How to Rebalance Your Portfolio in 2026 (Without Creating a Tax Mess or Doing It Every Quarter)

Last week I fixed a portfolio that had drifted almost ten points off where I set it. I didn't sell a single share. I didn't pay a cent in tax. The whole thing took about as long as a coffee.

Then I remembered the advice I used to follow. Rebalance every quarter, trim the winners, top up the laggards, repeat forever. And I thought about how much pointless tax and fiddling I'd signed up for over the years without ever asking whether it worked.

So here's how to rebalance your portfolio in a way that a normal person can actually live with. Most of the advice on this is either wrong or aimed at someone managing a pension fund, not someone autoinvesting a few hundred a month. The good news is that the version that actually matters is almost embarrassingly simple, and the single most useful move costs you nothing.

A couple weeks back I wrote about opening the hood on my own "diversified" index fund and finding roughly 37% of it sitting in ten companies. That post was the diagnosis. This one is the treatment.

Rebalancing, minus the jargon

You pick a target. Say 80% stocks, 20% bonds, or in my case a global stock core with a slice of something calmer around it. Time passes. The winners grow faster than everything else, so the thing that did well becomes a bigger and bigger share of the pie without you touching a button. That's drift.

Rebalancing is just dragging the pie back toward the shape you chose. That's the entire concept. If your 80/20 has quietly become 88/12 because stocks ran hot, you top up the 12 until it's 20 again, and you're back to the risk level you actually signed up for.

Notice what rebalancing is not. It's not a way to make more money. It's not market timing with extra steps. It's a way to stop your portfolio from turning into something riskier than you meant it to be, one good year at a time.

Why "rebalance every quarter" is mostly bad advice

This is the part that gets people, because it sounds responsible. Every three months, log in, tidy up, feel like a diligent adult.

Here's what the research actually says. Vanguard has looked at rebalancing frequency more than once, and the finding is consistent and slightly boring: checking more often does not give you better risk-adjusted returns than checking once a year or using drift bands. What it does do is rack up more trades, more transaction costs, and, if any of this is in a taxable account, more tax. You're paying real money for the feeling of being busy.

There's an old idea floating around called the "rebalancing bonus," the notion that the act of trimming and topping up somehow prints extra return. A financial writer named William Bernstein popularized a version of it back in 1996, and it's been quietly dismantled since. Treat it as a myth, or at best an accident of specific historical periods. You do not rebalance to get richer. You rebalance so a five-alarm market doesn't do more damage to you than you can stomach.

Once that clicks, the whole "every quarter" ritual falls apart. If the point is risk control, not returns, you only need to act when your risk has actually drifted. The calendar has no idea what your portfolio is doing.

Why 2026 is the year drift finally bites

For most of the last decade you could ignore all of this and be fine, which is exactly why so many of us did. But the market quietly changed shape under everyone's feet.

The ten biggest companies in the S&P 500 now make up around 40% of the entire index — the most top-heavy it's been since roughly 1972. Rewind to 2015 and the top ten were about 19%. By 2020 they were around 27%. The concentration has basically doubled in a decade, driven by a handful of mega-cap tech names.

Data: AInvest / RBC Wealth Management, "The Great Narrowing" (2026).

If you own a plain S&P 500 or total-market fund, this is your portfolio whether you chose it or not. The Information Technology sector alone is about 30% of the index. That's the mechanism nobody warns you about: you didn't buy a concentrated tech bet, but drift built one for you. I went deep on this in why your index fund isn't as diversified as you think, and it's the same handful of stocks behind the whole Magnificent Seven and AI bubble argument. The point for today is narrower: this is the year "I'll rebalance eventually" stopped being harmless.

But did the concentration even hurt? (the honest part)

I'm not going to pretend rebalancing is a free lunch, because that's exactly the kind of thing that makes these articles worthless.

Here's the uncomfortable truth. Over the last ten years, the concentrated, cap-weighted S&P 500 crushed its more diversified cousin. The regular SPY returned about 251% over the decade. The equal-weight version, RSP, which spreads your money evenly instead of piling it into the giants, returned about 207%. If you'd religiously rebalanced away from the winners the whole time, you'd have left a lot of money on the table.

Data: 24/7 Wall St, "RSP vs SPY" (2026).

Then 2026 showed up. Through the first half of the year the equal-weight index is ahead, up about 9.7% versus 8.4% for the cap-weighted one, as the market finally broadened out beyond the mega-caps. That's the whole argument for rebalancing in one chart. Concentration paid beautifully, right up until it didn't, and nobody rings a bell at the top.

