Portfolio Rebalancing: When, How Often and How to Do It Cheaply

Sixty thousand dollars in a stock fund, forty thousand in a bond fund. You place no trades all year. Stocks return 36 percent. Bonds lose 4 percent. You now own something you did not choose: a portfolio that is 68 percent stocks, carrying risk you never agreed to and picked up in silence while you were doing nothing in particular.

The year that redesigns your portfolio for you

Sleeve	Start	Return	End value	End weight
Stocks	$60,000	+36%	$81,600	68.0%
Bonds	$40,000	4% loss	$38,400	32.0%
Total	$100,000		$120,000	100%

Check it yourself: 81,600 / 120,000 = 0.68 and 38,400 / 120,000 = 0.32.

Drift always runs the same direction. Whatever has risen becomes a bigger share of your money, so your stock weight grows after stocks have already done well, which is when prices are higher and a large fall is more likely to be waiting, so your portfolio reaches its most aggressive shape immediately before aggression stops being rewarded.

Put a number on that. If stocks now fall 30 percent, the drifted portfolio loses 0.68 x 120,000 x 0.30 = $24,480. Held at target it would have lost 0.60 x 120,000 x 0.30 = $21,600. The extra $2,880 is what the drift cost you. No statement ever warned you it was building up.

The two trades

Work from the new total of $120,000. Reaching for the old $100,000 is the step people get wrong.

• Target stock value: 0.60 x 120,000 = $72,000
• Target bond value: 0.40 x 120,000 = $48,000
• Stocks sit at $81,600, so sell 81,600 - 72,000 = $9,600
• Bonds sit at $38,400, so buy 48,000 - 38,400 = $9,600

One sale, one purchase, $9,600 each way. The total is still $120,000 and the mix is 60/40 again. Nothing was added and nothing was withdrawn. The rebalancing calculator will produce both figures from your own holdings.

It will feel wrong while you do it. You are selling the fund that did well to buy the one that disappointed you, which is the opposite of what your instincts want, and that discomfort is the product working. If it ever feels natural, check whether you are chasing performance.

The version with no sale at all

Money still going in every month? Then you may not need to sell anything. Send the new contributions to whichever sleeve is short until the weights come back.

In the example, stocks are at $81,600. You want that to be 60 percent of the total, so bonds have to reach 81,600 x (40 / 60) = $54,400, they hold $38,400 today, and 54,400 - 38,400 = $16,000 of new money into the bond fund does the job. The portfolio then totals $136,000. Stocks sit at exactly 60 percent, and no taxable sale ever happened.

At $500 a month that takes over two years, so treat contributions as the thing that slows the drift and an annual trade as the thing that finishes it.

Dates, or bands, or both

Two triggers work. The third option, rebalancing whenever the news makes you anxious, is not a method.

Method	The rule	Trades per year	Who it suits
Calendar	Check on a fixed date, act if weights are off	Usually one	Anyone who wants one task a year
Absolute band	Act when a weight moves 5 points from target	Varies with markets	People who read their statements anyway
Relative band	Act when a weight moves 25 percent of its target	Rare	Large taxable accounts avoiding trades

The two kinds of band sound alike and behave nothing alike. With a 60 percent stock target, an absolute 5 point band triggers outside the range 55 to 65 while a relative 25 percent band means 25 percent of 60, which is 15 points, so it only triggers outside 45 to 75. The drifted portfolio above sits at 68 percent. It breaches the absolute band and sits comfortably inside the relative one.

Combining them is what I would do. One calendar reminder a year, and a trade only if a band has been breached, which means most years you open the spreadsheet, see that nothing moved much and close it again.

What it costs when the account is taxable

Inside a 401(k) or an IRA, rebalancing costs nothing in tax, so you can sell and buy freely, and those accounts should carry as much of the work as possible. The account types are in retirement portfolio basics.

In a taxable brokerage account, selling an appreciated fund realizes a capital gain. Say your $9,600 sale contains a $3,000 gain. If you held those shares longer than one year, the gain qualifies for long-term treatment, and at an illustrative 15 percent rate that is 3,000 x 0.15 = $450. Sell shares held for one year or less and the same gain is taxed as ordinary income, which at an illustrative 24 percent rate would be 3,000 x 0.24 = $720. Rates depend on your own situation and the IRS publishes the current ones, so read these figures as the size of the gap and look up your own.

Same trade. Two hundred and seventy dollars apart, decided by a date. Check purchase dates before you sell, and use contributions where you can. The wider picture is in tax-efficient investing.

The routine, start to finish

1. Pick one date a year and put it in your calendar. Your birthday works. You will not forget it, and it has nothing to do with the news.
2. On the day, write down what each holding is worth. Then work out its percentage of the total.
3. Compare each one with your written target. Nothing more than five points out? Close the file and go and do something else.
4. If a band has been breached, calculate the target dollar value for each sleeve from today’s total and trade the difference, keeping as much of it as possible inside sheltered accounts.
5. If you are trading exchange-traded funds, use limit orders during regular market hours and avoid market orders at the open, for the reasons in what an ETF is.

Why more often is worse

Monthly rebalancing produces a stream of small trades, each paying a bid/ask spread and, in a taxable account, a small tax bill, in exchange for a risk control benefit that annual rebalancing had already captured. Frequency is where people spend real money on almost nothing.

The same goes for precision. Getting back to 60.0 percent is no better than getting back to somewhere near 60, and chasing the decimal is how a once-a-year job turns into a hobby you eventually abandon. A rule you actually run for twenty years beats a more elegant rule you drop after the second year.

What rebalancing will not do

It will not reliably raise your returns. Over long periods when stocks beat bonds, trimming stocks to buy bonds slightly reduces the ending balance, and you accept that in exchange for a portfolio that stays near the risk you chose. That is a fair trade as long as you know you are making one.

It will not rescue a target that was wrong to begin with. Rebalancing to 60/40 every year for someone who needs the money in four years just keeps a mistake tidy, which is why asset allocation comes first.

And it offers nothing in a year when both sleeves fall together, as 2022 showed. Buying more of a falling bond fund with the proceeds of a falling stock fund felt absurd at the time, and the household that did it anyway reached the far side of that year still owning the same two funds at the weights it had chosen in a calm month. The discipline pays across whole cycles. That is why the rule goes in writing before the cycle starts, while you can still think straight.

Before the week is out, find out what your stock and bond weights actually are, compare them with the target you wrote down, and put one rebalancing date in the calendar with a reminder attached. If new money is what you plan to rebalance with, decide how much and how often in dollar-cost averaging versus lump sum, or check what stuck with the portfolio construction quiz.