How Often Should You Rebalance Your Portfolio? A Straight Answer

Once a year is enough for most people. Check the portfolio on a fixed date, and only trade if a holding has drifted past a band you set in advance, commonly 5 percentage points absolute or 25% of its own target weight. Rebalancing more often than that adds trades, costs and, in a taxable US account, tax bills, without meaningfully improving the risk control you are actually buying.

That is the answer. The rest of this is about why it holds, where it breaks, and how to decide your own number rather than borrowing someone else's. Everything below is educational content, not investment or tax advice.

How often should you rebalance your portfolio?

The practical answer for a long term investor is annually, with a drift band as an override. Pick a date you will remember, look at the allocation, and act only if something has moved far enough to matter. If you are still contributing regularly, direct new money to whatever is underweight in between, which often keeps the portfolio close enough that the annual check finds nothing to do.

Here is the full menu, with the honest trade-off for each.

What the research says about rebalancing frequency

The finding that keeps reappearing is that frequency barely matters and doing something matters a lot. Vanguard's published work on rebalancing has landed in the same place across multiple studies: annual or threshold based rebalancing captures essentially all of the risk control benefit that daily or monthly rebalancing provides, at a small fraction of the trades and costs. Their 2024 research on target date funds found threshold based approaches worth roughly 15 to 25 basis points a year over calendar based ones, which is a genuine edge and a small one.

Read that carefully, because it is easy to over-read. It does not say rebalancing makes you money. It says that once you rebalance at all, doing it more precisely buys you very little, and the extra precision costs real money in commissions, spreads and taxes. The gap between never rebalancing and rebalancing annually is large. The gap between annually and monthly is noise.

Calendar rebalancing versus rebalancing bands

Calendar rebalancing trades on a date. Band rebalancing trades on a condition. The problem with a pure calendar is that markets do not respect it: a portfolio can be perfectly on target on your review date and 8 points off two weeks later, and you would not touch it either time. The problem with pure bands is that they require you to actually watch, which most people do not do consistently.

Combining them solves both. Review annually on a fixed date, and rebalance only if a band has been breached. You get the discipline of a schedule and the sensitivity of a threshold, and in years where nothing much happened, you correctly do nothing. That is the version I would default to, and the one worth testing on your own holdings rather than accepting because someone wrote it down.

What is the 5/25 rule for rebalancing?

Rebalance a holding when it drifts either 5 absolute percentage points from its target or 25% of its own target weight, whichever threshold is smaller. It was popularized by Larry Swedroe and is widely used in the Bogleheads community, and the reason it works is that it scales.

Take a 40% target. Five points absolute triggers at 35% or 45%; 25% relative would mean waiting for 30% or 50%, which is too much drift. So the 5 point rule binds. Now take a 4% target sleeve. Five points absolute would mean waiting for it to hit 9%, more than doubling; 25% relative triggers at 3% or 5%. So the relative rule binds. One rule, sensible behavior at both ends, which is why it beats a single fixed percentage.

Does rebalancing more often produce better returns?

No, and frequently the opposite. Rebalancing systematically trims whatever is compounding fastest, so during a long run in one asset, every additional rebalance transfers money away from the thing that is working. Over the mega cap equity run of the last decade, portfolios that rebalanced frequently gave up return to portfolios that let the winners ride. That is not a failure of rebalancing; it is what rebalancing is supposed to do.

The return question is the wrong one to optimize. The number rebalancing actually improves is the worst drawdown, and the risk it removes is that you end up holding a portfolio far more aggressive than the one you chose, right before the market tests it. If you want to see the size of both effects on your own holdings, backtest the portfolio under two or three cadences over the same window and compare the drawdown column, not the return column.

How taxes change the answer in a taxable account

In an IRA or 401(k), rebalancing is free of tax consequences and the only cost is trading friction. In a US taxable brokerage account, every sale of an appreciated position realizes a capital gain that year, and that single fact should push your cadence out rather than in.

Two details matter most. Positions held a year or less are taxed at ordinary income rates rather than long term capital gains rates, so a quarterly schedule mechanically realizes more short term gains than an annual one. And net capital losses offset gains first, then up to $3,000 of ordinary income per year, $1,500 if married filing separately, with the remainder carrying forward indefinitely (IRS Topic 409). That is why rebalancing into a falling market is the cheapest kind: the sales you need to make are often at a loss.

The practical sequence for someone with both account types is to rebalance inside the tax advantaged accounts first, use new contributions second, and only sell in the taxable account when the drift is genuinely large. If the year's trading did leave you with a pile of realized gains and losses to reconcile, that is a good moment to get the return prepared properly rather than guessing at the carryforward.

When you should rebalance off-schedule

There are a few moments where the calendar is beside the point. A large market move that pushes an allocation well past its band is the obvious one. A change in your own circumstances is the better one: if your time horizon or your tolerance for a drawdown has genuinely changed, the target allocation itself needs revisiting, and rebalancing to an out-of-date target is precision applied to the wrong number.

The moment to be most suspicious of is the one that feels most urgent. Wanting to rebalance because an asset has been falling for three months and it feels bad is not a rule, it is a reaction, and reacting is what the written band exists to prevent.

Picking your own cadence

Write down three things before you touch anything: your target weights, the band that triggers a trade, and the date you will check. Then test that combination against real history before you adopt it, because the schedule that looks sensible on paper is not always the one you could have lived with. Running the same allocation under annual, quarterly and band-based rules over a window that includes a real bear market takes minutes and settles the question with your own numbers instead of a rule of thumb.

If the portfolio in question is an index you designed yourself rather than a handful of funds, the cadence question is even more consequential, because your weighting scheme depends on the reset to survive at all. That interaction is covered in portfolio rebalancing, and the mechanics of what a reset does to an equal weight index are worth reading alongside it. For how the big index providers handle the same problem on a published schedule, see how often index funds are rebalanced.

Indexes is educational and informational software. It never places trades, connects to a brokerage or holds assets, and nothing here is investment or tax advice. Backtests are hypothetical and past results do not predict future returns.