Rebalancing a portfolio: calendar vs threshold, drift and taxes

Every portfolio starts with a plan: so much in stocks, so much in bonds, so much in cash. Then markets move, and the plan quietly stops describing what you own. A year in which stocks rise 25% and bonds go nowhere turns a 60/40 portfolio into something closer to 65/35. Ten such years and you may own a portfolio you would never have chosen on day one.

Rebalancing is the habit of pulling the mix back to its target. It is simple to state and easy to neglect, and it raises practical questions: how often, how far to let things drift, which account to trade in, and how to avoid a tax bill. This guide answers them with a worked example, the two standard methods, and 23 years of real fund data that show what rebalancing did and did not do.

What portfolio drift is

Your target allocation is the mix you decided on, for example 60% stocks and 40% bonds (the Asset allocation: how stock, bond and cash mixes have behaved guide covers how people choose one). Your actual allocation is the mix you hold today, measured at market value. The gap between the two is drift.

Drift is not a mistake. It is arithmetic. If one holding grows 30% and another grows 2%, the first one's share of the total must rise. Stocks usually grow faster than bonds over long periods, so a portfolio left alone tends to become more stock-heavy over time. That also means it becomes more volatile, and a deeper fall becomes possible, right when you have more money at stake.

The example has a hidden lesson. Rebalancing makes you sell some of what just did well and buy some of what lagged. That feels wrong to most people, which is one reason it is easier to follow a written rule than to decide each time.

The two standard methods

The SEC's beginner guide to allocation describes both approaches: rebalance on a calendar, or rebalance whenever an asset class moves more than a set amount from its target.

Calendar rebalancing

You pick a date and check the mix on that date only: every quarter, every six months or every year. If the mix is off, you trade back to target. Many investors tie it to a date they already pay attention to, such as a birthday, tax season or the end of the year.

• Strengths: easy to remember, easy to automate, and it limits how often you look at the portfolio, which helps people who are tempted to tinker.
• Weaknesses: it ignores how far things moved. You may trade when drift is tiny, and you may sit still for months while a crash pushes the mix far off target.

Threshold (band) rebalancing

You set a tolerance band around each target and act only when a weight leaves it. A common convention is an absolute band of 5 percentage points: with a 60% stock target, you rebalance if stocks rise above 65% or fall below 55%. Some investors use relative bands for small holdings instead, for example 25% of the target weight, so a 10% holding has a band of 7.5% to 12.5%.

• Strengths: you trade only when drift is meaningful, and you react to big market moves rather than waiting for a date.
• Weaknesses: you have to check the portfolio regularly to know whether a band was crossed, and very narrow bands cause frequent trading.

Hybrid approaches

Many people combine the two: look at the portfolio on a schedule (say, quarterly), but trade only if a band has been breached. That keeps the monitoring light while avoiding pointless trades. Funds that manage the mix for you, such as target-date and balanced funds, do the rebalancing inside the fund under their own published policy.

What 23 years of real data show

To see drift on real numbers, we started a hypothetical $10,000 portfolio at the end of October 2003 with 60% in SPY, a fund that tracks the S&P 500, and 40% in AGG, a fund that tracks the broad US investment-grade bond market. AGG's first full month sets the start date. Each month we applied each fund's total return, with dividends reinvested at that month's close, and ran three policies through October 29, 2026 (276 months):

Volatility is the annualized standard deviation of monthly returns, and the deepest decline is measured on month-end values, so intra-month lows were somewhat deeper.

What the numbers say

The untouched portfolio stopped being a 60/40 portfolio. In the 2008 crash its stock share fell to 48% by February 2009, which made it more conservative at the bottom. Then a long bull market lifted the stock share year after year: about 77% at the end of 2019, 83% at the end of 2021, and 89.5% by October 2026. Someone who chose 60/40 for its risk level ended up holding something close to an all-stock portfolio.

Rebalancing cost return in this period. Stocks beat bonds by a wide margin over these 23 years, so the policy that let stocks grow largest finished with the most money. That is the honest result, and it will be true in any period when stocks strongly outperform. A rebalanced portfolio earns roughly what its target mix earns, no more.

Rebalancing kept the risk you signed up for. The annual and band versions had lower volatility and stayed close to 60%. The deepest-decline figures look similar only because the biggest crash in the sample, 2008, arrived when the untouched portfolio held just 67% stocks. Its risk today is much higher. As simple arithmetic, a 50% stock crash with bonds flat would cost a 90/10 portfolio about 45% and a 60/40 portfolio about 30%.

Calendar versus band made little difference. Annual and band rebalancing ended about 1% apart. The band approach traded 15 times instead of 23, because in calm years the mix never left its band. This matches the general finding that the exact schedule matters much less than having one.

