Rebalancing is selling high, buying low — by rule
It forces you to trim winners and add to laggards, which feels wrong but controls risk. Watch the fee and tax cost of each trade. Check concentration on the portfolio value calculator.
Rebalancing: selling winners on schedule
A 70/30 BTC/ETH target drifts every day the two move differently. Rebalancing sells the overweight side and buys the underweight one back to target — mechanically selling what pumped and buying what lagged. It's the only widely-used strategy that systematically forces buy-low-sell-high without prediction.
The empirical bonus is real but modest: for volatile, loosely-correlated assets, periodic rebalancing historically adds 0.5–2% annually over buy-and-hold of the same mix (the "rebalancing premium"), while cutting drawdowns somewhat. The cost side: every rebalance is a taxable event and pays fees, so frequency matters — monthly or threshold-based (rebalance when weights drift 5+ points) beats daily fiddling in nearly every backtest.
The uncomfortable part: in a strong single-asset trend, rebalancing underperforms — you kept trimming the winner all the way up. It's a bet on mean reversion between your assets. Choose it because you want the risk control, not because it beats holding in every market.