A portfolio built to a fixed mix does not stay at that mix, because the assets inside it move at different speeds. Rebalancing bands are the rule that decides when the drift has gone far enough to act on.
Drift is a mechanical consequence, not a decision
If one holding rises faster than another, its share of the total grows without anyone buying more of it. The mix shifts even though the investor has done nothing at all.
Left alone for long enough, a portfolio ends up dominated by whatever has performed best. The risk profile changes quietly, and it changes in the direction of whatever has already run.
That matters because the original mix was chosen to express a tolerance for loss. Drift means the portfolio no longer expresses the tolerance it was built around.
A band is a threshold, not a date
Calendar rebalancing checks the portfolio on a schedule regardless of what happened. Band rebalancing checks a condition instead, acting only when a holding moves outside a stated range around its target.
A band is usually written as an absolute or relative width. An allocation with a five point absolute band is corrected when it sits five percentage points above or below where it was set.
The difference is that a band responds to market movement rather than to the passage of time. Quiet stretches trigger nothing, and violent ones trigger action promptly.
Band width is a trade between two costs
Narrow bands hold the mix very close to target, but they fire often, and every correction involves transaction costs and, in a taxable account, realized gains.
Wide bands trade less and allow more drift between corrections. The portfolio spends more of its life away from the intended mix in exchange for fewer transactions.
Neither setting is correct in the abstract. The width encodes how much drift the owner is willing to carry against how much trading friction they are willing to pay.
Cash flows do much of the work
Contributions and withdrawals move the mix without any selling. Directing new money toward whatever has fallen below target shrinks the gap before a band is ever reached.
The same applies in reverse during withdrawals, where drawing from the overweight holding restores balance while raising cash that was needed anyway.
Portfolios that receive regular contributions therefore breach bands less often than static ones, because ordinary cash flow keeps absorbing part of the drift.
Rebalancing is not a performance strategy
Correcting the mix sells part of what has risen and buys what has lagged, which sometimes helps returns and sometimes hurts them depending on what happens next.
Its reliable effect is on risk rather than on return. It keeps the portfolio's exposure near the level that was deliberately chosen instead of near whatever the market drifted toward.
Treating it as a way to improve outcomes misreads the mechanism. The function is control, and the return consequence is a side effect that varies by period.