Rebalancing: schedules, drift bands and trading costs

Die Idee in einfachen Worten

Weights move with prices: a ticker that rises becomes a bigger part of the portfolio and its risk grows. Rebalancing sells some of what rose and buys what fell, back to the targets. It keeps the risk you chose and, in markets that swing back and forth, adds a small "rebalancing bonus".

So wird gerechnet

Each ticker's current weight is its value ÷ the portfolio's value; drift is current weight − target. A rebalance trades each ticker by (target − current weight) × portfolio value. Our portfolios do it on a schedule (weekly, monthly, quarterly), when any weight drifts more than your band (e.g. 5 points), or when you press Rebalance now; differences under the smallest trade you set are left alone.

wᵢ(now) = valueᵢ ÷ Σ value
driftᵢ = wᵢ(now) − wᵢ(target)
tradeᵢ = −driftᵢ × portfolio value

Anwendung

• Monthly or quarterly plus a 5-point drift band suits most portfolios: few trades, risk kept in check.
• Set the smallest trade to about 1% so that tiny differences do not generate costs.
• Rebalance to a new model (Compare models… on the portfolio) when you change your mind about the mix, not after every study.

Schlüsse ziehen

• Frequent small drifts mean volatile tickers; a wider band saves trades.
• Orders for a closed market wait and are carried out when it opens.

Worauf man achten sollte

• Every trade pays the spread; rebalancing too often turns a small benefit into a cost.
• In a long one-way trend, rebalancing sells the winner early; that is the price of keeping the risk constant.

Ein wenig Geschichte

André Perold and William Sharpe compared constant-mix rebalancing with buy-and-hold and portfolio insurance in "Dynamic Strategies for Asset Allocation" (Financial Analysts Journal, 1988): rebalancing wins in markets that swing back and forth and loses in strong one-way trends.