Rebalancing: When and How

Rebalancing Won’t Make You Richer

Rebalancing does not make you money. It costs you a little, in taxes or trading friction, in exchange for keeping your risk where you set it. If you’re rebalancing because you think it’ll juice your returns, you’ve got the wrong job description, and you’re the kind of person who ends up doing it wrong: too often, at the wrong moments, selling winners out of habit instead of need.

The right job description: your portfolio has a risk level you picked on purpose. Left alone, markets slowly reassign that risk level without asking you. Rebalancing is the maintenance that puts it back. No alpha, no edge involved, just keeping the plan you already made.

If you haven’t picked a target allocation yet, that’s a different article. This one assumes you have one and just need to keep it.

What Drift Actually Looks Like

Say you start with $100,000 at 80% stocks, 20% bonds. Clean split: $80,000 in stocks, $20,000 in bonds.

Now give it a good run. Assume stocks return 15% a year and bonds return 3%, both just modeling assumptions, not a forecast, and let four years pass without touching anything:

Nobody did anything wrong. Stocks just had four good years and bonds had four boring ones, which is what bonds are for. But you didn’t sign up for 86/14. You signed up for 80/20. The gap between “what you picked” and “what you now have” is 6.1 percentage points, and it happened without a single decision on your part. That’s drift: not a mistake, just math compounding unevenly across two assets that don’t move together.

To get back to 80/20 on that $162,431 total, you need $129,945 in stocks and $32,486 in bonds. You’re currently sitting on $139,921 in stocks and $22,510 in bonds. So you sell $9,976 of stocks and buy $9,976 of bonds. The two numbers match exactly, because rebalancing doesn’t add or remove money, it just moves it.

That’s a real trade with a real number attached. Compare that to how most people actually experience drift: as a vague feeling that their statement “looks different” without ever computing what changed.

Calendar Rebalancing: Once a Year Is Plenty

The simplest policy is a fixed schedule. Pick a date, birthday, January 1st, tax day, whatever you’ll actually remember, and check your allocation against target on that date every year.

Annual is the right cadence. Monthly or quarterly rebalancing doesn’t protect you any better and actively costs more: more trades, more chances to trigger short-term capital gains (taxed at your ordinary income rate instead of the long-term rate), more chances to sell into a dip. Drift takes time to build. Checking every 30 days mostly gives you a series of tiny reads that fall inside normal market noise, and each one is a reason to fiddle with a plan that doesn’t need fiddling.

The failure mode with pure calendar rebalancing is the opposite problem: on your rebalancing date, drift might be 0.4 percentage points. Trading anyway, and eating the transaction costs or tax bill to fix a 0.4-point gap, is busywork dressed up as discipline.

Threshold Rebalancing: Only When It’s Actually Off

The other policy is band-based: don’t rebalance on a schedule, rebalance when an asset class strays more than some threshold from target. A common band is 5 percentage points, absolute or relative. Absolute means a 20% bond target that drifts to 25% or 15% triggers a rebalance. Relative means a percentage move, so a 20% target moving by more than 25% of itself (to 25% or 15%, coincidentally the same numbers here) triggers it.

The appeal is that you only trade when there’s something meaningful to correct, and you skip the years where nothing happened. In the drift table above, you’d have crossed a 5-point band sometime in year four, right when the stock/bond split hit roughly 85/15.

The downside of pure threshold rebalancing on its own: it requires you to actually check the numbers regularly enough to notice when you cross the line, and most people don’t build that habit without a date on the calendar to anchor it.

The Practical Default: Check Annually, Act Only Outside the Band

Combine them. Check your allocation once a year, on the same date every time, and only trade if you’re outside your band, say plus or minus 5 points on any asset class. If you’re inside the band, do nothing and check again next year.

This gets you the discipline of a fixed check-in without the busywork of trading on years where the drift is trivial. In the four-year example above, you’d have checked in years one, two, and three, seen drift of 1.7, 3.3, and 4.8 points, done nothing each time, and then acted in year four when the gap passed 5 points and pulled the trigger on that $9,976 trade. One trade in four years, and it was the trade that mattered.

The Tax Problem With Selling to Rebalance

Here’s where the “just sell the winner and buy the loser” advice runs into a wall: in a taxable brokerage account, selling that $9,976 of stock realizes a gain, and the IRS wants a cut. If those shares have been held over a year, you’re paying long-term capital gains rates, 15% or 20% for most people depending on income, on the gain portion of that sale, not the whole $9,976. Still, it’s a real tax bill for the privilege of moving your own money from one pocket to another.

Two tools fix this before you ever need to sell:

• Direct new contributions to the underweight asset. If you’re adding $2,000 a month and bonds are underweight, put all of it into bonds instead of splitting it 80/20 the way you split your existing balance. Against a $9,976 gap, five months of $2,000 contributions aimed entirely at bonds closes it without selling a single share of stock.
• Direct dividends and interest the same way. Most people have dividends set to auto-reinvest in the fund that paid them. Redirect the underweight fund’s dividends toward itself and, more usefully, redirect the overweight fund’s dividends toward the underweight one instead of letting them compound the imbalance.

