Rebalancing
Keep a portfolio at its target weights · 7 min
- Weight in the growth asset, never rebalanced
- Target 60%
The growth asset slowly takes over, and so does its risk.
You start at 60% in a growth asset and 40% in a steady one and never rebalance. After years in which the growth asset did best, your portfolio is…
Target weights only hold for one day. After that, winners grow and losers shrink, and the portfolio drifts away from the risk you chose. Rebalancing trades it back: sell what is overweight, buy what is underweight.
There are two common schedules: calendar (monthly, quarterly, yearly) and bands (trade only when a weight is more than, say, 5 points off). Rebalancing is a mild reversion bet: it sells recent winners. That helps when assets wander and come back, and costs a little when one asset trends for years. Very frequent rebalancing mostly buys costs; never rebalancing gives up control of risk.
- Weight in growth asset, rebalanced
- Never rebalanced
- Target 60%
- Rebalanced
- Never rebalanced
Ten years, two generated assets: a growth asset (18% volatility) that falls in crises and a steady one (6%) that rises a little in them.
Generated markets, not real data. They show how a rule depends on the market it meets; they do not predict how real markets behave.
Rebalancing after a strong year
- Target
- 60% A, 40% B
- After a year
- A $78,000, B $42,000, total $120,000
- Target values
- A $72,000, B $48,000
- Trades
- sell $6,000 of A, buy $6,000 of B
- Weight of A before the trade
- 78 ÷ 120 = 65%
A 60/40 portfolio: the 60% part rose 30% this year and the 40% part was flat. What is the new weight of the first part?
What is the main cost of rebalancing every day instead of every quarter?
Target 50/50. You hold A $70,000 and B $30,000. What trades rebalance it?
Missed questions come back for review after one day, then three, then a week.
Your target is 60% in strategy A and 40% in B. After a year A is worth $50,000 and B $50,000. How much must you buy of A to rebalance (in $; negative means sell)?
Hint: Target value of A = target weight × total. The trade is target value − current value.
Three years of a fresh generated world are on screen; the next two years do not exist yet. Tune the rule on what you can see, commit to two predictions, then generate the future.
- Your rule so far
Over the next two years, after costs, this rule will…
Will the deepest drawdown after the split be worse than before it?
Generated markets, not real data. They show how a rule depends on the market it meets; they do not predict how real markets behave.