There are a few rules in investing that have become so widely accepted that we rarely stop to ask why we follow them. Diversify. Keep costs low. Stay invested. Rebalance your portfolio.
The first three are relatively easy to defend. The fourth is more interesting.
Imagine that you build a portfolio with 60% in equities and 40% in bonds. A decade passes, equities perform exceptionally well, and the portfolio gradually becomes 70/30. Conventional portfolio theory tells you what to do next: sell some equities, buy some bonds and restore the original 60/40 allocation.
It sounds perfectly sensible. In many circumstances, it is. But there is an assumption hidden inside that simple transaction: the allocation you chose ten years ago is still the allocation you should own today.
That assumption deserves more attention than it usually receives.
The SEC’s investor guide to allocation, diversification and rebalancing describes both calendar-based and threshold-based approaches, while emphasizing transaction costs and taxes. Rebalancing is often presented as though it were routine portfolio maintenance, something equivalent to servicing a car or checking the pressure in your tires. But a portfolio is not a mechanical object slowly drifting out of alignment. Its weights change because the assets inside it produce different returns. When we rebalance, we are deliberately reversing part of that process by selling some of what has appreciated and buying some of what has lagged.
Sometimes that is exactly what we should do. Sometimes allowing the portfolio to drift may also be perfectly rational. The interesting question is therefore not simply how often we should rebalance.
It is why we are rebalancing in the first place.
What Rebalancing Actually Does
The strongest argument for rebalancing has surprisingly little to do with increasing returns. Its primary purpose is risk control.
Suppose an investor deliberately chooses a balanced portfolio because that combination of equities and bonds represents an acceptable compromise between growth and stability. If equities subsequently outperform bonds for many years, doing nothing gradually changes the nature of the portfolio. What began as a balanced strategy can eventually become predominantly an equity strategy.
Historical data show how substantial this drift can become. Vanguard examined a globally diversified 50/50 stock-and-bond portfolio from 1926 through September 2020. When rebalanced annually, its equity exposure averaged about 51% and finished the period at 49%. When never rebalanced, equities eventually represented 98% of the portfolio. The unrebalanced portfolio produced the higher annualized return—8.8% versus 8.0%—but also considerably higher annualized volatility, 13.3% compared with 9.7%.
That example captures something important. The unrebalanced portfolio did not necessarily become a better 50/50 portfolio. It effectively stopped being one.
Its higher return was accompanied by a completely different risk exposure.
This is why asking whether rebalancing improves returns can be misleading. If an investor allows equities to grow from half the portfolio to almost all of it, and equities subsequently outperform bonds over the period being measured, of course the final return may be higher. But we are no longer comparing two different maintenance strategies applied to the same risk profile. We are comparing portfolios that evolved into substantially different things.
Seen this way, rebalancing is less about squeezing additional performance from a portfolio and more about preserving the risk decision that created the portfolio in the first place. Vanguard describes the primary function of rebalancing in essentially these terms: keeping portfolio risk aligned with its intended target as market movements cause allocations to drift.
That is a much less glamorous objective than “buy low, sell high,” but probably a more useful one.
There is, however, another side to the story.
When we rebalance, we systematically sell assets that have risen relative to the rest of the portfolio and purchase assets that have fallen behind. Under the right conditions, this can create what researchers sometimes call a diversification or rebalancing return. Scott Willenbrock, for example, described how a rebalanced portfolio can capture incremental return precisely because the process forces the investor to sell relative winners and buy relative losers, whereas a buy-and-hold portfolio increasingly concentrates itself in the assets that perform best.
But the words “under the right conditions” matter.
Rebalancing benefits from volatility between assets and, importantly, from some degree of reversal in their relative performance. If two assets repeatedly move away from and back toward one another, systematically selling one after it rises and buying it after it falls can be useful. If instead one asset enters a powerful and persistent long-term trend, repeated rebalancing can work in the opposite direction. You keep trimming the winner while continuously adding money to the laggard.
