Rand-Cost Averaging Explained, and Why Timing the Market Usually Fails

Rand-cost averaging is investing a fixed rand amount at regular intervals regardless of the price. If you put R2,000 into the same fund on the 25th of every month, you are doing it, whether or not anyone has given it a name.

The mechanism is simple arithmetic. The reason it works for most people is not arithmetic at all.

The arithmetic

Because you invest a fixed amount rather than buying a fixed number of units, a falling price automatically buys you more units and a rising price buys you fewer.

Say you invest R2,000 a month into a fund over four months, and the unit price moves R100, R80, R50, R80:

You have invested R8,000 and hold 110 units. Your average cost is R8,000 ÷ 110 = R72.73 — below the R77.50 simple average of the four prices, because more of your money went in at the low price.

That gap is the whole effect. It is real, and it is modest.

What it is genuinely for

The arithmetic is the smaller half. The larger half is that rand-cost averaging removes a decision you would otherwise make badly.

The alternative to investing monthly is choosing when to invest. In practice that means investing more when markets feel good and less when they feel frightening — which is buying high and selling low, arrived at honestly.

The gap between what funds return and what investors in those funds actually earn is well documented, and it is largely a timing gap. A monthly debit order that runs without your involvement closes most of it, because the hardest months to invest are precisely the months worth investing in.

What it is not

It is not a guarantee against loss. If the market falls and stays down, averaging in means you own more units at a lower price — but you are still down. It reduces the impact of buying everything at a single bad moment; it does not make a falling market profitable.

It is not mathematically optimal. If you have a lump sum to invest, the evidence generally favours investing it all at once, because markets rise more often than they fall and time in the market is doing the work. Spreading a lump sum over several months usually costs a little return.

That said, spreading it is often the better behavioural decision. Someone who invests a large inheritance in one go and watches it fall twenty percent may sell at the bottom, which costs vastly more than the small expected return they gave up. The optimal plan you abandon is worse than the good plan you keep.

Where it fits in South Africa

The natural vehicle is a monthly debit order into a low-cost fund, and the natural first home for it is a tax-free savings account: R46,000 a year of contributions, a R500,000 lifetime limit, and no tax on interest, dividends or growth inside it.

Two South African specifics worth holding in mind.

Contribution limits, not withdrawal-friendly. A tax-free account does not restore your allowance when you withdraw, so treat it as one-way money. It is an excellent home for a decades-long monthly contribution and a poor emergency fund. See tax-free savings accounts.

The rand cuts both ways. Averaging into a globally diversified fund means you are also averaging into the exchange rate. When the rand is weak your monthly amount buys fewer foreign assets, and when it is strong it buys more — the same mechanism, applied to currency. That is a reason to keep contributing steadily rather than trying to pick a rand level, which is a game professionals lose too.

A ten-year illustration

Take R2,500 a month for ten years — R300,000 contributed in total.

At a 9% average annual return, monthly contributions compound to roughly R480,000. The R180,000 of growth is not evenly earned: the contributions from the first three years do most of the work, because they have had the longest to compound. That is the practical argument for starting sooner with a smaller amount rather than waiting until you can afford a larger one.

Now change one thing. Suppose the same investor stops contributing for the eighteen months of a market decline — a very common response — and resumes once things "look better". They have missed the cheapest units of the entire decade, and typically resume at higher prices. The end figure lands materially below R480,000, even though total contributions barely changed.

The single largest determinant of the outcome is not the return, the fund, or the timing. It is whether the debit order kept running.

Averaging out, at the other end

The same logic applies in reverse when you eventually need the money, and it is rarely discussed.

Selling a large holding in one transaction exposes you to whatever the market happens to be doing that week. Where you have a known future need — school fees in three years, or the start of retirement — moving gradually out of growth assets into cash over the preceding couple of years does the same job as averaging in: it removes the single-date risk.

At retirement this matters most, because a sharp fall in the first years of drawing an income does lasting damage that a later recovery cannot fully repair. You are selling units to live on at exactly the wrong prices. Shifting a portion to lower-volatility assets before you need it is not market timing; it is matching the asset to the date the money is needed.

Making it work in practice

1. Automate it. A debit order the day after payday. The point is that it does not require a decision each month.
2. Keep the costs low. Fees are the one part of your return you know in advance. A percentage point of annual cost compounds against you exactly as returns compound for you.
3. Do not stop when it falls. Those are the months buying the most units. Stopping during a decline converts the mechanism into its opposite.
4. Increase it with your income, not with your confidence about markets.
5. Leave it alone. Check quarterly at most. Frequent checking produces action, and action is usually what costs money.
6. Fill the tax-advantaged space first — an employer fund with a match, then a retirement annuity if your marginal rate is high, then the tax-free account.

When it is the wrong tool

Rand-cost averaging assumes you are investing in something you intend to hold for many years, and that you can leave it alone through a decline.

It is not appropriate for money you need within three years — that belongs in cash or a fixed deposit, where the value does not move. It does not rescue a poor choice of investment: averaging into something that keeps falling for structural reasons buys more of a problem. And it is beside the point while you carry expensive debt, where paying off a balance at 21% is a certain return no market can promise.

Frequently asked questions

Is rand-cost averaging better than investing a lump sum?

Statistically, usually not — lump sums are more often ahead because markets rise more than they fall. Behaviourally it is frequently better, because it is the approach people actually stick to.

How much should I invest each month?

An amount you can sustain through a bad year without stopping. Consistency matters more than size, and a contribution you have to cancel is worse than a smaller one you keep.

Does it work with individual shares?

The mechanism does, but the risk is different. A diversified fund can recover from a decline; a single company may not. Averaging into a falling share can simply mean owning more of a failing business.