Most long-term investors do not need to rebalance more than once a year, and many can wait until an asset class drifts five percentage points from its target. Vanguard research published in 2019, covering U.S. markets from 1960 through 2018, found that monthly, annual, and threshold-based rebalancing of a 60/40 portfolio produced average annualized returns within roughly 0.1–0.2 percentage points of each other (per Vanguard's 2019 study). The differences were small; the discipline was what mattered.
NewsJay publishes information and education, not investment advice, and nothing here tells you what to buy or sell.
What does rebalancing actually do?
Rebalancing restores a portfolio to its target mix of stocks, bonds, and other assets after market moves push the weights away from it. A 60/40 portfolio that enjoyed a strong equity run might drift to 70/30, carrying more risk than the investor originally chose.
Its job is risk control, not return enhancement. Selling what rose and buying what fell mechanically trims exposure to whatever has grown most, which historically traded a little upside for a portfolio that behaved as designed.
In taxable accounts, rebalancing has a second dimension: every sale can realize capital gains. That is why the ordering of rebalancing moves matters as much as their timing.
Does rebalancing improve returns, or just manage risk?
Mostly it manages risk. The same Vanguard research, covering 1960–2018, concluded that no rebalancing strategy reliably produced higher returns than another; return differences across strategies were within hundredths of a percentage point on average, while the risk profile of the never-rebalanced portfolio drifted materially over time (per Vanguard, 2019).
The honest framing: rebalancing is a decision to keep holding the risk you chose. A portfolio left alone for decades does not stay 60/40 — it becomes whatever the markets made it, which is usually a stock-heavier, more volatile portfolio.
Calendar-based or threshold-based — which rule works?
Two rules dominate practice. Calendar-based: check on a fixed schedule, such as annually. Threshold-based: check periodically but act only when an asset class drifts beyond a band, commonly five percentage points from target.
Vanguard's 2019 analysis found the five-percentage-point threshold checked monthly or quarterly captured most of the benefit at low cost and low effort. Checking daily added trading costs without meaningful benefit; waiting years allowed risk to drift far from target. A practical sequence:
- Write down target weights for each asset class.
- Check the portfolio on a fixed date, or when markets move sharply.
- Rebalance only when an allocation breaches the agreed band.
- In taxable accounts, first direct new contributions and dividends to underweight assets, which rebalances without selling.
- Place any needed trades in tax-advantaged accounts first, where sales have no tax consequence.
What do taxes do to the rebalancing decision?
In taxable accounts, each rebalancing sale can realize gains taxed in the year realized. The SEC's investor materials note that turnover and frequency of trading are among the costs investors most often overlook (sec.gov). Using contributions, dividends, and bond interest to buy underweight assets rebalances the portfolio with no sale at all.
Where sales are unavoidable, specific-lot identification — choosing which tax lots to sell — can manage the size of the realized gain. The mechanics belong to tax rules that change; the investor's own records and current IRS guidance govern.
What are the classic mistakes?
The costliest mistake is abandonment at the worst moment. Rebalancing in a crash means buying the asset that just fell hardest, which is precisely when the rule feels wrong. Investors who abandoned their bands in 2008 and 2020 locked in the drift toward safety after the damage was done.
The second mistake is over-trading. A 2020–2021-style rally can tempt monthly tinkering that realizes gains and adds costs for a mix the annual rule would have fixed anyway. The evidence says pick one rule and let it be boring.
The third is not writing the rule down. A band that lives in memory becomes a negotiation, and negotiations happen at exactly the wrong times.
What does the evidence leave unanswered?
The studies cover U.S. markets across specific historical periods. Whether the same near-equivalence across strategies holds in future decades is unknown, and no rebalancing rule removes market risk itself. What the evidence does establish: the differences between sensible rules are small, and the difference between having a rule and having none can be a portfolio that no longer matches the risk its owner chose.
One concrete number to keep: five percentage points of drift, checked on a modest schedule, captured most of the measured benefit across six decades of U.S. market history, per Vanguard's 2019 research.
For more context, read What Dollar-Cost Averaging Is and How It Works.
For more context, read automatic.
