There is a line repeated in almost every article about investment costs: a 1%
annual fee will cost you about a third of your final pot.
It is not wrong exactly. It is missing the only word that makes it meaningful,
and without that word it roughly doubles the damage for a typical investor.
We computed it.
Rateweb analysis, 2026
What a 1% annual fee costs, expressed as the share of the no-fee outcome you
keep. Computed as (1 + r − f)ⁿ ÷ (1 + r)ⁿ.
Holding period
You keep
You lose
10 years
91.1%
8.9%
20 years
83.0%
17.0%
30 years
75.6%
24.4%
40 years
68.9%
31.1%
"A third" is roughly right at forty years. At twenty — closer to most
people's actual horizon — the real figure is about 17%, so the soundbite
overstates it by nearly double.
And the drag barely depends on returns. At ten years it is 9.0% if the
gross return is 6%, and 8.7% if it is 10%. Fee drag is driven by time, not
by performance.
The 6%, 8% and 10% rates used to test that are arbitrary illustrations. They
are not forecasts and this page quotes no performance figure of any kind.
The model is a lump sum held for the full period — the most generous case
for the popular claim. For somebody contributing monthly the real drag is
lower still, for the reason set out below.
Why the horizon is the whole answer
A fee compounds against you the same way a return compounds for you. That is
why the number climbs — 9%, 17%, 24%, 31% — rather than staying flat.
It also means the honest version of the advice is conditional. If you are
investing money you expect to need in eight years, a 1% difference in fees
costs you around 7%, not a third. That is worth having and it is not
catastrophic. If you are twenty-five and this money is for your sixties, the
same 1% is genuinely large.
The mistake is not caring about fees. The mistake is quoting a
forty-year number to somebody with a ten-year horizon.
That the drag is nearly independent of the return is the part almost nobody
mentions, and it is useful: you do not need to predict markets to work out what
a fee will cost you. You only need to know how long you are holding.
One caveat that makes the soundbite worse still
The table above models a lump sum left to grow. That is the arithmetic the
popular claim is built on, and it is the most generous case for the claim.
Most people do not invest that way. They contribute monthly or annually over
decades — which means the average rand is invested for considerably less than
the full period. A contribution made in year eighteen of a twenty-year plan
has been exposed to the fee for two years, not twenty.
So for a regular contributor, the real drag sits below even the figures in
the table. The forty-year row assumes a rand invested on day one and untouched
for forty years. Very few rands in a real portfolio have that history.
None of which means fees do not matter. It means the number people quote is
built on the most extreme case available and then applied to everybody.
What "1%" usually is not
The other reason the headline figure misleads: it is rarely one fee.
What an investor pays is typically several charges that sit in different
places and are disclosed in different documents — a fee for the product
itself, a fee for the platform holding it, sometimes a fee for advice, plus
transaction costs incurred inside the product that never appear on a
statement.
Three consequences worth understanding as mechanics:
A single quoted percentage may not be the total. A product fee quoted in
isolation tells you about one layer. The number that matters for the
arithmetic above is what leaves your money in total each year.
Costs inside a product are invisible by design. Trading costs incurred by a
fund are borne by the fund, so they reduce its return rather than appearing as
a charge to you. They are real and they do not show up as a line item.
Percentages and rand amounts behave differently. A fixed monthly rand
charge is punishing on a small balance and trivial on a large one; a
percentage is the reverse in felt terms — constant as a share, but growing in
rands as the balance grows. Which is worse depends entirely on your balance,
and the crossover is worth working out rather than assuming.
The practical question is not "is this fee 1%?" It is "what is the total
annual cost of holding this, in one number, and over how long am I holding
it?" Those two answers together give you the table above. Neither is
available from a headline percentage alone.
The two South African allowances worth knowing
Neither is advice about what to buy. Both are statutory, both are from SARS,
and both are commonly misunderstood.
Tax-free investments
Annual limit, 2026 tax year
R36,000
Annual limit from 1 March 2026
R46,000
Lifetime limit
R500,000
Penalty on excess
40%
The 2027 tax year is already running, so R46,000 is the current annual
figure, not R36,000.
