How often to rebalance: 98 years of 60/40 data

For a stock/bond portfolio, once a year — or only when an asset class drifts 5 percentage points or more from its target — is enough. In our test of a 60/40 mix of the S&P 500 and 10-year Treasuries over the 50 years 1976–2025, annual rebalancing, rebalancing every two or three years, and 5- and 10-point drift bands all landed between 9.9% and 10.2% a year. The band rules needed 16 and 8 trades instead of 49. What made the real difference was not rebalancing at all: that portfolio ended 2025 at 95.5% stocks.

That is the practical answer to "how often should I rebalance?" — the frequency matters far less than whether you have a rule and follow it. Below are the backtest, why a schedule and a band give such similar results, what rebalancing costs in a taxable account (with a worked example), and how to do it without selling anything. Data as of October 2026; this is a measurement guide, not investment advice.

The test: six rules, one 60/40 portfolio

We used annual total returns for the S&P 500 (dividends included) and 10-year US Treasury bonds from Aswath Damodaran's NYU Stern dataset, updated January 2026. Each rule starts with $10,000 at 60% stocks and 40% bonds and checks the mix once a year, at year end. A "band" rule trades only if the stock share has moved 5 (or 10) percentage points away from 60% — below 55% or above 65% for the 5-point band. No costs or taxes are deducted.

Rule, 1976–2025	Annual return	Ending value of $10,000	Volatility (std. dev.)	Worst year	Highest stock share	Rebalancing trades
Never rebalance	10.89%	$1,754,400	13.2%	−27.2% (2008)	95.5%	0
Every year	10.03%	$1,187,800	10.5%	−18.0% (2022)	68.6%	49
Every 2 years	9.93%	$1,135,800	10.5%	−18.0% (2022)	71.1%	24
Every 3 years	10.23%	$1,300,700	10.5%	−18.0% (2022)	72.8%	16
5-point band	10.13%	$1,243,900	10.6%	−18.0% (2022)	71.0%	16
10-point band	10.24%	$1,310,500	10.7%	−18.0% (2022)	76.4%	8

Our calculation from Damodaran's annual series. "Highest stock share" is the year-end weight before any rebalancing trade.

Three things stand out:

1. Every rule that rebalances lands in a 0.31-point range (9.93% to 10.24% a year), and the order changes with the period. Over 2000–2025 the three-year rule came first (7.23%) and annual rebalancing (6.94%) beat the never-rebalanced portfolio (6.77%). No frequency wins reliably.
2. Never rebalancing earned more because it stopped being a 60/40 portfolio. By 2007 it held 83.5% stocks, so 2008 cost it 27.2% against 13.9% for the annually rebalanced mix. The extra return is the return of a riskier portfolio, not a reward for doing nothing.
3. Bands do the same job with fewer trades. The 5-point band traded in 16 of 50 years — in clusters around big moves (1999–2003, 2008–09) and not at all in quiet stretches.

Vanguard's researchers reached the same conclusion with monthly data: in a 60/40 test over 1926–2009, monthly rebalancing returned 8.5% a year with 12.1% volatility, annual 8.6% with 11.9%, and never rebalancing 9.1% with 14.4% volatility and an average stock share above 84%. Their summary: "there is no optimal frequency or threshold." For most diversified portfolios they found that checking once or twice a year and acting on a 5% drift was a reasonable balance of risk control and cost.

What drift looks like without rebalancing

Drift is not always toward stocks. The chart follows two never-rebalanced 60/40 portfolios: one started in 1976, one in 2000.

Started 1976Started 2000

	Started 1976	Started 2000
1976	61.6
1980	71.6
1985	70
1990	73.1
1995	78.1
2000	85.7	53.9
2005	82.4	47.8
2010	80.1	44
2015	85.6	53.7
2020	90.5	65
2025	95.5	80.7

The portfolio started in 2000 fell to 34.4% stocks at the end of 2008 — after two bear markets it was a bond portfolio — and then climbed to 80.7% by 2025. The investor who set 60/40 in 2000 owned a 35/65 portfolio at the bottom, exactly when stocks were cheapest, and an 80/20 one after a long bull market. Started in 1928, the same buy-and-hold mix reached 99.6% stocks by 2025. Rebalancing is what keeps the risk you chose; the SEC's investor guide describes it simply as "bringing your portfolio back to your original asset allocation mix."