So rebalancing is insurance. Insurance costs something when the house doesn't burn down. You're not doing it because trimming winners makes you money — history says it often costs you money. You're doing it because you don't actually know whether it's 2015 or the top, and a 40%-in-ten-stocks portfolio is a much bigger bet than most people realize they're making.

Threshold bands: the only rule I actually use

If the calendar is the wrong trigger, what's the right one? Drift itself. You set a band, and you only act when the portfolio crosses it.

The version I like comes from Larry Swedroe and gets called the 5/25 rule. It sounds fussier than it is:

For any big chunk of your portfolio (anything that's 20% or more of the target), you rebalance when it drifts 5 percentage points. So a 60% stock target gets touched only when it hits 65% or drops to 55%. For smaller slices, you use a relative 25% band instead. A 15% position gets rebalanced when it moves about a quarter of its own size, roughly plus or minus 3.75 points.

Vanguard's own research landed somewhere similar with what they call a 200/175 approach — trigger a rebalance when you've drifted 2 percentage points, and when you do, only pull it back to within 1.75. The exact numbers matter less than the principle: pick a band, write it down, and let the portfolio sit inside it without touching it. Their threshold approach generated roughly a quarter of the trading costs of rebalancing every month, for the same risk control.

I check maybe twice a year. If nothing's outside its band, I close the tab and do nothing, which is the most underrated move in investing.

Rebalance with new money before you sell anything

Now the part that saved me from that tax bill, and the single most useful thing in this whole article.

Most people picture rebalancing as selling — dump some of what's grown, buy what's lagged. But if you're still adding money every month, you have a quieter tool that triggers no tax at all: point your new contributions at whatever's underweight. If your stock sleeve has ballooned and your bonds have shrunk, send the next few months of deposits into bonds until the shape is right again. You never sell the winner, so there's no capital gain to tax.

That's what I did last week. My equity side had drifted, so instead of selling, I just redirected my autoinvest for a while. Same result, zero tax, zero drama. If you reinvest dividends automatically, you can do the same thing: steer them to the underweight side instead of piling back into the thing that's already too big.

When new money isn't enough to close the gap and you genuinely have to sell, do the selling inside a tax-advantaged account (a 401(k), an IRA, whatever your country's sheltered wrapper is), where trades don't trigger a tax event. Selling appreciated shares in a plain taxable brokerage account is the expensive way, and it's the default that costs people hundreds or thousands for no reason. Picture trimming $10,000 of a fund that's up 60%; that's a $6,000 gain you just made taxable, purely to reshuffle money you weren't spending. Do that inside the sheltered account instead and the number is zero.

Rank it like this: new contributions first, then dividends, then selling inside tax-advantaged accounts, and only as a last resort selling in a taxable account. Most accumulators never get past the first two.

My dead-simple rebalancing routine

Stripped all the way down, here's the whole thing:

Twice a year, I open a view that shows what I actually own all the way through, not just "one world ETF" but the real breakdown underneath it, because that's where the concentration hides. The portfolio breakdown tool I built into MFFT does this for me, but a spreadsheet works too. Then I check each big piece against its band. If everything's inside, I close it and get on with my day. If something's breached, I fix it with new contributions first, and only sell if I have to, inside a sheltered account when I can.

That's it. No quarterly ritual, no spreadsheet with sixteen tabs, no watching the market. Twenty minutes, twice a year, and the most common outcome is that I do nothing at all.

If you're still choosing the core holding you'll rebalance around, it's worth getting that decision right first. I compared the usual suspects in VOO vs VTI, and any of the sane world ETFs will do the job. Rebalancing only makes sense once you know what you're rebalancing toward.

When not to bother

Plenty of people reading this don't need to do anything.

If your money is in a target-date fund or an all-in-one fund, it rebalances itself internally, and adding manual trades on top just creates tax and mistakes. If your portfolio is small, say you're early in the grind toward your first serious balance, drift barely matters yet and your savings rate is doing virtually all the work anyway. And if you're inside your bands, the correct action is genuinely nothing.

The mistake isn't rebalancing too little. For most people it's rebalancing too often, too anxiously, and in the most expensive account they own. Pick a target you can live through a bad year with. Set a band. Steer new money toward whatever's fallen behind. Look twice a year and be pleasantly bored most of the time.

I still enjoy my job, and I still autoinvest on the 1st without thinking about it. Rebalancing didn't change any of that. It just means that when the next reshuffle comes, and 2026 is starting to look like one, the portfolio I own is the one I actually chose, instead of the one a decade of quiet drift chose for me.