Three features people get wrong:
The lifetime limit is the real constraint. At R46,000 a year it takes
about 10.9 years of full contributions to reach R500,000 — ten full years
of R46,000, then R40,000 in the eleventh. After that you cannot contribute
again, ever.
Unused allowance is forfeited. SARS: "any portion of unused annual limit is
forfeited." It does not roll forward. A year you skip is a year gone.
The penalty is severe. Exceeding either limit attracts a penalty in the
form of normal tax payable of 40% on the excess. This is not a fee — it is
tax, and it is why contributing to more than one product without tracking the
total is genuinely dangerous.
That last point is the practical trap: the limits are per person, not per
account. Two products, each accepting R46,000, produce an excess and a 40%
penalty on it.
The interest exemption
Annual exemption
Under 65
R23,800
65 and older
R34,500
SARS shows these as unchanged across the 2022 to 2027 tax years, noted as "no
changes" as at 25 February 2026.
Worth pairing with the fee arithmetic above for one reason: an exemption
that does not move is an exemption that shrinks. It is a rand amount, and
the rand has
lost 37.8% of its purchasing power between 2015 and 2025.
An exemption held flat across five tax years is worth measurably less at the
end of them than at the start, without any announcement.
What this page will not do
We are being explicit, because this is the part of personal finance where
vagueness does damage.
We name no fund, platform, broker, index or security. Not one, anywhere on
this page.
We quote no performance, historic or expected. The percentages in the fee
table are arbitrary test values used to show that the answer barely depends on
them.
We do not tell you what to invest in, or whether to invest at all. Not
directly and not by implication — including the implication that low fees are
the only thing that matters, which they are not.
We are not licensed to advise you. What suits you depends on your
timeframe, your obligations, your tax position and your tolerance for loss —
none of which a web page knows.
What this page does is arithmetic on published numbers. That is a genuinely
useful thing and it is a different thing from advice.
The three questions the arithmetic actually answers
"How long is this money invested for?" That single answer determines what a
fee costs you, far more than any return assumption does.
"Am I tracking my tax-free contributions across every product I hold?" The
limits are per person. The 40% penalty applies to the excess, and nobody
aggregates it for you.
"Is a rand-denominated allowance keeping up?" The interest exemption has not
moved in five tax years. Whether that matters to you depends on how much
interest you actually earn — but the direction is not in doubt.
For anything beyond that, a licensed financial adviser and
SARS for the tax treatment. Limits and exemptions
change, and the figures here are as published in August 2026.
How does this affect YOUR Money OS?
Fees and allowances are the two parts of investing you fully control — you
cannot choose your returns, and you can choose what you are charged and whether
you use an allowance before it is forfeited.
Does a 1% fee really cost a third of your investment?
Only over about forty years. Over twenty it is roughly 17%, and over ten about
9%. Quoting "a third" without a holding period overstates it by nearly double
for a typical investor.
Does the fee drag depend on investment returns?
Barely. At ten years the loss is 9.0% on a 6% gross return and 8.7% on a 10%
one. Fee drag is driven by how long you hold, not by how well the investment
does.
What is the tax-free savings limit in South Africa?
R46,000 a year with effect from 1 March 2026 — so that is the current figure —
against R36,000 for the 2026 tax year. The lifetime limit is R500,000.
How long does it take to reach the R500,000 lifetime limit?
About 10.9 years of full contributions at R46,000 a year: ten years of R46,000,
then R40,000 in the eleventh.
What happens if I contribute too much to a tax-free investment?
A penalty in the form of normal tax payable of 40% on the excess above the
annual or lifetime limit. The limits are per person, not per account, so
holding two products without tracking the total is the common way people
trigger it.
Does unused tax-free allowance carry over?
No. SARS states that any portion of the unused annual limit is forfeited.
What is the annual interest exemption?
R23,800 for taxpayers under 65 and R34,500 for those 65 and older, shown by
SARS as unchanged across the 2022 to 2027 tax years.
Does this page recommend any investment?
No. It names no fund, platform, broker or security, quotes no performance, and
makes no recommendation. It is arithmetic on published figures.