This is why rebalancing is not a free return generator. It is a discipline imposed on a portfolio, and every discipline has trade-offs.
The Strange Obsession With the Calendar
Once an investor decides to rebalance, another question appears: when?
The traditional answers are wonderfully precise for something that has no naturally correct date. Rebalance quarterly. Every six months. Once a year. Perhaps every January.
But markets do not know that December has ended.
If a portfolio remains almost perfectly aligned with its targets for twelve months, there is little economic reason to trade simply because the calendar says it is time. Conversely, if a violent market movement dramatically changes the portfolio in three months, waiting another nine months simply because the annual rebalancing date has not arrived can seem equally arbitrary.
Research offers some useful perspective here. Vanguard’s earlier work using broad U.S. stock and bond data from 1926 through 2009 found no meaningful difference in risk-adjusted returns between monthly, quarterly and annual rebalancing, while more frequent rebalancing naturally produced more transactions and therefore potentially greater costs. The conclusion was not that one magical frequency existed, but that relatively infrequent monitoring combined with reasonable tolerance bands could provide a sensible balance between risk control and costs.
More recent Vanguard research reached a similarly interesting conclusion from another direction. In simulations comparing multiple calendar and threshold strategies, annual rebalancing performed well enough that Vanguard described it as a reasonable recommendation for many investors, with estimated advantages over more frequent strategies measured in relatively modest amounts—roughly 6 to 28 basis points depending on the portfolio and comparison.
Those numbers are useful precisely because they are not spectacular.
Investors can spend enormous amounts of intellectual energy optimizing whether to rebalance every three months, every twelve months or after a particular percentage deviation. Yet the differences between reasonable approaches may be much smaller than the behavioral damage caused by constantly interfering with the portfolio.
There is nevertheless an appealing alternative to calendar rebalancing: tolerance bands.
Instead of saying that the portfolio must be restored every December, an investor allows assets to move within predefined boundaries. A 60% equity target might be allowed to drift until it crosses a certain threshold before any action is taken. The portfolio is monitored periodically, but a review does not automatically produce a trade.
Conceptually, I find this more satisfying because the decision is triggered by something happening in the portfolio rather than something happening on the calendar.
Gobind Daryanani explored a more active version of this idea in his research on “opportunistic rebalancing.” His argument was essentially to look relatively frequently but trade relatively infrequently, using wider bands and acting when meaningful deviations actually occur. His simulations suggested that opportunistic approaches could capture more of the potential rebalancing benefit than simple annual resets, although—as with any historical or simulated strategy—the result depends on the behavior of the assets and cannot be treated as a permanent law of markets.
The principle, however, is interesting: monitoring and acting are not the same thing.
A portfolio can be reviewed without being changed.
That distinction alone would probably improve the behavior of many investors.
What If We Didn’t Sell Anything?
There is another form of rebalancing that receives less attention because it feels almost too simple: use the cash already entering the portfolio.
If equities have risen and bonds have become underweight, the investor does not necessarily need to sell equities. New savings, bond interest, dividends or other portfolio cash flows can simply be directed toward the underweight component.
This matters particularly in taxable accounts. Selling appreciated assets can create taxes, transaction costs and administrative friction. Using new contributions to gradually restore balance can accomplish much of the same risk-management objective without immediately realizing gains.
It also changes the psychology of the process.
Instead of repeatedly cutting back successful investments, the investor is deciding where the next coin should go.
That is a subtly different way of thinking about portfolio management and one that fits particularly well during the accumulation phase, when regular contributions can be substantial relative to the size of the portfolio.
As wealth grows, however, contributions become smaller relative to existing assets. A €10,000 annual contribution can materially rebalance a €100,000 portfolio; it has almost no effect on a €5 million portfolio. Eventually, meaningful drift may require actual sales if maintaining a specific allocation remains the objective.
And that brings us to the question I find more interesting than the mechanics of rebalancing itself.
What exactly are we trying to maintain?
