One all-in Retirement Annuity fee: Putting the whole bill in one number

The uncomfortable truth about retirement in South Africa - Rands and Sense by 10X [video]

We sit down with 10X Investment Consultant lead Andre Tuck and discuss the retirement savings crisis in South Africa. We also delve into living annuities, retirement annuities, TFSAs and everything in between. Read more

10X Investments publishes this series. It also sits inside it.                  

That has been true from the beginning, but this is the episode where it stops being in the background and becomes part of the test. So here is the number first, before anything else in this piece has a chance to soften it. On the same benchmark used throughout the series, the single all-in fee at 10X costs R842,000 over thirty years and leaves R6,092,000 at the end. That puts it third of five. The layered structure from episode four finishes about R500,000 ahead. The platform structure from episode five finishes about R251,000 ahead. Two of the four providers this series has taken apart are cheaper over three decades than the one publishing the results.

With that on the table, the same question asked of every other provider is asked again. What would you charge on R500,000, and what would that charge do over thirty years? For four structures, the answer runs to a list. For this one, the answer is short. Whether short is the same thing as good is what the rest of this piece is about.

The anatomy of the single all-in fee

Four episodes of this series have asked the reader to do arithmetic. Add the administration fee to the management fee. Model a charge that moves with performance. Price a fund the platform does not disclose. Work out what an advisor costs once one is appointed.

This structure asks for none of that. The retirement annuity is charged at 1.04% a year including VAT on the first R1 million, falling to 0.81% on the next R4 million, 0.58% on the next R5 million and 0.40% above R10 million. That single rate pays for administration, investment management, the platform, and the fund administration underneath it. 10X holds its own S13B fund administration licence, which is why work that appears as a separate line on a layered quote does not appear as one here.

Everything else is absent by design. The clean version looks like this:

That simplicity is real. It is also not the whole story. Three things sit around the single line, and each one changes how the number should be read.

• The benchmark investor is buying at the worst point on the scale. R500,000 sits inside the first tier, so it attracts the top rate from day one. The portfolio crosses R1 million in year eight on this model, and the blended rate softens from there. The projection below ignores that and holds 1.04% flat for three decades, as it does for every tiered structure in the series, which means the figure attributed to this provider is an overstatement rather than an understatement.
• Trading costs sit outside the 1.04%, except on the page where they do not. The fund page carries a fee table with a footnote saying trading costs of 0.12% to 0.18% a year apply, depending on the fund chosen. The retirement annuity FAQ page says the sliding scale starts from 1.04% including VAT and trading costs. One says trading costs are outside the headline. The other says they are inside it. Those statements do not sit comfortably together. The written response this series received in April 2026 points to the fee table, which puts the benchmark investor at roughly 1.16% to 1.22% all in.
• The other products carry their own prices, and one of them carries a date. The living annuity is charged at 0.86% a year on the first R5 million. The preservation fund was quoted in April 2026 below 0.70%, on a promotional rate running to 31 May 2026. Promotional pricing is a real saving for whoever catches it. It is also the one number in this structure with an end date, which makes it the one thing an existing investor would need to watch.

My retirement plan looked good in my head, but the numbers on the page disagreed

One day, finance journo Jan Bornman decided to actually write down the retirement savings numbers that had been kicking around inside his head for years. And the picture wasn't as pretty as he thought it would be. Read more

What one number costs over thirty years

Run through the same model as every other structure in the series, holding the headline rate flat and excluding separately disclosed trading costs, this structure produces the following.

Third of five. Two structures in this comparison take less and leave more. It’s worth noting that these are nominal rands and the thirty-year fee total is a plain sum of what is charged each year. In today's money, the gaps are smaller. The ranking stays unaffected, since every provider is modelled on the same basis.

The layered fee is the instructive comparison, because it is the closest thing this series has to a like-for-like rival. Neither charges for advice or performance. Both publish their fees without being asked. On paper they are the same proposition, and at year thirty they are R500,000 apart.

The reason is not disclosure. This structure discloses more thoroughly. Fold in the trading costs and it comes to roughly 1.16% to 1.22%, with nothing further to find. The layered provider's own quotation puts its Effective Annual Cost at 0.90% in year one, transaction costs included. One charges roughly thirty basis points less than the other. 

Thirty basis points is the sort of difference that sounds like nothing when a salesperson says it out loud. Left alone for three decades, it is the difference between the two figures at the bottom of that table.

