Rebalancing: When to Do It and When to Leave It Alone

Rebalancing has a reputation as the responsible thing to do, which is how it becomes the thing people do too often and in the wrong account. Compare what you hold to what you meant to hold, then close the gap. The hard part is knowing when the gap is worth closing.

What rebalancing actually does

Rebalancing restores the risk level you chose. When one part of a portfolio outruns the others, the whole thing quietly becomes a different portfolio, more concentrated in whatever ran and more exposed to the next bad stretch there. Moving money out of the winner and into the laggard puts the mix back where it belongs.

That is the entire benefit, and it is easy to expect more. Rebalancing does not add return on command, call a top or a bottom, or protect you from a broad decline, since a portfolio at target weights still falls when everything falls. It does not repair an allocation that was wrong from the start.

What it does is keep the risk you signed up for. A 60% stock, 40% bond portfolio that drifts to 75/25 is no longer the portfolio whose worst case you accepted when you chose it.

Calendar or bands: two ways to run it

The calendar method picks a date and checks the mix then, once or twice a year. Its strength is that it removes discretion: the date arrives, you look, you fix, and a headline never makes the decision for you. Its weakness is timing. A large move the month after your check date waits almost a year, and in a quiet year it pushes you to trade when nothing needed it.

The threshold method sets a band around each target and acts only when a weight leaves it. It responds to what happened, so a big move gets corrected while it is big. The price is that a volatile stretch can trigger several rounds of trading.

A common convention sets each band at 25% of the target weight, capped at 5 points. On a 60% target, 25% of 60 is 15 points, so the cap applies and the band runs from 55 to 65. On a 15% target it runs from 11.25 to 18.75. The logic: a small slice can wander far in relative terms without changing risk much.

Treat that as one default among several. A flat 5 point band works, and so does a 10% relative band. The number matters less than writing it down before a market move makes one answer feel obvious.

Close the gap with new money first

Selling to rebalance can create a tax bill, and a tax bill is a certain loss taken to fix an uncertain problem. Contributions are cheaper, because new money can be aimed at whatever is underweight without selling anything.

Take a hypothetical portfolio. You start with $240,000 split 60% US stock, 25% international stock, and 15% bonds: $144,000, $60,000, $36,000.

You leave it alone. US stock grows to $186,000, international falls to $58,000, and bonds sit at $36,000. The total is $280,000, and the mix is 66.4% US, 20.7% international, 12.9% bonds. Your 60% target has a band of 55 to 65, so US stock is outside it. Nothing has gone wrong, but it carries more stock risk than you chose.

At $280,000, the targets would be $168,000, $70,000, and $42,000. You are $18,000 heavy in US stock, $12,000 light in international, $6,000 light in bonds. Instead of selling, you redirect a year of contributions: $1,500 a month, sent $1,000 to international and $500 to bonds, nothing to the US fund.

The holdings end at $186,000, $70,000, and $42,000 on a total of $298,000. The mix is 62.4% US, 23.5% international, 14.1% bonds. Every weight is back inside its band, no shares were sold, and no gains were realized.

The contrast matters. Had that $18,000 gone in at target weights, the portfolio would have landed near 66% US, 21% international, 13% bonds. A full year of saving would have moved the mix half a point.

Selling still has a place. When contributions are too small against the gap, or a band breaks hard, trimming is the only tool left. Do it inside a retirement account when you can.

Asset location decides what rebalancing costs

The same trade costs different amounts depending on where it happens. Inside a 401(k) or an IRA, selling one fund to buy another triggers nothing. In a taxable account, a sale can realize a gain, taxed in the year it happens whether or not you reinvest a dollar.

That asymmetry is the practical argument for holding your least tax-efficient assets in retirement accounts and broad stock index funds in taxable ones. Bond interest is taxed at ordinary rates, so a bond fund inside a 401(k) grows without that drag. A broad stock index fund in a taxable account can qualify for lower rates on dividends and is likely to be the position you hold for decades, where a step-up in basis at death does its work. Location is a default, not a law. Municipal bonds in a high bracket will bend it.

The useful consequence: rebalance first where it is free. If bonds are underweight and your 401(k) holds bonds, shift inside the plan and leave the taxable account alone. If the only place to fix the mix is taxable, use contributions and dividends, and sell the lots with the smallest gains.

How many funds is too many

A portfolio is too complicated when you cannot state your weights without opening a statement. That test beats any count.

Complication usually arrives as overlap, not as fund count. A total market fund plus a large cap fund plus a growth fund can look like three positions and behave like one, so your real exposure is hidden and your math is wrong. A target date fund beside sector funds has the same problem: it already contains everything, and the additions are bets on a plan you are no longer following.

Every fund should have a job no other fund does. Large US stock, international stock, bonds, and perhaps a small or value tilt if you have a reason. That is often three to five funds, and a sixth rarely adds diversification. If a fund needs another fund to explain it, it is probably overlap.

The trap: a fall is not a discount

The most expensive rebalancing mistake is emotional, and it shows up right after a decline. The lagging asset is down, the plan says buy more of it, and the story writes itself: it is on sale. That framing feels like discipline and is actually a forecast.

Your target weight is not a prediction but a risk instruction. You buy the laggard because holding less than planned means the portfolio no longer matches the risk you chose, not because you concluded it will recover. It may not recover for years, and the rebalancing still did its job.

The mirror image is as common. Trimming the winner feels like punishing your best performer, so people let it run out of loyalty to a good decision. Letting it run is a decision too, just an unrecorded one that raises your risk without asking you. Both failures treat rebalancing as a judgment about the future rather than a response to the present.

Write the rule down while you are calm, band and date included, so the decision is made before the numbers move. And separate rebalancing from changing your target. If a decline makes you want less stock permanently, that is a new plan, not a quiet edit in a bad week.

When your mix is inside its bands, the correct action is nothing. Sit still, keep contributing, check again on schedule. Doing nothing on purpose is the part nobody writes about, and most of the benefit lives there.

What to do this week

• Write down target weights and the band for each, in a file you will find again.
• Set a check date twelve months out and put a reminder on the calendar.
• Note your current weights beside the targets and which accounts hold each fund.
• If anything is out of band, point your next contributions at the shortfall before selling.
• Route future bond buying into a retirement account, where the interest is not taxed each year.