A portfolio left alone stops being the portfolio that was chosen. The holdings that rise fastest come to dominate it, and correcting that means selling exactly what has been working.
Drift is arithmetic, not neglect
If two assets are held equally and one doubles while the other is flat, the mix is no longer even. Nothing was done wrong; the weights simply moved with the prices.
Over several years of divergent returns, an allocation set deliberately at the start can end up concentrated in whichever asset class had the strongest run.
The drift is silent because nothing about the account looks unusual. The balance has grown, and the change in composition only appears if someone measures it.
What rebalancing restores is risk, not return
The case for correcting the weights is that the original mix encoded a tolerance for loss. Letting the riskiest holding grow unchecked quietly raises the exposure beyond it.
A portfolio that has drifted heavily toward equities will fall harder in a downturn than the one its owner agreed to hold.
Rebalancing is therefore best understood as risk maintenance. Any effect on returns is a by-product, and over some periods it reduces them.
The discipline runs against instinct
Selling the strongest position to buy the weakest is uncomfortable, because recent performance feels like information about the future.
The rule exists precisely to remove that judgment. It acts on the deviation from target rather than on any view about which asset will do better next.
Investors who abandon rebalancing usually do so in the middle of a strong run, which is the moment the correction would have mattered most.
Calendars and thresholds are the two triggers
One approach checks the portfolio at fixed intervals and restores the weights regardless of how far they moved. It is simple and easy to automate.
The other acts only when a holding drifts beyond a set band. It trades less in quiet markets and more when moves are large.
Threshold rules respond to what actually happened rather than to the calendar, at the cost of requiring the portfolio to be monitored more often.
Costs decide how often it is worth doing
Every correction incurs trading costs, and in a taxable account it may realise gains that would otherwise have stayed deferred.
Frequent rebalancing tightens the tracking to target but pays those costs repeatedly. Wide bands and long intervals accept more drift in exchange for fewer transactions.
Directing new contributions toward whichever holding is underweight achieves much of the same correction without selling anything, which is why regular savers often need fewer explicit trades.