Best Retirement Annuity in South Africa: 2026 Fund Performance Guide

What is a retirement annuity?

A retirement annuity is best understood as a structure rather than a single investment. You contribute money to a retirement fund, which invests it in one or more underlying portfolios that comply with applicable retirement-fund rules.

Two investors can both say that they have an “RA” while owning very different investments. One might use a low-cost index-oriented balanced fund; another might combine several active managers. Their providers, asset allocations, charges and investment results can all differ.

For someone without an employer retirement fund, an RA can be a principal vehicle for retirement saving. For someone already contributing to a pension or provident fund, it may provide supplementary saving capacity. Whether additional contributions are sensible depends on the wider plan, not just the deduction.

Tax benefits and the Two-Pot system

Retirement annuities have important structural tax advantages. For the 2026/27 South African tax year, qualifying contributions to pension, provident and retirement annuity funds are generally deductible up to 27.5% of the greater of remuneration or taxable income, subject to a R430,000 annual cap and the full section 11F calculation. The limit applies across qualifying retirement-fund contributions; it is not a separate R430,000 allowance for each RA.

Contributions that cannot be deducted immediately may generally be carried forward, subject to the detailed rules. Investment growth in the fund does not incur the same ongoing personal income-tax, dividend-tax and capital-gains-tax treatment as many directly held investments, although retirement benefits are taxed under the retirement-fund rules.

The Two-Pot system

Since 1 September 2024, new contributions are generally allocated one-third to a savings component and two-thirds to a retirement component. The savings component permits limited qualifying withdrawals before retirement, which are generally taxed at the member’s marginal income-tax rate. Any amount withdrawn also loses its future opportunity to compound.

Rights built up before the reform can sit in a vested component with different rules. It is therefore no longer accurate to say that no RA savings can ever be accessed before age 55. For the 2026/27 tax year, the threshold under which compulsory annuitisation is not required increased to R360,000. Retirement choices still depend on the components, vested rights and other statutory rules.

For more on access and the implications of withdrawal, read our Two-Pot retirement system guide.

Regulation 28 and why balanced funds feature so prominently

Retirement annuities do not offer unrestricted asset allocation. Regulation 28 imposes prudential investment limits intended to limit excessive exposure to particular assets and risks. Listed equity exposure is generally limited to 75%, while the separate prudential foreign-exposure limit applicable to pension funds is 45%.

Within these boundaries, managers still have considerable flexibility: they can take different positions in South African and offshore shares, bonds, cash, property and asset-allocation decisions. That is why funds in the South African Multi Asset High Equity category make a useful peer comparison — but being in the category does not mean every fund has the same risk profile or is available on every RA platform.

Nor does it mean every long-term RA investor should use the highest-equity balanced portfolio. Horizon, capacity for loss, existing retirement assets and the job that the capital must perform all matter.

How we compared balanced-fund performance

The tables below use the Morningstar data, with returns measured to 30 September 2026.

• Peer category: (ASISA) South African MA High Equity
• Three-year universe: 210 ranked share classes
• Five-year universe: 191 ranked share classes
• Ten-year universe: 136 ranked share classes
• Return basis: compound annualised percentage return, reported by Morningstar for the particular fund class, in fund base currency (rand for the selected South African funds)

One underlying fund can appear repeatedly because it has different fee or unit classes. We therefore show each underlying strategy only once in the top-five tables, using a more representative retail or higher-cost class where multiple versions of the same strategy appear. The Morningstar peer rank remains the unadjusted raw class rank, so gaps are intentional. The peer-group count is also not reduced after our de-duplication.

These are fund-class returns, not all-in investor outcomes. Class expenses and fee arrangements affect the reported series; extra platform, administration and financial-advice fees may apply outside the fund class. The export does not establish availability on every RA platform, and we have not adjusted the figures for a particular investor’s tax position or cash flows.

Rankings are historical. They are neither forecasts nor a reason to switch investments simply because the order has changed.

