Rebalancing Without Triggering a Tax Bomb: Four Ways to Fix a Drifted Taxable Portfolio

A taxable portfolio that has drifted from 60/40 to 72/28 is not a crisis. It is a tax problem wearing a portfolio problem’s clothes. If the account holds large embedded gains, the fastest way back to target can be the most expensive way back, because every sale can realize gains the IRS will tax at 0%, 15%, or 20%, plus a possible 3.8% net investment income tax ^[1][2].

That is why rebalancing in a brokerage account is different from rebalancing in an IRA. In taxable, the question is not just “what mix do I want?” It is “which tax bill do I want to pay now, and which one can wait?” Kitces has repeatedly argued that tax drag is a real, measurable headwind in taxable investing, especially when turnover and distributions are ignored ^[5].

A 72/28 portfolio is usually a tax problem first, not a risk problem

Start with the math. A 60/40 portfolio that has drifted to 72/28 has not merely become “a little more aggressive.” It has roughly 20% more equity exposure than intended, and that changes drawdown behavior in a way investors feel quickly. If stocks fall 30%, a 72% stock portfolio loses about 21.6%; a 60% stock portfolio loses 18.0%. That 3.6-point gap is not abstract. It is the difference between discomfort and panic for many households. For a plain-language refresher on how allocation drives outcomes, see asset allocation and drawdowns.

But in taxable accounts, the bigger issue is that the portfolio may be sitting on years of unrealized gains. Selling the winner to buy the laggard can trigger a tax bill immediately. IRS Publication 550 explains that capital gains are generally realized when you sell, and the tax rate depends on holding period and income level ^[1]. Long-term gains are taxed at preferential rates, but “preferential” does not mean free. For many investors, the federal rate is 15%, and high earners can face 20% plus the 3.8% NIIT ^[2].

That is the uncomfortable implication: a portfolio can be too risky and too tax-inefficient at the same time. Most investors treat those as separate problems. They are not.

That table is the reason the rest of this piece exists. Rebalancing is easy in theory. In taxable, it is a sequence problem.

The 60/40-to-72/28 case: four paths, one portfolio, very different tax bills

Assume a taxable account worth $1,000,000. Five years ago it was 60/40. Today it is 72/28 because stocks outperformed bonds and the investor made no meaningful rebalancing. The stock sleeve is now $720,000 and the bond sleeve $280,000. To return to 60/40 without adding new money, the investor would need to move $120,000 from stocks to bonds.

Now assume the stock lots sold have a 50% embedded gain ratio. That means $60,000 of the $120,000 sale is gain and $60,000 is basis. If the investor is in the 15% long-term capital gains bracket and owes no NIIT, the immediate federal tax on that sale is about $9,000. If NIIT applies, the bill rises to $11,280. That is not a rounding error ^[1][2].

Here are four ways to deal with the drift. None is magic. Each trades speed for tax efficiency.

*Assumes long-term gains taxed at 15% or 15% plus 3.8% NIIT. This is an illustrative calculation, not audited performance. It ignores state taxes and assumes a 50% embedded gain ratio on the shares sold.

The point is not that one number is “right.” The point is that the tax bill changes the ranking of the strategies. A clean rebalance is often the most expensive rebalance.

For readers who want the mechanics of how rebalancing itself can help or hurt, AIBROKER’s rebalancing guide and rebalancing bonus myth piece are useful companions.

Technique 1: Use new contributions to buy the laggard and let time do the work

This is the cheapest tool in the box. If you are still adding money, direct every new dollar to the underweight asset until the portfolio drifts back toward target. No sale. No realized gain. No tax bill. Vanguard’s research on dollar-cost averaging is often discussed as a market-timing question, but the same cash-flow logic applies here: new money can repair an allocation without forcing a taxable event ^[6].

In the case study, suppose the investor contributes $20,000 this year. Buying bonds with that cash moves the portfolio from 72/28 to roughly 71.2/28.8. That is not dramatic, but it is free. Add another $20,000 next quarter and the drift narrows again. The catch is obvious: if the portfolio is large and the contribution is small, this method is slow. Very slow.

That slowness is not a flaw. It is the price of avoiding realized gains. Investors often overestimate how much rebalancing precision they need and underestimate how much tax they are paying for it. If you are still in accumulation mode, this is usually the first lever to pull. If you are retired and withdrawing, it may not be available at all.

That table is why contribution routing is underrated. It is boring. It works. And it keeps the IRS out of the room.

For investors building a repeatable process, AIBROKER’s automatic investing and lump sum vs. dollar-cost averaging pieces connect naturally to this step.

Technique 2: Redirect dividends, but do not pretend dividends are free money

Dividend redirection is the second low-friction tool. If the stock sleeve throws off dividends, you can send them to bonds instead of reinvesting them into more stock. That nudges the portfolio toward target without selling appreciated shares. It also avoids the false comfort of “income” thinking. A dividend is not a bonus; it is a distribution that changes the composition of your account and may be taxable in the year received ^[1].

Qualified dividends are generally taxed at long-term capital gains rates, while nonqualified dividends are taxed as ordinary income ^[1]. That matters because the tax treatment can be better than selling appreciated shares, but it is not zero. Kitces has argued that taxable investors should think in terms of after-tax compounding, not headline yield, because tax drag can quietly eat the spread between a good asset allocation and a bad one ^[5].

Suppose the portfolio generates a 2% dividend yield on the $720,000 stock sleeve, or $14,400 a year. Redirecting that cash to bonds reduces the stock weight a bit and costs no extra trading tax. But the dividend itself may still be taxable. If the dividend is qualified and the investor is in the 15% bracket, the federal tax is about $2,160. If it is nonqualified, the bill can be much higher ^[1][2].