Maybe the Original Allocation Is the Thing That Needs Rebalancing
Consider again the investor who started with 60% equities and 40% bonds ten years ago.
Why 60/40?
Perhaps it reflected the investor’s age, income, tolerance for losses, available bond yields, future spending plans and total wealth at that particular moment.
Ten years later, all of those things may have changed.
The investor may have accumulated far more capital than expected. Retirement may be closer. Future liquidity needs may be clearer. Bond markets may offer very different opportunities. The investor’s ability to tolerate risk may have increased because the portfolio is now much larger—or decreased because employment income is about to disappear.
Automatically restoring the old 60/40 allocation assumes that the answer remains correct even though the question may have changed.
This is where I think we should distinguish rebalancing a portfolio from redesigning a portfolio.
Rebalancing says: my desired system remains correct, but markets have moved me away from it.
Redesigning says: my life, opportunities or objectives have changed, so perhaps the system itself should evolve.
Confusing the two can create strange outcomes. An investor might religiously rebalance to an allocation created twenty years earlier without ever reconsidering whether that allocation still serves a useful purpose. At the opposite extreme, another investor might call every tactical market opinion a “redesign” and continuously change the portfolio based on forecasts.
Neither is particularly attractive.
A sensible investment system needs enough discipline to resist unnecessary changes, but enough flexibility to recognize when the underlying objectives genuinely change.
This is also where a system-based approach differs from traditional percentage-based portfolio management. If every component of a portfolio has a defined job, the relevant question is not necessarily whether its percentage has moved away from an arbitrary target. The more useful question may be whether it can still perform the job for which it exists.
If bonds are intended to provide future liquidity, for example, perhaps the quantity that matters most is not whether they represent exactly 30% of the portfolio, but whether the planned future liquidity remains adequately funded. If equities exist to provide long-term growth, allowing them to become a larger percentage after years of successful compounding may not automatically constitute a problem. It becomes a problem when that increased exposure creates a level of risk inconsistent with what the overall system is designed to tolerate.
That distinction takes us beyond rebalancing as a mechanical ritual.
The Chill Perspective
So, should you rebalance?
Probably—but not because a textbook says that portfolios should always return to their original percentages, and certainly not because December has arrived.
Rebalancing makes the most sense when we understand it for what it primarily is: a method of controlling portfolio drift and keeping risk consistent with an intentional design. Historical evidence shows that refusing to rebalance can allow a diversified portfolio to evolve into something radically different, while research also suggests that obsessing over the perfect rebalancing frequency adds remarkably little value.
The interesting middle ground is therefore not between “rebalance” and “never rebalance.” It is between disciplined maintenance and unnecessary tinkering.
New contributions and portfolio cash flows can often correct small deviations without selling anything. Wider tolerance bands can allow markets to move without demanding constant intervention. Periodic reviews can determine whether risk has drifted meaningfully rather than automatically generating trades. And occasionally the review may reveal something more important: not that the portfolio has moved away from the plan, but that the plan itself no longer reflects the investor’s life.
That may be the most important distinction of all.
A portfolio should not remain unchanged merely because we once wrote some percentages on a spreadsheet. Nor should it be continuously modified because markets are moving. The purpose of an investment system is to provide enough structure that we do not need to make decisions every time prices change, while still allowing deliberate changes when our actual circumstances require them.
Rebalancing, viewed this way, is neither a source of magical excess returns nor an outdated ritual. It is simply one tool for keeping risk where we intended it to be.
Rebalancing also raises a broader question about how much complexity an investor actually needs. A portfolio can become increasingly difficult to manage when every adjustment introduces another rule, another allocation target, or another reason to intervene. The relationship between simplicity and effective portfolio construction is explored in Why Most People Overcomplicate Their Portfolio, where the emphasis is on understanding the purpose of each investment decision rather than accumulating additional layers of complexity.
And perhaps that is the most Chill way to approach it: review occasionally, trade reluctantly, use new money intelligently, and never confuse activity with improvement.
Leave a Reply