The South African retirement fund market has for some time been mainly defined contribution-based, with members taking the risk if investment returns are less than are required to build up sufficient resources for retirement.
And studies find that members are not building up sufficient resources to maintain their standard of living after retirement.
It would be easy to link these two statements and deduce that poor investment returns are to blame for poor member outcomes, but this would be a fallacy.
There are many reasons why retirement is “not working”, and investment returns have not been one of the main culprits in recent years.
For some time now, we have been talking about the realities of a low-return environment, but that hasn’t been the case over the past 15 years when the FTSE/JSE SWIX earned a 13% annualised real return, the All Bond Index a 4% annualised real return and the FTSE/JSE SA Property Index a 14% annualised real return.
These are great investment returns and should have translated into huge fund credits that would enable people to buy annuities that would be sufficient for their retirement needs.
However, this has not been the case, with studies finding that only 6% of people can retire with sufficient resources to replace a high proportion of their salary after retirement.
Targeted vs actual net replacement ratios
A lot of defined contribution funds set up investment strategies that set out to target a net replacement ratio (NRR).
The NRR is the income after retirement that a member can expect to purchase using their fund credit, expressed as a percentage of their pre-retirement income.
While funds aim to target NRRs of 60% to 70% of pensionable salary at retirement, studies show that the average actual replacement ratio of retiring members is closer to 30%.
The graph on the right illustrates the challenge – younger members still have the opportunity to attain high levels of NRRs, but older members tend to have little chance of retiring comfortably due to their low level of accumulated benefits as they approach retirement.
Drivers of the NRR
The actual NRR that members achieve is driven by many factors, including the accumulation period over which fund credits build up, investment returns earned over the accumulation period, the member’s salary increases, their level of pensionable salary vs actual salary, the level of contribution toward retirement, the member’s age at retirement, the type of annuity that members buy at retirement, and the cost of this annuity at the date of retirement.
The longer the period that members add to their defined contribution fund credits, the higher the expected NRR.
But it is critical that members not only save toward retirement by adding to their retirement fund credit while they work for an employer, but also that they preserve this money in a preservation fund, or in the retirement fund of their new employer.
If, instead, members take their exit benefits in cash, they are then reducing their period of accumulation, which will have a massive reductive impact on their ultimate retirement benefit.
Similarly, if a member retires earlier than expected, then this reduces their accumulation period and also increases the cost of purchasing an annuity due to the longer period they will spend in retirement.
Thus, their NRR will decrease due to a combination of these two factors.
The level of contribution to retirement also has a significant impact on retirement benefits, with a one percentage point increase adding an estimated five percentage points to the expected NRR at retirement age.
The fact that some employers and funds allow members to choose different levels of pensionable rather than actual salary can mean that a high level of NRR based on pensionable salary translates into a much lower NRR versus actual salary.
Any projected benefit statements based on this overstated NRR will give information that will be misleading for members.
Over the last couple of decades, real interest rates have decreased, resulting in an increase in the cost of annuities to be purchased by retiring members.
Thus the fund credit required to purchase a desired level of income in retirement has increased.
Many funds that set up investment strategies with targeted levels of NRR before these interest rate decreases have had to review whether their NRRs are achievable, and potentially reduce them to be more realistic in a lower real interest environment.
Are any of us average?
Around 60% to 70% has been a traditional target level for an NRR, even in a defined benefit environment. It is based on contributions of the order of 15% to 17% of salary per annum toward retirement, with 30 to 40 years of accumulation.
While these accumulation assumptions may be appropriate for other, non-South African, environments where an “average” employee works for a company for most of their working lives and then retires in their 60s, the reality for the “average” member retiring from a fund in South Africa today could differ significantly from this scenario.
Retiring members are very unlikely to have worked for a single company for 30 to 40 years, but will probably have changed jobs several times.
Their contribution rate to their retirement funds will have varied significantly.
They may have to retire well before age 65 because their employment contract requires this, or because their employment is terminated early.
They may have faced several crisis periods where their only source of funds was their retirement savings, and they had to access these savings when leaving an employer.
But most significantly, the reason that these “average” assumptions will not be widely applicable is due to the vastly different economic circumstances that South African retirement fund members face, including the potential for members to be supporting a wide extended family, including parents and grandparents, siblings and children.
The 2018 Alexander Forbes Benefits Barometer refers to a diversity metric from a study prepared in 2003 by JD Fearon.
The metric measures ethnic fractionalisation as being “the probability that two randomly selected individuals from a country will be from different ethnic groups”.
Using this metric, Fearon’s study found that 17 of the 20 most diverse countries in the world are African and that South Africa ranks eighth.
With this level of diversity it is not surprising the “one size fits all” retirement and savings solutions will not be able to meet many South African members’ needs.
A different, South African way of looking at retirement and savings
The reality for many in South Africa is that they need to be encouraged to save as much as they can, when they can, but that they may also need to access these savings during crisis periods and to invest in their families, particularly in their children’s education.
And when they retire they may well be starting to run their own businesses, or working in a family business, or requiring support from the children that they have helped to educate with their savings.
The Benefits Barometer sets out the need for different solutions that are “increasingly flexible, individualised and targeted at specific needs” so as to deal with the complexity of challenges posed by South Africa’s diverse environment.
Because let’s face it – retirement planning and saving toward retirement is complex. Many retirement funds, and the asset managers that they appoint, have been focused on the delivery of the right investment solutions, but this is not sufficient to ensure that desired retirement outcomes are achieved for members.
Members face a vastly broader set of issues over the whole period of their retirement savings journey, and this reality needs to be taken into account in designing retirement benefit and savings schemes that are “fit for purpose” for the members that access them.
Janina Slawski is principal investment consultant at Alexander Forbes Investments.