Why I'm saving for my kids' university with a Unit Trust and not a Tax-Free Savings Account

A unit trust is more flexible than a tax-free savings account, and the tax burden might not be as bad as you think. Don't make the wrong choice when it comes to medium vs long term savings for your kids. Read more

The case for one number

There is a real case for this structure, and it does not depend on the fee being low. It depends on the fee being knowable.

• The quoted figure is close to the actual bill Every other structure in this series keeps something outside the headline. Here the list runs to trading costs, published to the basis point. That matters because investors routinely assume the fund's costs are already covered. In a 2021 State Street survey cited in Vanguard's October 2025 analysis, 47% believed exactly that, and fewer than a third said they fully understood what they were paying.
• Nothing is conditional on performance or on anyone's involvement A performance fee moves with the benchmark. An advice fee starts when an advisor is appointed and stops if they leave. Brokerage arrives with every trade. None of those apply here. The rate does move, but only with the size of the portfolio, and only downward as it grows.
• No performance fee means no unknown Episode three set out how complicated a performance-linked charge is to forecast, since the fee depends on returns that have not happened yet. This structure has no such component. Whether performance fees are well designed is an argument the reader can have elsewhere. Here it does not affect the bill.
• No one is paid on what you pick The RAND working paper cited in episode two reviewed the evidence and concluded that advisors steer clients toward worse outcomes when their own remuneration depends on it, and that clients can rarely tell when it is happening. That describes a risk this structure does not carry.
• The fee is easy to find Assessed on the same basis as the other four websites in April 2026, this provider’s fee disclosure was the most findable in the comparison. The platform provider’s retirement annuity marketing page, by contrast, claimed low fees and competitive rates without publishing a single figure. Findability is a low bar. Two of the five did not clear it.

That is the strongest version of the case. It is simple, legible, and there is nothing further to find. It is also where the easy part ends.

Retirement Annuity vs Tax-free Savings Account vs Bond: Where Should Your Extra Money Go?

Got extra money to invest? Compare three powerful options: Retirement Annuity, Tax-Free Savings Account (TFSA), or paying off your bond faster. Learn the pros and cons of each to make an informed choice for your financial future. Read more

Where one number stops being enough

The weakness of this structure is not that it is hard to read. It is that a single number feels complete even when it is not.

• It is not the cheapest, and the gap is half a million rand Everything in the case above is a claim about disclosure. None of it is a claim about price. On price, over thirty years, two structures in this comparison beat it.
• The returns are published before the fee comes off The March 2026 investment report shows the 10X Your Future Fund returning 12.1% a year over five years, before 10X's fee and after other expenses. The layered and performance-linked providers publish theirs net. So a reader comparing three fact sheets holds two numbers with the fee removed and one without, and the gap to the layered provider closes once that is corrected. The marketing rests on the same basis, with the peer comparison in the investment report measured before fees.
• The absence of an advisor is a cost as well as a saving The case above counts no advisor as an advantage, and for a confident investor it is. For everyone else it removes the part of advice worth the most. Vanguard's October 2025 analysis of 23,000 advice offerings puts the value of behavioural coaching at 100 to 200 basis points a year or more, meaning the work of talking an investor out of selling at the bottom. A retirement annuity blunts that risk by design, since the money is locked until 55 and cannot be traded freely. What it does not blunt is everything short of selling, from stopping contributions to switching at the wrong moment. Nothing in this structure does that job.

None of that makes the structure bad. It makes it a trade rather than a win. The number is easy to see. The compromises around it are easier to miss.

The investor experience

The strongest service record in this series belongs to the publisher. That does not make it perfect. It makes it worth reading properly.

The rating is 4.28 out of 5 across 177 reviews in the twelve months to April 2026, with a Net Promoter Score of 48 and ninth place in HelloPeter's Investments category. No other provider in this comparison clears 2.11. The sample sits between the 542 reviews behind the advice-led provider and the 18 behind the performance-linked one, which makes it a strong signal rather than a settled verdict.

Reviewers praise speed, clarity, and specific staff by name, which is worth noting for a provider with no advisors. Complaints fall into three groups. There is no mobile app and reviewers want one. Manual processes are slow, with fund switches and onboarding drawing repeated criticism. And a recurring theme concerns pressure and urgency in the sales approach, including one reviewer pushed to register before fees increased.

Every review gets an answer, which is rarer than it should be. The average wait for that answer is 148 hours. The layered provider, sitting on one of the worst ratings in this series, manages 14. Both figures are means published by HelloPeter, and a mean is easily dragged by a handful of very slow replies, so neither is what a typical reviewer waits.

