You built the portfolio — a target allocation, low-cost index funds, automatic contributions from every paycheck. Then the market moves, the percentages drift, and the question arrives that stalls more physician investors than any other piece of portfolio maintenance: how often should you rebalance? The honest answer has two parts. First, the precise frequency matters far less than the marketing around it suggests. Second, having a written rule — any reasonable rule — matters far more, because the moment rebalancing becomes genuinely important is exactly the moment you will not want to do it. This article states what rebalancing actually buys you, gives you the two trigger systems that work, shows where to execute so it costs nothing in tax, and explains why the real payoff is behavioral.
Rebalancing controls risk; it does not reliably add return
Start by discarding the most common sales pitch. The "rebalancing bonus" — the idea that systematically selling winners and buying losers adds meaningful return over time — is mostly a myth in diversified stock-and-bond portfolios. Because stocks carry a higher expected return than bonds, rebalancing out of stocks after they rise usually trims long-run expected return slightly rather than enhancing it. A 2022 analytical paper from one of the largest asset managers reaches the same conclusion industry research has converged on for years: the purpose of rebalancing is to manage risk, not to maximize return, and rebalancing too frequently — monthly or quarterly — adds transaction cost and tax drag without improving outcomes. Roughly annual attention is enough.
Rebalancing is the maintenance fee you pay to keep the risk level you chose; it is not a return strategy. Judge it by whether your portfolio still matches your written target, never by whether it beat anything.
That reframing settles most of the anxious questions. Did you miss the optimal window? There is no optimal window worth chasing. Should you rebalance before an election, a rate decision, a rumor? The rule does not care, which is the point.
Left alone, your portfolio becomes a different portfolio
Drift is not neutral. Each year you skip rebalancing during a rising market, the portfolio quietly reallocates itself toward stocks — which means toward more risk — without asking you.
Example calculation
Assumptions, stated explicitly: this is a labeled illustration, not verified index history — cumulative returns chosen to resemble a strong six-year bull run. Stocks +190% cumulative, bonds +7% cumulative, no contributions, starting from a 90/10 target.
- Starting portfolio: $90,000 stocks / $10,000 bonds = 90/10
- Stocks after the run: $90,000 × 2.90 = $261,000
- Bonds after the run: $10,000 × 1.07 = $10,700
- New total: $271,700
- New stock weight: $261,000 ÷ $271,700 = 96%
- The 90/10 portfolio is now a 96/4 portfolio — nobody decided that.
For an early-career physician at 90/10, six points of drift is survivable. The same mechanism is not survivable everywhere: a 60/40 portfolio that drifts to 75/25 across a long bull run turns a 35% equity decline from a 21% portfolio loss into a 26% portfolio loss — a difference measured in years of contributions for someone close to retirement. And drift compounds with career stage in the wrong direction: the longer the bull market runs, the closer you are to needing the money and the more equity risk you are unintentionally carrying. Your target allocation encodes a decision about how much decline you can absorb; the process for making that decision is worked in the asset allocation module. Drift silently unmakes it.
Two triggers work: the calendar and the 5/25 bands
Every workable rebalancing rule is one of two types, and both are defensible.
The calendar trigger. Pick one date — your birthday, your contract anniversary, the week after your CME conference — and rebalance annually on that date if allocations have drifted. Research on rebalancing frequency supports annual as a sensible default: often enough to contain drift, infrequent enough to keep costs and taxes near zero. The date is arbitrary; the consistency is not.
The threshold trigger. Instead of watching the calendar, watch the drift. The most widely used convention is the 5/25 rule, popularized by financial author Larry Swedroe: rebalance an asset class when it moves 5 absolute percentage points from target, or, for allocations under 20%, when it moves 25% in relative terms — whichever band is tighter. A 60% stock target acts at 65% or 55%. A 10% REIT or international sleeve acts at 12.5% or 7.5%, because 25% of 10 is 2.5. The relative band exists so small allocations do not drift into irrelevance while waiting to move five whole points.
| Feature | Annual calendar | 5/25 threshold bands |
|---|---|---|
| What you monitor | Nothing until the date | Allocation, checked occasionally |
| Trades per decade | Roughly 5–10 | Often fewer; only when bands break |
| Responds to crashes mid-year | No — waits for the date | Yes — a sharp decline can trip a band |
| Effort | Lowest | Slightly higher |
| Failure mode | A big move goes unanswered for months | Checking becomes watching becomes tinkering |
A practical hybrid many physicians settle on: check once a year on a fixed date, but act only if a 5/25 band is broken. One calendar reminder, few trades, and a written answer for the crash year.
Where you rebalance decides whether it costs anything
The execution location matters more than the trigger. Selling an appreciated fund in a taxable brokerage account realizes capital gains; the identical trade inside a , , 457(b), , or realizes nothing. So the priority order is fixed:
- Rebalance inside first. Exchanges between funds in your 403(b) or IRA have no tax consequence. If your overall portfolio is stock-heavy, sell stock funds and buy bond funds inside the retirement account, even if the drift "happened" in taxable — the portfolio is one portfolio.
