Rebalancing: When, How, and Whether It's Worth It

A portfolio that started at 80% stocks and 20% bonds doesn't stay there on its own. After a good few years for stocks, that same portfolio might drift to 88/12 without you doing anything, simply because the stock portion grew faster. Rebalancing is the act of selling some of what's grown and buying more of what hasn't, to bring the mix back toward your original target. It sounds like a small mechanical chore. Whether and how you do it has real consequences, especially once taxes enter the picture.

Why drift happens, and why it matters

Different asset classes grow at different rates over any given stretch of time. Left alone, a portfolio's actual mix quietly shifts away from the one you originally chose, and that original mix was presumably chosen for a reason tied to your asset allocation and risk tolerance. An 80/20 portfolio that's drifted to 88/12 isn't wrong, exactly, but it's now carrying more stock-market risk than you signed up for, and you may not notice until a downturn hits and the drop feels larger than expected, because the portfolio actually is riskier than it was.

Threshold-based rebalancing

Under this approach, you rebalance whenever an asset class drifts a set number of percentage points from its target, say 5. Hit that band in either direction, and you trade back to target regardless of the calendar date. This responds directly to the thing you actually care about (drift), and it can mean going years without touching anything if markets stay roughly in line with your target, or rebalancing several times in a single volatile year if they don't.

Calendar-based rebalancing

Under this approach, you check and rebalance on a fixed schedule, commonly annually, sometimes semi-annually or quarterly, regardless of how much drift has actually occurred. It's simpler to remember and to automate, and it avoids the temptation to "just wait a little longer" that a pure threshold approach can invite when markets are moving in a favorable direction. The trade-off is that you might rebalance a trivial 1% drift on schedule, or let a large drift run for months before your next scheduled check.

Many long-term investors use a hybrid: check on a calendar schedule, but only actually trade if drift exceeds a threshold at that check-in. That combines the simplicity of a fixed date with the discipline of not trading over noise.

A small illustrative example

Say a target allocation is 70% stocks, 30% bonds, on a $100,000 portfolio: $70,000 and $30,000 respectively. After a strong year for stocks, the mix has drifted to $80,000 stocks and $28,000 bonds (a $108,000 portfolio now sitting at roughly 74/26). Under a 5-point threshold, that drift hasn't yet crossed the band and you'd leave it alone. If stocks kept climbing and the mix reached $88,000 stocks and $27,000 bonds, closer to 77/23, that's crossed the threshold, and you'd sell enough stock (or direct enough new bond purchases) to bring the split back toward 70/30. The point of the exercise isn't the specific numbers; it's that the trigger is the drift itself, not a date on the calendar.

Treat your household as one portfolio, not several

If you and a spouse hold accounts separately, or you have money spread across a 401(k), an IRA, and a taxable brokerage account, it's worth evaluating your allocation across all of them together rather than trying to hold 70/30 inside every single account. This is also where the tax-account ordering matters most: it's often more efficient to hold the bond portion inside tax-advantaged accounts and let a taxable account run more heavily toward stocks, then do the actual rebalancing trades wherever they're cheapest from a tax perspective, rather than mechanically rebalancing each account to the same target in isolation.

The tax problem rebalancing creates in a taxable account

Here's the part that a purely mechanical description of rebalancing skips over. Selling an appreciated asset to rebalance realizes a capital gain in a taxable account, and per IRS Topic 409, that gain is taxable in the year it's realized, at rates that depend on your income and how long you held the position. Rebalance a taxable account every year on a fixed schedule, selling winners each time, and you can generate a recurring tax bill purely from maintaining your target allocation, on top of whatever you'd owe from other activity.

This isn't a reason to avoid rebalancing altogether; an allocation that's drifted far from target carries real risk of its own. It's a reason to rebalance taxable accounts more deliberately than tax-advantaged ones, where trades inside a 401(k) or IRA don't trigger this problem at all.

A practical order of operations

If you're still contributing to your accounts, the first tool to reach for in a taxable account, before considering an outright sale, is directing new contributions toward whichever asset class has fallen below target rather than selling the one that's grown. This achieves the same rebalancing effect without realizing any gains at all.

Rebalance inside tax-advantaged accounts first and most freely, since trades there don't create a tax event. In taxable accounts, use new contributions and dividend reinvestment to correct drift before selling anything. If a taxable account still needs a sale to get back on target, consider pairing it with tax-loss harvesting elsewhere in the portfolio in the same year, so a realized loss can offset the gain from rebalancing rather than the two events happening independently and each showing up separately on your return.

When it's genuinely not worth doing

A small amount of drift, a few percentage points, rarely justifies transaction costs and tax consequences on its own; that's the entire logic behind using a threshold in the first place rather than trading on every tiny move. And if your entire portfolio sits inside tax-advantaged accounts, most of the tax argument above simply doesn't apply, and you can rebalance as often as you'd like without a tax cost, though transaction costs and simplicity still argue for not overdoing it.