What drift actually does to you
Say you decide on a mix of 70% shares and 30% bonds. That is a deliberate choice about how much risk you are willing to carry.
Then shares have a strong three years while bonds go sideways. Nobody has done anything, but your portfolio is now roughly 82% shares and 18% bonds. Your risk went up by a meaningful amount, silently, and you never made that decision. When the next correction arrives, you take a materially bigger hit than the one you originally signed up for.
| Target | After a strong run in shares | What it means | |
|---|---|---|---|
| Shares | 70% | 82% | Much more exposed to a crash |
| Bonds | 30% | 18% | Far less cushion when it comes |
| Decision made? | None. The market changed your risk profile for you. | ||
Drift is not limited to shares versus bonds. It happens between regions when one market outperforms, between sectors when one runs hot, and inside a share portfolio when a single holding doubles and quietly becomes a third of everything you own.
What rebalancing is really for
It is risk control, not return enhancement. In a long bull market, rebalancing will usually cost you a little return, because you keep trimming the thing that keeps rising. What it buys you is a portfolio that behaves the way you expected when markets turn - and an investor who does not panic, because the drop was within what they had planned for.
The two methods
Calendar rebalancing
Pick a date - annually, or every six months - and on that date, adjust everything back to target. Simple, predictable, and it works. Annual is fine for most people. Monthly is unnecessary and expensive.
The strength of this method is that it removes judgement entirely. You are not deciding whether now is a good time; the calendar decides.
Threshold rebalancing
Set a tolerance band - commonly 5 percentage points - and act only when something drifts beyond it. A 70% share target with a 5-point band means you do nothing until shares hit 75% or 65%.
This responds to what markets actually do rather than the calendar, so it tends to require fewer trades in quiet years and act promptly in violent ones. The trade-off is you have to check periodically, which for some people means checking constantly.
| Calendar | Threshold | |
|---|---|---|
| Effort | Once or twice a year, then forget it | Requires periodic checking |
| Discipline needed | Low - the date decides | Higher - you must not act early |
| Responds to crashes | Only at the next scheduled date | Promptly |
| Best for | Most long-term investors | Those already reviewing regularly |
A blended version works well: check once or twice a year, but only actually trade if something has drifted beyond your band. You get the discipline of a schedule and the trade efficiency of a threshold.
The cheap way: rebalance with new money
Here is the technique that matters most, and it is the one beginners rarely hear.
Instead of selling what is overweight, buy more of what is underweight.
If you contribute monthly, direct those contributions toward whichever asset has fallen behind its target, rather than splitting them according to your original percentages. Over a year or two, this pulls your allocation back toward target without a single disposal.
Why this is so much better
- No capital gains tax. You never sell, so nothing is disposed of, so no CGT event occurs.
- No selling commission. You were going to buy something anyway.
- Fewer trades overall, which means lower total costs and less to get wrong.
- It happens automatically if you simply adjust your debit order allocation once or twice a year.
Contribution-based rebalancing is slower - it will not fix a large drift quickly - and it only works while you are still adding money. Once contributions are small relative to your total portfolio, or once you are drawing down rather than contributing, you will need to trade. But for anyone in the accumulation phase, this should be the default method and selling should be the exception.
Keeping the cost down when you do have to trade
- Rebalance inside tax-sheltered accounts first. Trades inside a TFSA or retirement annuity do not trigger capital gains tax. If you hold the same asset classes in both taxable and sheltered accounts, do your selling in the sheltered one and leave the taxable holdings alone.
- Use your annual CGT exclusion. South African individuals have an annual capital gain exclusion. Spreading a large rebalance across two tax years can keep gains within it. See how investment tax works in South Africa for the detail.
- Rebalance broadly, not precisely. Getting from 82% back to roughly 72% captures nearly all the benefit. Chasing exactly 70.0% costs extra trades for essentially nothing.
- Redirect dividends. Rather than automatically reinvesting dividends into the holding that paid them, use them to top up whatever is underweight.
- Do not rebalance a tiny portfolio. If a 5% drift on your balance amounts to a few hundred rand, the trading cost outweighs the benefit. Just direct your next contributions sensibly.
The part nobody warns you about
Rebalancing correctly feels wrong every single time. It means selling the thing that has been performing beautifully and buying the thing that has been disappointing you. Every instinct says do the opposite.
It feels worst precisely when it matters most. During a crash, rebalancing means buying more shares while headlines announce the end of the world - which is exactly when it does the most good and is hardest to do.
This is why the rule must be written down before you need it. Judgement made during a crash is not judgement, it is fear with a spreadsheet.
If following your own rebalancing rule during a downturn feels genuinely unbearable, that is not a discipline problem - it is a signal that your target allocation is too aggressive for you. Fix the target, not the discipline. Our guide to allocation over time covers how to set one you can live with.
A simple rebalancing policy
Write this down somewhere you will find it again:
- My target allocation is: ___% growth assets, ___% defensive assets, with ___% offshore.
- I review it: every [six / twelve] months, on [a fixed date].
- I act when: any asset class is more than [5] percentage points from target.
- I rebalance by: redirecting new contributions first; selling only if the drift is too large for contributions to correct within a year.
- When I sell, I sell: inside tax-sheltered accounts wherever possible.
Five lines. That is the whole discipline, and having it in writing before a crash is worth more than any amount of analysis during one.
Track the drift, not just the balance
Rebalancing only works if you can see your actual allocation. The EZvest portfolio view shows your real mix across assets, sectors and regions, so drift is visible before it becomes a problem. The Learning Hub covers risk and allocation in short lessons, and FinBot can explain what has moved and why - as an educational tool, with your decisions staying yours.
Frequently asked questions
How often should I rebalance my portfolio?
Once or twice a year is enough for most long-term investors. Research consistently finds that rebalancing more frequently adds costs without materially improving results. The key is having a rule you follow, rather than rebalancing whenever markets make you nervous.
Does rebalancing improve my returns?
Usually not - its main job is controlling risk, not raising returns. Selling winners to buy laggards often slightly reduces returns in a long bull market. What it does is stop your portfolio drifting into a far riskier position than you intended, which is what protects you when the market turns.
How do I rebalance without paying capital gains tax?
Direct new contributions toward whatever is underweight instead of selling what is overweight. This shifts your allocation over time without any disposal, so no capital gains event occurs. You can also rebalance inside a tax-free savings account or retirement annuity, where trades do not trigger CGT.
What is threshold rebalancing?
Instead of rebalancing on fixed dates, you act only when an asset class drifts more than a set amount from its target - commonly 5 percentage points. It responds to what markets actually do rather than the calendar, but it requires you to check your allocation periodically.
Should I rebalance during a market crash?
In principle yes, and that is precisely when it is hardest, because it means buying the thing that just fell. This is why a written rule matters more than judgement. If following it feels unbearable during a crash, your target allocation was probably too aggressive to begin with.
See your drift before it matters
Track your actual allocation across assets and regions in one view, and ask FinBot to explain what has moved and why.
Launch EZvest - it's free to start ->Educational content only. This article is general information, not financial advice. Tax treatment depends on your circumstances and current legislation, and rebalancing can trigger capital gains tax in a taxable account. Consult a licensed financial professional or tax practitioner before making decisions. How we research and review these guides.