In early March 2026, a reviewer reported the platform still showing the annual tax-free savings limit as R36,000 after it rose to R46,000 on 1 March. 10X acknowledged the error. That is the second time in this piece that published information has lagged the business, after the two fee pages that disagree about trading costs.

Transferring out takes six to eight weeks and costs nothing, which matches the performance-linked provider and beats the advice-led model's stated 180 business days by roughly seven months.

Who it’s for

This structure suits an investor who does not want to assemble a price from parts. Someone who does not want an advisor, does not want to choose a fund from an open shelf, and would rather have the decision narrowed for them than widened. The retirement annuity opens at R500 a month, so that investor does not need capital to start. The living annuity is a different proposition, with a R500,000 minimum, the highest in this comparison.

It is a less suitable fit for anyone who wants to hold something the provider does not manage, anyone who wants regulated financial advice included in what they pay, and anyone whose only test is the thirty-year cost.

What this structure asks of an investor is very little. What it charges for asking very little is more than two of its competitors charge for asking more.

How it compares

Here they are together for the last time.

This structure comes third on cost, first on service by a distance, and first on how much of the price it shows you before you commit. Three different competitions, and it loses the one most readers came for.

What six episodes actually measured

This series set out to compare five prices and discovered that it could not. The five providers do not disclose on a comparable basis, and no amount of careful reading fixes that. What could be compared was architecture, which turned out to be the more useful thing to measure anyway.

The result in one picture

The capital, the assumed return, and the thirty years are identical across all five. The only thing that changes is how each provider charges for it.

What each provider is actually selling

Nobody sets out to buy a fee structure. Everybody ends up with one.

What the series found

Six episodes, five providers, and five things worth carrying out of it.

• The R1.8 million is structural Only the charging structure varies across the five. National Treasury reached a version of this in 2013, modelling a saver who cuts recurring charges from 2.5% to 0.5% a year and ends up around 60% better off after forty years. Treasury notes the result barely moves when you change the return assumption, which answers the obvious objection to any thirty-year projection, this one included.
• Year one tells you almost nothing about year thirty The cheapest opening fee in this series finishes third. Treasury put it more starkly than we can. Take five per cent off every contribution as it goes in and the investor ends up with about five per cent less at retirement. The damage stops there. A recurring annual fee compounds for as long as the money is invested, which is why it does so much more.
• What a headline leaves out varies more than the headline itself Rates run from 0.55% to 2.88%. What sits behind them varies by more than that. Treasury named the mechanism thirteen years ago, describing how charges get shuffled between asset management, performance, platform, administration and advice until the total looks palatable. That is a list of the five structures in this series, written before we measured them.
• Price and service do not move together The cheapest structure here has the weakest service record in the comparison. The most expensive has the second-weakest. The mid-priced publisher has the strongest by a distance. Whatever a bigger fee buys, it is not better treatment.
• One number makes all five comparable, and almost nobody leads with it The Effective Annual Cost standard has required a single all-in figure in point-of-sale documentation since October 2016, split into investment management, advice, administration and everything else. Ask for it and the comparison stops being guesswork.

What to do with that

We asked five providers what they would charge on R500,000. One produced an EAC unprompted. The rest answered in their own units, and 10X sent a link to its online calculator. ASISA's standard treats a provider-produced EAC and a calculator the investor fills in themselves as different things, and only the first satisfies the obligation to make an EAC available on request.

So ask for it, on your own amount over your own horizon. A provider who cannot produce one quickly has told you something.

Publishing a clear number is not the same as changing what people do with it, and the research on that is blunt. When Beshears, Choi, Laibson and Madrian handed investors the SEC's simplified fund disclosure and real money to allocate, nothing changed. Asked afterwards what had mattered most, the subjects ranked past performance first and fees sixth out of eleven. Treasury reached the same conclusion in 2013. Publishing the number is necessary. Reading it is the part that is up to you.

Which leaves the part no table can do. This series could not measure thirty years of actual returns, because they have not happened yet. It could not measure whether advice earns its fee for any particular person, because that depends on the person. It could not tell whether a service record reflects a provider’s structure or merely its size.

It is worth being blunt about how little of the future is knowable here. Over the ten years to June 2025, S&P Dow Jones Indices found that 75% of actively managed South African funds trailed their benchmarks, and that 35% of funds did not survive the decade at all. Three in four lost the race. One in three left the track.

What can be measured is the cost, and the cost is knowable before you commit. Divers do not argue about how the water looks from the boat. They read the gauge before they go in.

Ask for the EAC before you decide.