Only sell when contributions and dividend redirection can’t close the gap fast enough, and when you do sell, sell inside tax-advantaged accounts first: a 401k or IRA has no capital gains tax at all, so a rebalancing trade there is free in a way it never is in a taxable account.

Asset Location: The Idea Next Door

This is a related but separate lever: which account holds which asset, not how much of each you hold. Bonds throw off interest taxed as ordinary income every year, so they’re a natural fit for a 401k or IRA where that income isn’t taxed as it’s earned. Stock index funds are naturally tax-efficient already (low turnover, mostly unrealized gains, qualified dividends), so they tolerate living in a taxable account better.

Get the location right and something nice falls out: your bonds, the asset most likely to need rebalancing trims after a stock rally, are sitting in the account where trading them costs nothing.

Treat Every Account as One Portfolio

Here’s the mistake that trips up people who are otherwise doing everything right: rebalancing each account separately instead of looking at the whole picture.

Say your 401k is 100% bonds (because that’s the only place you’ve put bonds, following the asset location logic above) and your taxable brokerage is 100% stock index funds. Look at either account alone and it seems wildly out of balance. Panic-rebalance the 401k by adding stocks, or the brokerage by selling stock to buy bonds, and you’ve just broken an allocation that was fine.

Add them together instead. $200,000 in the 401k (all bonds) plus $800,000 in the taxable brokerage (all stocks) is a $1,000,000 portfolio sitting at exactly 80% stocks and 20% bonds. Perfectly on target, held across two accounts that each look “wrong” in isolation. The target allocation applies to your net worth, not to each account individually. Once you start managing multiple accounts (401k, IRA, taxable, maybe a spouse’s accounts too), pull the balances into one spreadsheet and rebalance against the combined total, not account by account.

The Wash Sale Trap, If You’re Also Harvesting Losses

If you’re tax-loss harvesting (selling a fund at a loss to book the tax deduction, then buying a similar-but-not-identical replacement), rebalancing trades can quietly step on the wash sale rule. That rule disallows your loss if you buy the same or a “substantially identical” security within 30 days before or after the sale, in any account you own, including a spouse’s and including your IRA.

The collision happens like this: you harvest a loss in your taxable account by selling a total market fund and buying a similar fund. A week later, your annual rebalancing check tells you to buy more of that original fund in your 401k because bonds ran hot and you’re topping up equities. If the 401k purchase is close enough to count as the same security, the IRS disallows the loss you just harvested, and because it happened in a tax-advantaged account, you don’t even get a basis step-up to show for it. The loss just evaporates.

The fix is boring but effective: before executing any rebalancing trade, check it against any tax-loss harvest you’ve made (or plan to make) in any account in the last 30 days, not just the one you’re trading in.

Automatic Options: Let Someone Else Do It

Target-date funds and robo-advisors rebalance for you, automatically, inside the fund or account, with no capital gains consequence to you because the trades happen inside the fund wrapper. This is a real, non-trivial argument for using them. If the idea of tracking bands and directing contributions sounds like homework you’ll quietly never do, a target-date fund in your 401k does this exact job, continuously, for an expense ratio around 0.1 to 0.15%. A robo-advisor does the same in a taxable account, and the better ones also manage the tax-loss harvesting wash sale problem above programmatically.

Trading a little control and a little in fees for the guarantee that rebalancing actually happens is a completely reasonable trade for a lot of people. Doing it yourself is not morally superior, it’s just cheaper if you’ll actually stick with it.

When Not to Rebalance

• Drift under your band. If you’re checking annually with a 5-point band and you’re sitting at 82/18 against an 80/20 target, that’s noise. Leave it.
• The tax bill would exceed the risk you’re correcting. A 2-point drift on a highly appreciated taxable position can trigger a tax cost bigger than the risk reduction is worth. Run the actual numbers before selling, not a gut feeling.
• Trading costs eat the benefit. Less common now that most brokerages charge $0 commission on stock and ETF trades, but still check for bid-ask spread costs on thinly traded funds, and for account-specific redemption fees on some mutual funds if you sell within a short holding window.

The Annual Checklist

1. On your chosen date, pull every account’s balance, retirement and taxable, into one place.
2. Compute your combined stock/bond (and any other asset class) percentages against your target.
3. If everything’s inside your band (5 points is a reasonable default), stop here. You’re done for the year.
4. If something’s outside the band, close the gap with new contributions and redirected dividends first.
5. If that’s not enough, sell inside tax-advantaged accounts before touching taxable ones.
6. If you must sell in a taxable account, check the last 30 days for any tax-loss harvest that could turn into a wash sale.
7. Write down what you did and the date. Next year, do it again.

That’s it. Not exciting, not a growth strategy, just the thing that keeps the risk you signed up for from quietly turning into a different risk you didn’t.