Why a portfolio stops being what you chose
Set 60% equity and 40% debt, then do nothing. Equity rises faster, and after a strong run the split is 72/28.
Nobody decided that. The portfolio is now materially riskier than the one you chose, and the change happened silently, during the period when it felt best. Then the market falls, the split becomes 52/48, and the portfolio is more conservative than intended at precisely the moment cheap equity is available.
Drift always moves you towards more risk after a rise and less risk after a fall — the opposite of what you would choose deliberately in either case.
What it does and does not do
The usual pitch is that rebalancing systematically buys low and sells high. That is literally true and it is not the same as improving returns.
In a long trending market, rebalancing costs you. Trimming equity every year through a sustained bull run means holding less of the thing that went up. The “rebalancing bonus” appears mainly when asset classes are volatile and mean-reverting, and it is modest even then.
So the honest case is not about return:
- It holds risk constant. You continue owning the portfolio you chose rather than one the market assembled.
- It is a rule, not a judgement. It forces the two actions nobody takes voluntarily — selling after a rise and buying after a fall.
- It protects the goal. A portfolio that drifted to 80% equity two years before a goal is exposed in exactly the way the preservation phase warns about.
Anyone promising rebalancing as a return-enhancement technique is describing the smaller and less reliable half of the case.
Calendar or threshold
| Calendar | Threshold | |
|---|---|---|
| Trigger | A fixed date — annually is typical | The split moves beyond a band, say ±5 points |
| Effort | Check once a year | Requires periodic monitoring |
| Transactions | Predictable, sometimes unnecessary | Only when something actually moved |
| Responds to a crash | Only at the next date | Immediately, which is when it matters most |
| Risk | Rebalancing when nothing needed it | Over-trading in a volatile period |
The combination is what most careful practice uses: check on a fixed date, and act only if the split is outside the band. That gives you the discipline of a calendar with the transaction economy of a threshold, and it avoids both failure modes.
Annual is enough for most households. More frequent rebalancing adds cost and tax without materially changing risk.
The cheapest way is not to sell at all
The practical point most descriptions omit.
Rebalance with new money. Instead of selling the overweight asset, direct your next contributions entirely to the underweight one until the split is restored.
Why it is better: no sale means no capital gains event, no exit load, and no transaction cost. You are correcting the drift with money that was going in anyway.
For anyone still in the accumulation phase and contributing monthly, this handles most ordinary drift on its own. Selling becomes necessary only when the gap is too large for contributions to close in reasonable time — after a very strong run, or once contributions are small relative to the portfolio.
The same logic applies to withdrawals in reverse: draw from the overweight asset, which rebalances as a side effect of spending.
The costs of doing it by selling
- Capital gains. A switch between funds is a redemption and a fresh purchase, so it realises gains. Treatment depends on rules that change — verify the current position, and note that this cost is often larger than the rebalancing benefit for a small drift.
- Exit load, where units are inside their load window. Rebalancing recently-purchased units is worth avoiding for this reason alone.
- Time out of the market, if the switch is not same-day.
These costs are the argument for a wide band rather than a tight one. A ±5 point tolerance is not laziness; it is the recognition that correcting a 2-point drift costs more than the drift does.
Rebalancing within a single fund — which is what a hybrid fund does — avoids the tax event entirely, and that is a genuine structural advantage of the wrapper rather than a marketing claim.
Doing it
- Know the target. Written down, from asset allocation.
- Pick a date. The same week each year.
- Compute the actual split across every holding, including any employer or retirement account.
- Compare against the band. Inside it, do nothing — that is a valid outcome and the most common one.
- If outside, correct with contributions first, and only sell if contributions cannot close the gap.
- Rebalance the glide too. If a goal is closer, the target itself should have moved, per goal-based investing.
Step four is where discipline is tested, because doing nothing feels like not managing the portfolio. It usually is the correct answer.
Where it goes wrong
- Rebalancing on performance rather than allocation. Selling a fund because it lagged is not rebalancing; it is chasing, and it usually happens at the wrong point in a cycle.
- Too tight a band. Costs exceed the benefit for small drifts.
- Too frequent. Monthly rebalancing generates cost and tax without materially changing risk.
- Ignoring accounts you do not look at. Retirement accounts are part of the allocation whether or not you think of them as such.
- Abandoning it in a crash. The rebalance after a fall is the one that requires buying into weakness, and it is both the hardest and the one the whole rule exists to force.
Seeing the drift
Rebalancing needs one number you probably do not have to hand: the actual current split across everything you own.
FNOTrader's Mutual Funds app values holdings against the full AMFI NAV history — around 34 million NAV rows — so the current allocation is a figure rather than an estimate, and drift is visible before it becomes large. Maximum drawdown alongside it answers the related question: whether the drifted allocation is one you would still choose.
FNOTrader is not a SEBI-registered investment adviser and this is not investment advice.
Common questions
What is rebalancing?
Restoring your portfolio to its target asset allocation after market movement has changed it. Drift always moves you towards more risk after a rise and less after a fall — the opposite of what you would choose deliberately in either case.
Does rebalancing improve returns?
Usually not. In a long trending market it costs you, because trimming equity through a sustained rise means holding less of the thing that went up. Its real function is holding risk constant, which is a better reason than the one usually given.
Should I rebalance on a calendar or a threshold?
Most careful practice combines them: check on a fixed date, act only if the split is outside a band such as ±5 points. That gives calendar discipline with threshold economy and avoids both over-trading and delayed response.
How often should I rebalance?
Annually is enough for most households. More frequent rebalancing adds cost and tax without materially changing risk.
What is the cheapest way to rebalance?
With new money. Direct your next contributions entirely to the underweight asset instead of selling the overweight one — no sale means no capital gains event, no exit load and no transaction cost. Reverse it for withdrawals by drawing from the overweight asset.
What does rebalancing by selling cost?
A switch is a redemption and a fresh purchase, so it realises capital gains; exit load may apply on recently bought units; and there may be time out of the market. These costs are the argument for a wide band rather than a tight one.
Is it rebalancing to sell a fund that underperformed?
No — that is chasing performance, and it usually happens at the wrong point in a cycle. Rebalancing acts on the allocation drifting from target, not on which fund lagged.
What is the hardest rebalance?
The one after a fall, which requires buying into weakness. It is also the one the entire rule exists to force, since it is the action nobody takes voluntarily.
Continue reading
More in Mutual Funds · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article