Portfolio rebalancing, and when to actually do it

Rebalancing is how you keep the risk level you chose after markets move. Here's the mechanics, the two rules that work (calendar and threshold), and why doing it more often isn't better.

By StevenPG April 17, 2026 Reviewed May 21, 2025 5 min read

Educational content only — not financial, tax, or legal advice.

Rebalancing is the simplest risk-control tool in investing, and the one most people either ignore or overdo. The idea: you pick a target mix of assets (say 80% stocks / 20% bonds), markets move the mix around, and once or twice a year you nudge it back to the target.

That’s the whole strategy. No charts to read, no news to follow.

Why the mix drifts

Suppose you start with $100,000: $80,000 in a US stock index and $20,000 in a bond index. After a great year, stocks are up 25% and bonds are flat.

	Start	After	% of portfolio
Stocks	$80,000	$100,000	83.3%
Bonds	$20,000	$20,000	16.7%
Total	$100,000	$120,000

Your “80/20” portfolio is now 83/17. Not dramatic — but do this for five years of stock outperformance and you can end up at 93/7, running far more risk than you chose.

The opposite also happens after a crash. If stocks drop 40% and bonds hold steady, your 80/20 becomes roughly 71/29. You’re now under-exposed to the recovery at exactly the worst time.

Rebalancing forces you to sell what went up and buy what went down — exactly what disciplined investors are supposed to do, and exactly what most people can’t bring themselves to do by feel.

The two rules that work

There are two rebalancing policies with evidence behind them. Both work. Neither dominates the other. Pick one and stop thinking about it.

Calendar rebalancing

Pick a date — your birthday, January 2, the day after Thanksgiving — and rebalance once a year on that date. That’s it.

Once-a-year calendar rebalancing has been shown by Vanguard and others to capture essentially all the benefit of more frequent schedules while minimizing transaction costs and tax drag. Monthly or quarterly rebalancing adds complexity without meaningfully improving returns.

Threshold rebalancing

Only rebalance when an asset class drifts more than a fixed percentage from its target — commonly 5 absolute percentage points. An 80/20 portfolio gets rebalanced only when stocks cross 85% or fall below 75%.

Threshold rebalancing trades fewer times in quiet markets and responds faster to big moves. It requires checking in (quarterly is fine).

What doesn’t work

Rebalancing continuously or every month: more tax and trading cost, no better risk control
Rebalancing based on predictions about what will happen next: that’s market timing, not rebalancing
Ignoring rebalancing entirely: you end up with the wrong risk profile over time

Taxable vs. tax-advantaged accounts

Rebalancing inside a 401(k), IRA, or Roth IRA is free: no tax consequence, just a few clicks. Do it on schedule without thinking.

Rebalancing inside a taxable brokerage sells appreciated assets and triggers capital gains tax. The workaround that works for most investors:

Rebalance with new contributions. If stocks are over target, direct new contributions entirely into bonds until the mix is back to target. No sales, no tax.
Rebalance with dividends. Turn off automatic reinvestment and redirect the cash to the underweight asset class.
Rebalance across accounts. Sell in your IRA (tax-free) to fix the mix; keep the taxable account holding long-term appreciated positions.

Selling for cash in a taxable account is the last resort, not the default.

How rebalancing actually affects returns

Rebalancing is primarily a risk-control tool, not a return enhancer. Most of the time, a “let it ride” portfolio with no rebalancing will deliver slightly higher long-run returns than a rebalanced portfolio, because stocks beat bonds over long horizons and rebalancing trims the winning position.

So why do it? Because “let it ride” means your risk steadily ratchets up over decades, and the worst-case drawdown gets worse every year you don’t rebalance. A disciplined investor with an 80/20 target who wakes up in a 95/5 portfolio the year before retirement is exposed to a catastrophe they didn’t sign up for. The trade: give up a small amount of expected return for a meaningful reduction in worst-case outcomes. That trade is almost always worth taking.

When to change the target (not just rebalance to it)

Rebalancing keeps you at your target allocation. Separately, your target allocation should shift as your situation changes. Roughly:

10+ years from a goal: lean heavily toward stocks (80–100%)
5–10 years from a goal: begin adding bonds (60–80% stocks)
2–5 years from a goal: shift toward capital preservation (40–60% stocks)
< 2 years from a goal: mostly cash and short-term bonds

A “target date” or “lifecycle” fund does both jobs — holds a target allocation and shifts it over time. If you want to delegate the whole problem, they’re a reasonable default.