Calendar or band: which rule to pick

Every year49

Every 2 years24

Every 3 years16

5-point band16

10-point band8

Over the full 1928–2025 period the 10-point band traded 15 times in 98 years — the same number of rebalancing events Vanguard's researchers reported for annual monitoring with a 10% threshold in their own test.

A common combination is to check on a fixed date — once or twice a year — and trade only if a band is breached. With annual data that is exactly what our band rows simulate.

Percentage points or percent? "5%" can mean two different rules. Absolute: 60% stocks may move between 55% and 65%. Relative: 5% of the target, so 57% to 63% — a much tighter band that trades more often. For small allocations (say a 10% international slice), an absolute 5-point band is very loose, so a relative band is often used for the small pieces. Write down which one you mean.

What rebalancing costs in a taxable account

In an IRA or 401(k), selling inside the account is not a taxable event; the IRS taxes traditional IRA money when it is withdrawn, and qualified Roth distributions not at all. In a regular brokerage account, every sale that realizes a gain is reported. The IRS taxes gains on assets held more than one year at long-term rates of 0%, 15% or 20%, depending on taxable income; gains on assets held one year or less are taxed as ordinary income.

Most investors with regular contributions can do much of their rebalancing the second way. The SEC lists the three methods: sell over-weighted assets, buy under-weighted ones, or redirect new contributions to the under-weighted category, and to check first whether your method "will trigger transaction fees or tax consequences."

How to rebalance in five steps

1. 1

   Write down the target and the rule

   For example: 60% US stocks, 40% bonds; check every January; trade only outside ±5 points

2. 2

   Count every account together

   Your 401(k), IRA and brokerage account are one portfolio — the mix that matters is the combined one

3. 3

   Measure the current weights

   Use market values on one date, including cash; a broker's per-account pie chart is not enough

4. 4

   Fix it in the cheapest place first

   Redirect new contributions and dividends, then trade inside tax-advantaged accounts, and only then sell in a taxable account

5. 5

   Record what you did

   Keep the trade dates and cost basis — they decide next year's holding periods and the tax on later sales

The second step is where most do-it-yourself rebalancing goes wrong: an investor with bonds in a 401(k) and stocks at two brokers cannot see the real mix in any single app. GetHossa's broker import pulls holdings from several brokers into one view, and realized gains and transactions shows the cost basis and holding period of each lot before you sell.

Common mistakes

• Judging rebalancing by return. Its job is to hold risk steady. In our 1976–2025 test it cost 0.86 points a year against never rebalancing — the price of not ending at 95% stocks. Compare it with the max drawdown you could live with; our S&P 500 drawdown history shows how deep a near-100% stock portfolio has fallen.
• Rebalancing each account separately. Bringing every account to 60/40 creates sales in the taxable account that a trade in the IRA would have avoided.
• Trading on tiny drift. Moving from 61% back to 60% in a taxable account realizes gains for almost no change in risk.
• Forgetting that 2022 happens. Bonds do not always rise when stocks fall: in 2022 the S&P 500 returned −18.0% and the 10-year Treasury −17.8% in Damodaran's series, so rebalancing that year moved money between two losers.
• Measuring performance with the account balance. Contributions used to rebalance inflate the balance; use a time-weighted return when you compare your portfolio with the S&P 500.

Methodology and sources

• Data: annual total returns of the S&P 500 (with dividends) and of 10-year US Treasury bonds, 1928–2025, from Aswath Damodaran's "Historical Returns on Stocks, Bonds and Bills" (NYU Stern, data updated January 1, 2026), downloaded in this run.
• Rules: every portfolio starts at 60/40 on January 1 of the first year. At each year end the stock weight is measured; "every year/2 years/3 years" rebalance back to 60/40 on that schedule, the band rules rebalance only if the stock weight is 5 or 10 percentage points or more away from 60%. A trade is counted when the weight is not already 60%. Volatility is the population standard deviation of annual returns; worst year is the lowest calendar-year return.
• Other periods: 1928–2025 annual returns were 9.45% (never), 8.34% (every year), 8.30% (every 2 years), 8.36% (every 3 years), 8.37% (5-point band) and 8.38% (10-point band); 2000–2025: 6.77%, 6.94%, 6.75%, 7.23%, 7.03% and 7.14%.
• Limits: year-end data cannot show intra-year drift, so a band checked monthly would trade more often. No trading costs, taxes or fund fees are included. Two asset classes only; results for international stocks, small caps or short-term bonds differ. Past returns do not predict future ones.