A portfolio drifts away from the weights you chose. The standard fix is to sell whatever has grown too large. There is a second fix that uses only buys, and the choice between them decides whether correcting drift costs you a realised gain.
Drift is the gap between the weights you set and the weights you currently hold. You decide on 30% BTC and 10% LINK. BTC doubles and LINK does not, and without a single trade you now hold 42% BTC and 7% LINK. Nothing went wrong — that is arithmetic. But the portfolio you are holding is no longer the one you designed, and it is more concentrated in whatever has already run.
Threshold rebalancing sells what is above target and buys what is below it, whenever the gap passes a limit you set. If BTC has grown past its share, the tool sells just enough BTC to bring it back to 30% and puts the proceeds into whatever is underweight.
It works, it is fast, and it needs no new money — which is its real advantage. It also has two costs. The first is that it sells the position that was working, on the way up, by design. The second is that a sale is a disposal: in many jurisdictions that is a taxable event, and vendors who sell threshold rebalancing say so in their own documentation. What that costs depends on where you live and how long you held, which is a question for someone qualified to answer it.
Yes. Point new money at the holdings furthest below their target instead of splitting it evenly. This is sometimes called cash-flow rebalancing, and it corrects the same drift from the other direction: rather than cutting the holding that grew, you top up the ones that did not.
Nothing is sold, so no gain is realised and no position is cut short. The correction happens at whatever pace you are adding money.
Buy-only rebalancing is slower, and it has a limit. If you stop adding money, drift stops being corrected — there is nothing to correct it with. And if one holding runs far enough, no realistic contribution brings it back to target; a coin that is 20 points overweight stays overweight until you either sell some of it or add a great deal to everything else.
Selling to rebalance is the only method that works with no new money, and the only one that can bring a large overweight back quickly. Whether that is worth realising the gain is the trade-off, and it is yours to make.
Buying is drift-aware by default, and nothing is ever sold to correct drift on its own:
The one thing the engine can sell without you pressing anything is the optional scheduled trim, and it ships switched off. When it is on, it sells a holding's profit only — never the money that bought it — and never touches a position at a loss, however overbought it gets.
Yes. Instead of selling what has grown past its target weight, direct new contributions to the holdings furthest below target. This is called cash-flow rebalancing. It corrects drift using buys only, so no gain is realised, but it works only while you are still adding money and cannot quickly correct a very large overweight.
Selling one asset to buy another is a disposal, and in many jurisdictions a disposal is a taxable event even when the proceeds never leave the exchange. Buy-only rebalancing avoids the disposal because nothing is sold. Tax treatment depends on your country and your holding period — ask someone qualified about your own situation.
Threshold rebalancing triggers a correction whenever a holding drifts more than a set number of percentage points from its target weight. It sells the overweight positions and buys the underweight ones to restore the target allocation. It is faster than cash-flow rebalancing and requires no new money, but it realises gains on the positions that performed best.
No. Rebalancing by buying is the default and nothing is sold to correct drift on its own. Hard and positive rebalances exist as manual actions behind a preview and a confirmation. The only automatic sell in the product is the optional scheduled trim, which ships off, sells only a holding's profit, and never sells a position at a loss.
There is no single correct interval and Trade Leopard makes no recommendation. Common approaches are calendar-based (quarterly or annually) and threshold-based (when a weight drifts past a set number of points). With buy-only rebalancing the question mostly answers itself: the correction happens every time you add money.
Set your weights, your budget and the gate. It buys on your schedule at Kraken, OKX or Coinbase, using an API key that cannot withdraw. $15 a month, cancel anytime.