How to Rebalance a Portfolio Using New Investments First

Neha's portfolio corrects itself, for now

Neha has been investing for about a year. She has a monthly instruction going into two funds, a target split she chose at the start, and a growing suspicion that she is supposed to be doing something called rebalancing.

She is, but her situation is unusually forgiving and it is worth understanding why. Her portfolio is small and her monthly contribution is large relative to it. Every payment she makes is a meaningful fraction of the whole, which means that where she directs those payments largely determines where the portfolio ends up. She can hold her allocation almost exactly on target without ever selling anything, simply by sending each month's money to whichever side has fallen behind.

This will not last. It is worth using while it does, and worth knowing when it stops.

Why selling is the expensive option

Rebalancing by sale has three costs, and only one of them is obvious.

The obvious one is transaction cost, which for a fund is usually small and occasionally not. The second is tax: a sale at a gain is a taxable event in a way that a purchase never is, and it can convert a paper gain into a real bill years before the money was going to be spent. The third is the least discussed and possibly the largest — a sale requires a decision, and decisions get postponed. Someone who has to sell an asset that has been doing well in order to buy one that has not will find reasons to wait, and the waiting is where the risk control is lost.

Directing new money has none of these. There is no transaction beyond the one Neha was making anyway, no gain is realised, and no decision is required in the moment because the rule can be set in advance. It is rebalancing that happens by default rather than by effort, which is the only kind that reliably happens at all.

The arithmetic

The method is simple enough to do on paper, and doing it on paper once makes it obvious.

Value everything on the same day. Add it up. Work out what each holding would be worth if the portfolio were exactly on target, and subtract to get the gap for each one — some positive, some negative. Then split this month's contribution across the holdings that are short, in proportion to how short they are.

If the contribution is larger than the total shortfall, the portfolio lands exactly on target and the remainder is split in the target proportions. If it is smaller, the portfolio moves closer without arriving, which is fine. Neha does this again next month.

Two refinements are worth adding once she is comfortable. Dividends and interest, if she is taking them rather than reinvesting automatically, are contributions she did not have to make and should be directed the same way. And once she has a bonus or any irregular sum, that is by far the most efficient rebalancing instrument she will ever hold, because it is large enough to close a gap that twelve monthly payments could not.

The same trick in reverse

The method does not stop working when the contributions do. It inverts.

Someone taking money out — Lakshmi, living off a portfolio rather than adding to one — has the same instrument available. Each withdrawal comes from whatever is currently overweight, which means the act of funding her spending is also the act of restoring her allocation. It is one of the few genuinely free things in decumulation, and it is why the withdrawal question and the rebalancing question should be answered together rather than separately.

When it stops being enough

Cash-flow rebalancing has a limit, and Neha will reach it without noticing unless she is watching for it.

The method works because her contributions are large relative to her portfolio. As the portfolio grows, that ratio falls. A monthly payment that could shift the allocation by a couple of points in year one will shift it by a fraction of a point in year fifteen, and at that point the flows are no longer an instrument. They are a rounding error.

The other limit arrives faster and matters more. After a sharp market move — the kind that takes one side of a portfolio down by a fifth in a few weeks — the gap is far larger than any plausible contribution. Neha would need years of payments to close it, and years is exactly what she does not have if the point of the target was to control her risk.

When the drift is beyond the band and the flows cannot close it, the sale is the instrument, and declining to make it is a decision to run a risk you already rejected. Deferring tax is a real benefit and it is not worth an allocation you would not have chosen. The order of operations is: decide what the portfolio should be, then find the cheapest compliant way to get there — not decide what is cheap and let that determine the portfolio.

What gets set up

The setup is a written target and a band around it, and a date each year to check where things actually are. That is covered in calendar or threshold rebalancing.

Between those dates, every new rupee points at whichever side is behind. Many platforms will not do this automatically, so in practice it means adjusting the standing instruction once or twice a year rather than genuinely doing it monthly, and that is close enough — the precision here is worth much less than the habit.

And this arrangement can be expected to work for a good number of years and then quietly stop being sufficient. The signal is not a date; it is the first annual review where the contributions obviously cannot close the gap. When that happens she has not done anything wrong. She has simply accumulated enough that the portfolio needs managing rather than merely feeding.

What to take away

New money is the cheapest rebalancing tool available and the only one that requires no decision at the moment it is used. Direct contributions, dividends and bonuses at whatever is underweight, take withdrawals from whatever is overweight, and a great deal of drift never accumulates in the first place.

But do not let the cheapness of the tool decide the allocation. When the portfolio has moved beyond its band and the flows cannot bring it back, sell — and treat the tax as the cost of holding the risk you actually chose.