Top-performing balanced funds over 3 years to 30 September 2026

Granate Balanced leads the three-year group at 21.33% p.a. Among the five distinct portfolios shown here, the return range is more than 4.6 percentage points a year. PSG Investment Management Growth FoF and PPS Managed follow, with New Road Enhanced Opportunities and Prowess Balanced completing the de-duplicated top five.

These latter two names are a reminder that a multi-asset peer category covers more than funds with “Balanced” in the title. All five were classified in the same ASISA high-equity category in this export; that is not the same as confirming the investment is offered in every retirement annuity.

Three years can favour a particular investment style or market environment. A high ranking is an invitation to examine the investment process and risk, not proof of future leadership.

Top-performing balanced funds over 5 years to 30 September 2026

The five-year leaderboard looks different. Granate remains in first place at 16.28% p.a., but ABAX Balanced moves into second in our retail comparison, followed by Edgestone Prime Growth, PSG Investment Management Growth FoF and PSG Balanced.

ABAX demonstrates the share-class problem especially clearly. Cheaper classes occupied Morningstar positions two, three and four over five years. The A1 class used here returned 14.92% p.a. and ranked fifth in the raw 191-class universe. Counting those different classes as separate funds would have made the table less useful.

Five years tells us more about the durability of a strategy than three, but it is still a selected period with a particular starting point. Returns should be considered alongside how much risk was taken and whether the process remains credible.

Top-performing balanced funds over 10 years to 30 September 2026

Over ten years, the order changes again. ABAX Balanced A1 returned 11.40% p.a. and ranks third among the full Morningstar share-class universe — but first after our underlying-fund de-duplication and choice of retail class. Aylett Balanced, ClucasGray Equilibrium, PSG Balanced and Centaur Balanced follow.

The less expensive ABAX B3 and B1 classes ranked first and second in the raw export, with higher reported returns than A1. They represent the same underlying fund, not three different long-term winners. The A1 figures are therefore used consistently in this article.

Notice that Aylett’s three-year position is much lower than its ten-year position. A single rolling return window can tell a very different story from a longer one. This is why a list of past leaders should inform due diligence rather than substitute for it.

How passive and core balanced funds compare

One of the more striking features of the September data is how competitive several low-cost, rules-based and core strategies have been against a substantial active-management universe. These are comparison examples, not a homogeneous set of passive funds.

Satrix Balanced Index ranks 7th / 210 over three years, 19th / 191 over five years and 8th / 136 over ten years, an unusually strong relative profile across the three periods. Prescient Balanced ranks tenth over three years and thirteenth over ten years; Sygnia Skeleton Balanced 70 ranks twelfth over ten years. Nedgroup Core Diversified sits 35th over five years and 21st over ten.

By contrast, 10X Your Future ranks substantially lower within the peer group over the three- and five-year periods shown, and does not have a ten-year return in the September export.

These observations do not prove that passive management always beats active management. They do make it harder to dismiss well-constructed, lower-cost core strategies. The comparison also shows why fees, benchmark exposures, asset allocation and manager-specific risks need to be evaluated together rather than treating any one label as sufficient.

Some well-known active funds in context

Brand recognition and peer-group rank are not interchangeable. The same September 2026 data provides useful context for three established balanced-fund names:

Allan Gray Balanced is the strongest of these three by five-year rank, at 12th / 191. Coronation Balanced Plus and Ninety One Opportunity occupy materially lower positions in the three- and five-year comparisons, although their relative positions differ over ten years.

This is not a judgement on the current suitability or future prospects of any manager. It is a reminder that a familiar brand and a recent top-five ranking are both inadequate shortcuts for choosing a long-term investment.

What the performance tables really tell us

1. The same ASISA category contains a wide range of outcomes

Among the selected fund classes discussed here, the three-year annualised returns range from 21.33% for Granate Balanced to 10.17% for Ninety One Opportunity — a difference of about 11.2 percentage points a year over that period. The corresponding five-year examples range from 16.28% to 9.43%, and the ten-year examples from 11.40% to 8.46%.

