The Two-Component Retirement System, introduced effective 1 September 2024 allows members of retirement funds to take some retirement savings out of their defined benefit or defined contribution fund without leaving employment or the fund. The amount that can be encashed (once every tax year until retirement) is the amount in the savings component which consists of a seeding capital of 10% (subject to a maximum of R30 000) of the vested benefit at 31 August 2024 and one third of the benefit accrued post 1st September 2024.
For a Defined Contribution (DC) fund this is a relatively simple calculation (although the maintenance of the components is not). The benefit now consists of a vested pot, a savings pot and a retirement pot. The investment risk remains with the member. Members will be able to encash up to one third of their future contributions without having to withdraw from the fund (although this will certainly be disadvantageous from a tax perspective – savings withdrawals are taxed at marginal rates). To the extent that members do this, they will be
poorer at retirement. It remains to be seen whether this will be a feature of managing a DC fund, and whether funds will adopt a more conservative investment strategy for these savings pots – to the extent that they do, this could also result in the risk of lower growth for the member.
Defined Benefit (DB) funds will face a much more difficult transition, partly because of how these components are defined and linked to pensionable service. This makes the calculation of components and benefits tricky and therefore requires more actuarial input.
The FSCA has issued communication 16 OF 2024 (RF) in which it is stated that if a part of the savings component is paid to the member prior to retirement, the pensionable service of the member needs to be reduced to make this payment financially neutral to the fund at the time of payment. DB Funds offer a promised pension at retirement and are required to pay a benefit on early exit, which is determined as per the rules of the Fund, but is subject to a minimum referred to as the Minimum Individual Reserve (MIR). This is based on the assumptions set in the latest statutory actuarial valuation and incorporates the net discount rate published by the FSCA at each month, based on the basis adopted by the Fund. The pension is payable for the lifetime of the member and often the rules will include a provision that should the member pass away a portion of the pension will revert to the spouse. Actuarial assumptions used to place a value on the benefit therefore include the discount rate referred to above, as well as demographic assumptions such as the percentage of members married at retirement and the age of the spouse and mortality applicable.
DB Funds operate on the basis of cross-subsidisation and have a sponsor who assumes the underlying risks. When a calculation is determined to be “financially neutral” to a DB fund, this can only be “ïf the assumptions are borne out in practice” – which will not happen exactly. The calculation is based on assumptions making a “best estimate” of the future, and encashing a portion of service prematurely could impact the costs of the sponsor. The ultimate costs will always be determined by the actual experience of the fund – longevity, investment returns, salary increase rates and other demographic factors. Anti-selection in the context of a defined benefit arrangement could occur when an individual has information that the fund doesn’t have, or there is an advantage to the member, because of the way of the operation of the DB Fund.
The Two-Component Retirement System is intended to balance a need for “emergency cash” and encourage compulsory savings for retirement. In doing so, there may be increased risks, both in the DB and the DC environment.
This can manifest when calculating the savings withdrawal benefit which can be encashed on an annual basis. If the actuary makes assumptions about long term trends in economic and demographic data. and the member is aware that the assumptions in the calculation are favourable relative to the current economic environment, the member could make a large cash withdrawal. If many members do this on a regular basis, the fund would be paying away up to one third of the savings intended for retirement and this could result in a financial strain on the fund.
Another opportunity for anti-selection in a DB fund is because of demographic factors. At retirement the service in the savings component may be commuted for a lump sum. The value of the commutation will be determined by the fund valuator. Often commutation factors are applied based on the most recent actuarial valuation, and the pension commuted is based on a single life annuity. The value of the pension in the retirement component will be paid as a pension and may include a reversion to a spouse for married member, which is often adjusted to assume that no portion is commuted. As an example, if the reversion is 50% this would be adjusted to 75% to allow for the reversion of the (1/3rd) amount commuted for a single life /annuity. Overall, the benefits at retirement will not change for members. They will just be paid from various components.
Anti-selection against the fund is possible in the following way: Commuting a portion of the pension for a single life annuity would likely be lower than the MIR which could include a reversion to the spouse, based on the assumptions that a percentage of members will be married at retirement and the assumed age difference. This is especially true for unmarried members where the balance of the pension will have no reversion. Instead of commuting members could elect to take the saving withdrawal benefit and the sponsor would be required to fund any shortfall. The risk of this may be low as tax benefits on commuting at retirement may act as a deterrent, but if members do have other retirement products the tax-free lump sum benefits could be applied there.
A risk for both DB and DC funds to consider is the liquidity and investment risk. The quantification of voluntary benefits payable is unknown. Funds would need to have sufficient liquidity at all times which could impact on long term investment objectives. DB members could encash the service at times when relevant yields used for discounting are low which could result in a larger benefit and a financial strain for the Fund. Certainly, the costs of providing an annual opportunity to members to withdraw savings will increase the administration and actuarial costs associated with the two-component system.
The Two-Component Retirement System is intended to balance a need for “emergency cash” and encourage compulsory savings for retirement. In doing so, there may be increased risks, both in the DB and the DC environment. DC members may have lower retirement benefits, and DB Funds could potentially be paying more than the actuarial reserve held for a member and this could financially impact the sponsor and required contribution rates. It will be interesting to observe the long-term impact on the industry as trustees, members and their advisors navigate these risks.