- In taxable, rebalance with new money, not with sales. Redirect contributions to the underweight asset class until the target is restored. Switch dividend reinvestment off and point the cash at whatever is underweight. This is where a physician income is a structural advantage: a household adding $60,000 to $100,000 a year in savings can correct most drift without selling a single appreciated share.
- Sell in taxable only as a last resort — and if you must, harvest losses against the gains where available; the mechanics pair naturally with tax-loss harvesting.
Key insight
A strong savings rate makes rebalancing nearly free. If annual contributions are large relative to the portfolio — true for most attendings in their first decade — new money alone can hold the allocation on target for years, and the "how often should I sell" question mostly answers itself: rarely, and almost never in taxable.
Which funds belong in which account type is its own decision — bonds and REITs are usually better housed in tax-advantaged space — and that asset-location layer is covered in the taxable account strategy module.
If you do not want the job, buy the automation
There is a fully legitimate way to never think about any of this: hold a fund that rebalances itself. A target-date fund maintains its stock/bond split continuously and shifts it along a glide path as the date approaches; a fixed-allocation balanced fund does the same without the glide path. Inside a 403(b) or 401(k), where there are no tax consequences to the fund's internal trading, a single target-date fund is a complete, self-rebalancing portfolio — and for a resident or new attending with a five-figure balance, it is frequently the best available answer. The trade-offs are real but modest: you accept the fund's glide path rather than your own, and you give up asset-location control across account types. Automation you will actually leave alone beats a sophisticated policy you will abandon in a bad quarter.
Important
Target-date funds belong in tax-advantaged accounts, not taxable ones. In a taxable account the fund's internal rebalancing and glide-path shifts generate distributions you cannot control or time, and you cannot separate its bond sleeve into more tax-efficient placement.
The real function is behavioral: the rule decides so you do not have to
Here is the part the frequency debate misses. In a rising market, rebalancing is a mildly annoying chore. In a falling market, it is a demand that you sell the asset that feels safe and buy the asset that is on fire — during a 30% decline, with headlines forecasting worse, a threshold rule will tell you to move money into stocks. Almost no one does that by feel. The physicians who rebalanced into the March 2020 decline were, almost without exception, following a rule written years earlier.
That is what the written rule is actually for. It converts the hardest decision in investing — acting against your own fear, in real time, with real money — into clerical work specified in advance. One page is enough: target allocation, trigger (annual date, 5/25 bands, or the hybrid), execution order (tax-advantaged first, new contributions in taxable), and a sentence stating that the rule applies in declines. A rebalancing rule you wrote in calm conditions is a decision your future self does not have to make in a drawdown — that, not any return bonus, is what it is worth.
Quick takeaway
Rebalance to control risk, not to chase return. Pick one trigger — an annual date, the 5/25 bands, or check-annually-act-on-bands. Execute inside tax-advantaged accounts and with new contributions in taxable. Or hold a target-date fund and let it do the whole job. Then write the rule down, because its real value is redeemed in the year you least want to follow it.
Common questions
How often is too often?
Monthly or quarterly rebalancing is more than a diversified stock-and-bond portfolio needs; research finds it adds trading cost and, in taxable accounts, tax drag without improving risk control relative to annual attention. If you are rebalancing more than once or twice a year, the activity is probably serving a desire to do something rather than the portfolio.
Should I rebalance in taxable if it means realizing capital gains?
Almost never as a first resort. Check whether the drift can be corrected inside retirement accounts or with new contributions — between those two, most households never need to sell in taxable at all. If a large one-time drift genuinely requires taxable sales, spread them across tax years and pair them with any available loss harvesting.
Do I need to rebalance a target-date fund?
No — internal rebalancing is the product. The mistake to avoid is holding a target-date fund alongside separate stock funds, which silently overrides the glide path you paid for. One-fund solutions work when they are actually the one fund.
What should I do about rebalancing during a crash?
Whatever your written rule says, and nothing more. If you use bands, a crash may trip them — that is the rule telling you to buy equities at lower prices, which is the entire design. If you use an annual date, you are permitted to do nothing until the date. Both are correct; improvising mid-decline is the only wrong answer.
What to do next
- Write down your target allocation in one line. If you cannot, start with the allocation decision itself before any rebalancing rule.
- Pick your trigger — annual date, 5/25 bands, or check annually and act on bands — and write it under the target.
- List your accounts and mark where rebalancing is free: every 403(b), 401(k), 457(b), IRA, and HSA.
- Redirect new taxable contributions and switch off dividend reinvestment in taxable so incoming cash can flow to whatever is underweight.
- Set the calendar reminder for your chosen date, and check your current drift today against your bands.
- Add one sentence to the rule: "This applies in declines." Sign it, figuratively or literally.
A rebalancing policy is a page you write once and consult yearly, and the protocol above works with or without us. This is education, not individualized financial advice.