Most investors get this wrong in one of two ways. They either reinvest dividends automatically and then complain that the portfolio drifts, or they treat dividend redirection as a tax shelter. It is neither. It is a control knob.

For a broader discussion of how distributions affect total return, AIBROKER’s dividend investing article is the right companion.

Technique 3: Tax-loss harvesting pairs can finance rebalancing without adding new tax

Tax-loss harvesting is the most powerful tool here, but only when the portfolio has something to harvest. Sell a position at a loss, buy a similar but not substantially identical replacement, and use the realized loss to offset gains elsewhere. IRS rules and Publication 550 make the wash-sale restriction the main trap: if you buy the same or substantially identical security within 30 days before or after the sale, the loss is disallowed for now ^[1].

That means the technique is not “sell anything that is down.” It is “harvest a loss without breaking the wash-sale rule.” In practice, investors often pair broad-market ETFs with close substitutes. For example, one U.S. large-cap ETF can be swapped for another with a different index provider, or a total-market fund can be paired with a similar fund that tracks a different benchmark. The exact substitute matters. A sloppy substitute can create tracking error or a wash sale. For a deeper mechanics discussion, see AIBROKER’s tax-loss harvesting guide and ETFs vs. mutual funds.

In the 72/28 case, imagine the investor has a $15,000 unrealized loss in an international equity sleeve or a bond fund. Harvesting that loss can offset part of the $60,000 gain needed for a full rebalance. If the investor realizes a $15,000 loss and a $60,000 gain in the same year, the net taxable gain falls to $45,000. At a 15% long-term rate, that saves about $2,250 in federal tax; with NIIT, the savings are larger ^[1][2].

The catch is that tax-loss harvesting is not free alpha. It is tax deferral, not tax elimination. If the replacement security later rises, the deferred gain comes back. Still, deferral has value because money paid to the IRS today cannot compound for you tomorrow. That is the whole point of tax drag.

Do not harvest losses just to feel active. Harvest them when they can offset gains you were going to realize anyway, or when they create a bank of losses for future rebalancing.

That is the hidden tradeoff: the best tax-loss harvest is often the one that supports a rebalancing plan you already needed.

Technique 4: Partial rebalance with a capital-gains budget is the adult compromise

Partial rebalancing is what disciplined taxable investors often end up doing after they stop pretending taxes are optional. You set a capital-gains budget for the year, then sell only enough appreciated shares to stay inside it. That may mean moving from 72/28 to 68/32 this year, then 65/35 next year, and so on. It is slower than a full reset. It is also usually smarter.

Suppose the investor is willing to realize only $30,000 of gains this year. At a 15% long-term rate, the federal tax is about $4,500; with NIIT, about $5,640 ^[2]. If the investor sells $60,000 of stock with a 50% gain ratio, that is the tax cost. The portfolio does not return all the way to 60/40, but it moves meaningfully closer. The investor has bought risk reduction without lighting the whole gain pile on fire.

This is where a capital-gains budget becomes a real planning tool. It forces a tradeoff that many investors avoid naming. You can have faster rebalancing or lower taxes, but not both. Pick one. If you want both, you need losses, new contributions, or time.

Kitces’ work on tax drag is relevant here because the drag is not just the tax rate; it is the timing. A tax paid today is more expensive than the same tax paid later, because the deferred dollars keep compounding ^[5]. That is why a partial rebalance can beat a full rebalance on after-tax wealth even when the full rebalance looks cleaner on paper.

That table is the practical heart of the article. The budget is not a tax dodge. It is a throttle.

Which sequence usually wins: a simple decision tree for taxable rebalancing

Most investors should not choose one technique and ignore the others. The better move is to sequence them. Start with the cheapest source of drift correction, then move to the next cheapest, and only then sell appreciated shares. That sequence usually looks like this: new contributions first, dividend redirection second, tax-loss harvesting third, partial rebalance with a gain budget last.

Here is a simple decision tree for the 72/28 case:

1. Are you still contributing? If yes, route new cash to bonds until the gap narrows.
2. Do you receive dividends or interest in taxable? If yes, redirect those flows to the underweight sleeve.
3. Do you have unrealized losses in similar holdings? If yes, harvest them and use the losses to offset gains.
4. Do you still need to reduce risk? If yes, sell appreciated lots only up to your capital-gains budget.

This sequence is not elegant. It is effective. And it respects the fact that taxable rebalancing is constrained by lot selection, holding periods, and the wash-sale rule ^[1].

For investors who want to think more systematically about portfolio maintenance, AIBROKER’s monthly review process and three numbers that matter pieces fit well with this framework. The point is to make rebalancing a process, not a mood.

One more judgment call: if your taxable account is large and your embedded gains are huge, a full return to target may be less important than a stable, repeatable tax policy. Investors who chase perfect allocation often end up with perfect tax bills.

After-tax outcomes: what changes if you wait, harvest, or sell now?

The table below compares the four techniques using the same drifted portfolio. The numbers are illustrative, but the ranking is realistic. The cheapest path is usually the one that avoids realizing gains. The most expensive path is usually the one that ignores them.

There is no free lunch here. There is only a less expensive lunch. If you need to restore risk quickly, you may have to pay some tax. If you can wait, you should. That is the uncomfortable truth most “rebalance now” advice skips.

For investors who want to understand how taxes affect long-run compounding, the broader literature on tax-efficient investing is worth reading alongside this piece. The IRS rules are the floor. The real planning work starts above that floor ^[1][2][5].