How Portfolio Rebalancing Controls Risk
A portfolio left alone does not stay where you put it. The asset that has done well grows into a larger share of the whole, and the risk you agreed to becomes a risk you did not.
Updated 9 September 2026
Kabir has more in shares than he chose to
Kabir decided years ago on a split between shares and bonds. He wrote it down, he was comfortable with it, and he has not changed his mind about it since.
He also has not looked at it in a while. When he does, the proportion in shares is noticeably higher than the one he picked. He did not buy any more shares. He made no decision at all. The shares simply did better than the bonds, so they grew into a larger part of the total, and the split drifted while he was busy with other things.
This matters more than it sounds. Kabir now holds a portfolio he never agreed to, and it will fall further in a bad year than the one he did agree to. The drift always runs in the same direction — towards whatever has recently done well — which means an unattended portfolio quietly becomes riskier precisely after the period that made risk feel comfortable.
What rebalancing actually is
Rebalancing is the act of putting the proportions back. Sell some of what has grown too large, buy some of what has shrunk too small, and the portfolio is once again the one that was chosen.
It is worth being precise about why this is done, because the popular version of the argument is wrong. Rebalancing is not a way of buying low and selling high, and it is not a reliable source of extra return. Its purpose is control: it keeps the amount of risk you are running equal to the amount of risk you decided to run. Everything else claimed for it is a side effect that may or may not appear.
The mechanical part is what makes it trustworthy. Kabir does not have to form a view about whether shares are expensive. He compares where the portfolio is against where he said it should be, and if the gap is large enough he closes it. No forecast is involved, which is the whole appeal for someone who has learned to distrust his own forecasts.
Watching the drift happen
Because drift is invisible month to month, it helps to see a run of years laid out at once.
The table below applies a sequence of annual returns to a portfolio split sixty-forty between shares and bonds. The returns are assumed, not market history — they are an illustration of the mechanism rather than evidence about any real period, and a different assumed sequence would produce different amounts at the end.
| Shares did | Bonds did | Untouched: share in shares | Rebalanced value | Untouched value | |
|---|---|---|---|---|---|
| Year 1 | +30% | +5% | 65.0% | ₹12.00 lakh | ₹12.00 lakh |
| Year 2 | -20% | +6% | 58.4% | ₹10.85 lakh | ₹10.69 lakh |
| Year 3 | +25% | +6% | 62.3% | ₹12.74 lakh | ₹12.52 lakh |
| Year 4 | +8% | +5% | 63.0% | ₹13.60 lakh | ₹13.38 lakh |
| Year 5 | -12% | +6% | 58.5% | ₹12.95 lakh | ₹12.67 lakh |
| Year 6 | +18% | +5% | 61.3% | ₹14.61 lakh | ₹14.26 lakh |
Read the fourth column first, because it is the one that does not depend on the assumption. The untouched portfolio's share in shares moves every single year, drifting up after good years and back down after bad ones, and at no point does it sit at the target except by accident. The rebalanced portfolio returns to the target every year by construction. That is the entire finding, and it would hold for any return sequence you cared to substitute.
The last two columns are a different matter and should be read with more suspicion. On this particular path the rebalanced portfolio happens to end slightly ahead. That is an artefact of the order these invented returns came in, not a result. Reverse a couple of them and the untouched portfolio wins. Anyone who tells you rebalancing raises returns is quoting a path, not a property.
Choosing when to do it
Two rules are in common use, and Kabir needs one of them rather than both.
The simplest is a calendar: look once a year, on a date chosen in advance, and act on what you find. Its virtue is that it demands nothing of him between those dates, and the great majority of the benefit of rebalancing comes from doing it at all rather than from doing it at the ideal moment. Its weakness is that it is indifferent to what actually happened — it will trade after a year in which almost nothing moved, and it will leave a large drift in place for eleven months if the movement came just after the last review.
The alternative is a band: pick a range around the target and act only when the portfolio leaves it. This ties the action directly to the thing that matters, which is how far the risk has strayed. It costs more attention, and it can demand trades at exactly the moments trading feels worst, which is also when it is most necessary.
There is a fuller treatment of the trade-off in calendar or threshold rebalancing. The short version is that either works and switching between them does not, because a rule abandoned whenever it becomes uncomfortable is not a rule.
Do it with new money where you can
Before selling anything, the money already arriving is worth looking at.
Every contribution, every dividend, every maturing deposit is money that has not yet been allocated, and directing it at whatever is currently underweight moves the portfolio back towards target without any sale at all. No tax event, no transaction cost, no decision to regret. For someone still contributing steadily relative to the size of the portfolio, this can do most of the work on its own, and the same trick runs in reverse in retirement — withdrawals taken from whatever is overweight rebalance the portfolio on the way out.
This has a limit worth knowing in advance. Once the portfolio is large relative to the money flowing in, contributions cannot correct a serious drift, and after a sharp market move they will not come close. At that point a sale is the only instrument left, and postponing it to avoid tax means choosing to run a risk you rejected in order to defer a bill. That is the wrong way round. Rebalancing with new investments first covers the mechanics in detail.
Rebalancing is not the same as de-risking
One distinction causes a great deal of confusion, and it is worth separating cleanly.
Rebalancing restores today's target. It assumes the target is still right and only the portfolio has moved. De-risking changes the target itself, because the goal has come closer and the portfolio should now be carrying less risk than it was. The first is maintenance; the second is a deliberate change of plan, laid out in what a goal-date glide path is.
Kabir needs both, in that order. A target appropriate to the goal, a rule that keeps the portfolio near it, and a schedule for lowering the target as the money gets closer to being spent. Confusing the second with the third is how people end up either taking too much risk near a deadline or abandoning growth a decade too early.
What to take away
A portfolio that is never touched does not stay as it was built. It follows whatever has recently done well, which means it takes on the most risk at the point when the last few years have made risk seem least frightening.
Rebalancing fixes that, and it is the one portfolio action that requires no view about the future at all. Choose a rule — a date or a band — write it down, use incoming money to do as much of the work as it can, and expect the rule to feel wrong at exactly the moments it is doing its job. Do not expect it to make you money. Expect it to keep you in the portfolio you chose, which is worth more.
Educational content only. This is not personalised financial, investment or tax advice. Figures quoted are historical or illustrative and are not forecasts. Consult a qualified professional before acting on anything you read here.