The one-page rebalancing plan

Pick a target allocation based on your horizon and tolerance for drops.
Pick a rule: once a year on a fixed date, or 5% threshold.
When you rebalance, use new contributions first, IRA trades second, taxable sales last.
Write the rule down somewhere (account nickname, sticky note, calendar event).
Don’t touch the portfolio between rebalances.

Five minutes, once a year. That’s the entire exercise.

Portfolio drift — a concrete example with the math

Start with a $200,000 portfolio targeting 70% stocks / 30% bonds. Stocks return 10% per year; bonds return 3% per year. No rebalancing.

End of year	Stocks value	Bonds value	Total	Stock %	Bond %
Start	$140,000	$60,000	$200,000	70%	30%
Year 1	$154,000	$61,800	$215,800	71.4%	28.6%
Year 3	$186,340	$65,563	$251,903	74.0%	26.0%
Year 5	$225,470	$69,556	$295,026	76.4%	23.6%
Year 7	$272,927	$73,796	$346,723	78.7%	21.3%
Year 10	$362,950	$80,635	$443,585	81.8%	18.2%

After 10 years of doing nothing, you intended a 70/30 portfolio and you have an 82/18 portfolio. You’re carrying 17% more stock exposure than you agreed to when you set the allocation. In a 40% stock crash, your 70/30 portfolio was supposed to lose about 28%. Your actual 82/18 portfolio loses about 33%. That’s $10,000 more in losses on a $300,000 portfolio — from inaction alone.

Rebalancing the 5-year mark. At year 5, stocks are $225,470 (76.4%) and bonds are $69,556 (23.6%). Target is 70/30. Total portfolio: $295,026.

Target stock balance: $295,026 × 70% = $206,518
You have: $225,470 in stocks
Sell $18,952 of stocks, buy $18,952 of bonds.

One transaction. Five minutes. You’re back to 70/30 and the risk level you chose.

Rebalancing methods compared

There are really only two defensible approaches. Both work. The choice is mostly about temperament.

Factor	Calendar rebalancing	Threshold rebalancing
How it works	Rebalance on a fixed date once a year	Rebalance when any asset class drifts ±5% from target
Minimum trades per year	1	0 (quiet markets may need no action)
Maximum trades per year	1	Unlimited (but usually 1–3)
Responds to crashes quickly	No — waits for the date	Yes — a 20% crash triggers rebalancing immediately
Good for	Set-and-forget investors, retirement accounts	Active monitors, volatile portfolios
Main risk	Missing a major drift between annual dates	Over-checking, leading to reactive behavior

A hybrid policy — “rebalance annually, but also if any asset drifts more than 10%” — captures most of the benefit of both. You stay disciplined most years and respond to extreme moves when they happen.

What research says. Vanguard’s analysis found that once-a-year calendar rebalancing produced outcomes nearly identical to monthly rebalancing in terms of risk control, while incurring significantly fewer trades and lower tax drag. More frequent rebalancing is not more rigorous — it’s just more work.

Tax-efficient rebalancing

Inside a 401(k), IRA, or Roth IRA: rebalance freely. Selling appreciated assets inside a tax-advantaged account has zero tax consequence. If your 401(k) has drifted from 70/30 to 80/20, just sell the over-weight fund and buy the under-weight fund. Done.

Inside a taxable brokerage: every sale of appreciated shares triggers a taxable event. Selling $18,952 of a stock index fund that has a cost basis of $8,000 creates $10,952 of capital gains. At 15% long-term rate, that’s $1,642 in tax — just to rebalance.

Strategies to rebalance taxable accounts without (or with minimal) selling:

Direct new contributions to the underweight asset. If stocks are over-target, send your next 3–6 months of contributions entirely into bonds. No sale, no tax, account drifts back toward target naturally. This works until the overweight is too large to fix with contributions alone.

Use dividends and distributions. Turn off automatic dividend reinvestment on over-weight funds. Route those distributions to buy the under-weight asset. Fidelity, Vanguard, and Schwab all support this per-fund.

Rebalance across account types. Hold more of your target bond allocation inside your IRA or 401(k) (where selling is tax-free), and let the taxable account hold equities long-term. When you need to rebalance, sell bonds inside the IRA and redirect new taxable contributions to stocks (or vice versa). You adjust the total portfolio allocation without triggering gains in the taxable account.

The last resort: sell in taxable. If none of the above keeps pace with the drift, selling in taxable is still often worth it — especially for assets held long-term (15% capital gains rate) and especially when the rebalancing trade is large. The reduction in portfolio risk may be worth more than the tax cost.