higher yielding assets such as equity in the investment fund.
The problem with operating any type of risk sharing or smoothing mechanism is that we need to distinguish between signal and noise. The signal of concern in this case being that the long-term returns are genuinely below the contractual accrual rate, and the noise the animal spirits evident in markets.
Before considering, the mechanisms which may facilitate that distinction, we need to cover the payment of pensions when the scheme is in surplus. There are two variants possible here: 1) the pension paid may be limited to that targeted, and any surplus is retained within the scheme, or 2) the pension payment may be augmented by the proportion of the surplus. This is a matter for scheme rules.
As the risk-sharing rules operate on the equitable interests of members, and their totality, we should explicitly state that the equitable interest of a member represents their fair claim on the assets of the scheme. It is the accrued value of awards, which may include revisions to the projected benefits and any prior adjustments made as compensation for participation in risk-sharing arrangements. Risk-sharing ensures fairness among members by operating on the equitable interests of those members not receiving support. Risk-sharing takes place among members. The most obvious classification, whereby this may occur, is between pensioners in payment and other members, actives and deferreds. In the case of a scheme where there are only pensioners in payment this could be those with claims extending most remotely in time supporting those more immediate; this obviously would be the case for CDC schemes as a decumulation vehicle. It is also possible to design a support mechanism based upon the relative magnitudes of claims, and indeed some combinations of the two
In order to address the signal from noise distinction problem and to avoid catastrophic spirals into non-recoverability, two dimensions to the risk-sharing rules are needed, of time and amount. At this point we should state that the precise or optimal time and amount rules are a function of the membership make-up of the scheme, and notably the number of pensioners and the proportion of scheme assets needed for pension payments. The rules we describe below work well for a scheme with the sort of membership profile we might associate with an open DB scheme.
The magnitude rule is simply that the total amount of support made available to pensioners, since the last position of balance or surplus, should be limited, cumulatively, to 10% of the value of non-pensioner members’ equitable interests. The supporting non-pensioner members see their equitable interests augmented by an amount similar in proportion to the support provided to pensioners. If support would breach the 10% limit, pensions and the equitable interests of all members are immediately cut to bring the scheme back into balance.
The time rule operates independently of the amount rule. The time permitted for remedy and deficit cure, during which risk-sharing rules and support of pensioners continue to operate, is inversely related to the magnitude of the deficit. A 10% deficit needs remedy, which can come only from investment performance, within ten years; a 20% deficit within five years and a 50% deficit within two years. If this is not achieved within the time-spans set, the pensions and equitable interests are cut to bring the scheme into balance.
Simulation suggests that cuts to pensions in payment for a scheme which has 60% active and deferred members and 40% pensioners uniformly distributed in accrual and following the ONS national survival projections in retirement are likely to be rare, less frequent than the default or insolvency likelihood of a AA corporate imply.
Many have suggested that new members will be unlikely to join a scheme which is in deficit. This is a misconception inasmuch as new awards are priced by trustees to reflect their expectation of future investment returns, and following joining, these new members would have their interest immediately augmented by an amount reflecting the support received by pensioners.
The effect of support is to reduce the value of the residual asset portfolio by the amount of that support, but it is also, by way of the operation of the risk-sharing rules to increase the relative share of the non-pensioner claims on those residual assets. The effect of operation these risk-sharing rules is, ceteris paribus, to worsen the declared deficit. This should not surprise as the procedure has involved the payment of an otherwise unwarranted amount to pensioners.
It is important that the risk-sharing rules, whatever they may be, operate automatically; there should be no room for trustee discretion or other source of uncertainty.
In terms of disclosures, members may be informed of both the projected value of accrued benefits, as a retirement income and the present level of their equitable interest as well as the current level of (pseudo) funding.