Ways to rebalance

There is more than one way to move a portfolio back to target, and the order matters for cost and taxes.

1. Direct new money. Put new contributions into whatever is underweight. In the drift example, adding $1,000 entirely to bonds brings the stock share from 65.7% down to $7,800 ÷ $12,880 = 60.6% with no sale at all.
2. Redirect income. Instead of reinvesting dividends and interest automatically in the fund that paid them, send them to the underweight asset.
3. Trade inside tax-advantaged accounts. Sales inside a 401(k) or IRA do not create a current tax bill (the Retirement accounts explained: 401(k), IRA, Roth and 2026 limits guide explains why), so that is often the cheapest place to sell the overweight asset.
4. Sell and buy in a taxable account. This works, but it can realize capital gains.
5. Take withdrawals from the overweight asset. Retirees who are drawing money can rebalance simply by choosing what to sell for spending.

Tax considerations in the US

Rebalancing inside a taxable brokerage account usually means selling something that has gone up, and that is a taxable event.

• Holding period. As of 2026, gains on assets held more than one year are long-term and taxed at 0%, 15% or 20% depending on taxable income. For a single filer in 2026, the 0% rate applies up to $49,450 of taxable income and the 20% rate above $545,500 (IRS Rev. Proc. 2025-32). Gains on assets held one year or less are short-term and taxed as ordinary income.
• Net investment income tax. Higher earners may owe an extra 3.8% on investment income above $200,000 of modified adjusted gross income (single) or $250,000 (married filing jointly).
• Selling losers. If an asset you are trimming is below what you paid, selling it realizes a loss that can offset gains, and up to $3,000 a year of ordinary income.
• Wash sales. If you sell at a loss and buy the same or a "substantially identical" security within 30 days before or after the sale, including in your IRA, the IRS disallows the loss for now (Publication 550). This can trip up people who rebalance and reinvest automatically at the same time.

Two habits keep the bill small: rebalance with cash flows before selling, and when you do sell in a taxable account, sell lots held more than a year. Your broker's cost-basis records show which lots are which.

Pitfalls

• Rebalancing too often. Monthly trading of tiny drifts adds spreads, possible fees and tax events for almost no benefit.
• Never rebalancing. The most common failure is simply forgetting, which lets risk climb in long bull markets, as the data above show.
• Treating it as market timing. Rebalancing follows a rule, not a forecast. Skipping it because "stocks look strong" turns a rule into a guess.
• Ignoring the whole picture. Rebalance across all your accounts together. A 401(k) that is all stocks and an IRA that is all bonds can still add up to 60/40.
• Forgetting that the target can change. As goals near, many people lower their stock target on purpose. That is a change of plan, not rebalancing, and it deserves its own decision.

How Finabase helps

A Finabase portfolio records your positions and trades and values them at live prices, with each position's cost, current value and gain, plus a value history over time. Adding up the value of your stock funds and bond funds there gives you your actual mix to compare with your target. The comparison tool charts several symbols from the same start date, which makes it easy to see how far a stock fund and a bond fund have diverged since you last rebalanced, and every fund has a symbol page with its price history and dividends. Finabase does not hold accounts or place trades; any rebalancing happens at your broker or plan provider.

Frequently asked questions

How often should I rebalance my portfolio?

Once a year, or whenever a holding drifts about 5 percentage points from its target, is a common convention. In the SPY/AGG data above, annual and band-based rebalancing produced nearly identical results, so having a rule matters more than its exact frequency.

Does rebalancing increase returns?

Not reliably. In a period when stocks beat bonds, as from 2003 to 2026, rebalancing lowered returns because it kept trimming stocks. Its job is to keep risk at the level you chose.

Should I rebalance during a market crash?

A rule-based plan says yes: a crash lowers your stock share, and rebalancing buys stocks to restore it. In early 2009 the untouched portfolio had fallen to 48% stocks. Whether you follow the rule in a crash is the real test of the plan.

Do I have to pay taxes when I rebalance?

Only if you sell at a gain in a taxable account. Trades inside 401(k)s and IRAs are not taxed when made, and steering new contributions or dividends to the underweight asset avoids selling altogether.

Do target-date funds rebalance automatically?

Yes. They rebalance inside the fund and also shift their mix toward bonds over time on a schedule the fund publishes, called the glide path.

Summary

A portfolio drifts because its parts grow at different rates, and left alone it tends to become more stock-heavy and riskier over time. In Finabase data, a 60/40 SPY/AGG portfolio started in 2003 drifted to about 90% stocks by 2026, while yearly or band-based rebalancing held it near 60%. Rebalancing did not add return in that rising market; it kept the risk at the level chosen. Pick a simple rule, use new money and tax-advantaged accounts first, and mind holding periods and wash sales when you sell in a taxable account.