That difference matters. But we know these historical outcomes only after the fact, and the tables do not identify who will lead next.

2. Leadership changes depending on the period

Granate leads the three- and five-year retail comparisons, while ABAX leads over ten. PSG Balanced appears in both the five- and ten-year top-five lists, but is thirteenth over three years. Aylett, by contrast, is fourth among raw ten-year classes and much further down over three years.

3. Smaller managers deserve investigation, not automatic selection

Several managers near the top are not the household names an investor might expect. That challenges an assumption that the largest provider must have the best investment strategy. Size, capacity, liquidity, business stability, team depth and succession are still relevant due-diligence questions.

4. Cost is consequential, but not the whole answer

Repeatedly paying unnecessary fees creates a compounding drag. The Satrix data demonstrates that lower-cost core approaches can also be highly competitive on reported return. But costs must be weighed against diversification, the investment process, appropriate risk and the total charges actually borne by an investor.

5. Return tables say little about the difficult periods

These tables do not show maximum drawdown, recovery time, downside participation or the consistency of the return. Two funds with similar ten-year returns may have delivered very different experiences — with implications for investor behaviour and retirement risk.

6. Share-class rankings can create false precision

ABAX is a vivid example of one underlying fund appearing in multiple ranking positions. A class-by-class league table is not automatically a league table of different strategies. Always identify the exact class, fees, investor eligibility and measurement period before treating a ranking as meaningful.

How to evaluate funds inside an RA

1. Start with the role of the money

A person accumulating retirement savings over several decades and someone close to retirement face different risks. Time horizon, required return, existing assets, cash-flow needs and capacity for loss should frame the decision.

2. Compare more than one measurement period

Three, five and ten years reveal different things. Shorter periods say more about what has recently worked; longer records may span several market environments. Neither is a stand-alone verdict.

3. Understand the source of performance

Was it asset allocation, equity selection, offshore positioning, currency exposure or a particular investment style? A return becomes more useful evidence when one understands how it was achieved.

4. Evaluate risk alongside return

Consider drawdowns, volatility, concentrations, liquidity and how the fund behaved when markets were under pressure.

5. Include every layer of cost

Fund expenses are not the whole price of an RA. Platform, administration and advice charges can be additional. A realistic comparison looks at the combined cost of owning and maintaining the strategy.

6. Think about manager and style diversification

Using several portfolios can reduce reliance on one investment philosophy, although simply adding more funds does not automatically improve diversification.

7. Consider Regulation 28 and the wider portfolio

Existing employer-fund savings, other retirement capital, discretionary assets, offshore exposure and future income requirements help determine the job the RA should do.

8. Avoid switching just because a rank changes

Ranking changes are prompts for investigation. They are not automatic buy, sell or switch instructions; understanding the underlying reasons matters more than the change in a league-table position.

Where an RA fits into the wider retirement plan

An RA can be an effective retirement-saving structure: qualifying contributions may attract a tax deduction, capital benefits from preservation rules and Regulation 28 provides a prudential framework.

Those advantages do not imply that every spare rand should go into an RA. Planning also requires accessible capital, emergency reserves, discretionary investments, other retirement vehicles and sometimes offshore assets with different liquidity and estate-planning characteristics.

The 45% institutional foreign-exposure prudential limit means that offshore exposure in retirement funds is not unrestricted. Other assets may therefore complement an RA, depending on the overall circumstances. This is a planning decision, not a blanket argument for or against retirement annuities.

At death, retirement-fund benefits are governed by retirement-fund death-benefit provisions rather than simply passing according to a will; beneficiary nominations matter but are not the entire legal process.

The better question is not simply “RA or something else?” It is which structure gives each part of the capital the right role in the bigger plan.