Every three months, you sell whatever has outgrown its target and buy whatever has fallen behind. That's the whole idea. It takes about twenty minutes on Kraken once your target percentages are written down, and it removes the hardest part of investing: guessing what happens next.
The alternative is doing nothing and hoping your original allocation survives contact with the market. It usually doesn't. Bitcoin's 2021 run turned plenty of "balanced" portfolios into accidental bitcoin bets, and the 2022 drawdown punished everyone who never trimmed.
Why three months beats weekly and yearly
Waiting a full year lets drift compound. A 60/40 bitcoin-ethereum split can become 80/20 after one strong quarter, which means you're carrying risk you never agreed to. Checking weekly is the opposite problem — you generate fees and taxable events chasing corrections that would have reversed on their own by Tuesday.
Quarterly sits in the middle. You catch real drift, skip the noise, and make four decisions a year instead of fifty-two.
What drift looks like with real numbers
Start with $10,000 split 50/50 between BTC and ETH. Bitcoin runs up 40% over the quarter; ethereum goes nowhere. You now hold $7,000 in BTC and $5,000 in ETH — a 58/42 portfolio, not the 50/50 you chose. Nothing broke. Your risk simply shifted without asking you first.
Rebalancing here means selling roughly $1,000 of BTC and moving it into ETH to get back to $6,000 each.
Running the rebalance on Kraken
Kraken handles this well because both assets sit on one exchange. You check balances in one place and trade without shuffling funds between platforms. The process is deliberately boring.
- Write your target percentages somewhere permanent — not in your head.
- Pull current balances and price them in your base currency.
- Calculate the dollar gap between where each asset sits and where it should sit.
- Place the smallest trades that close the gap.
- Log the date and the trades. Future you will want the record.
Set a drift band before you start
Trading every quarter over a 1% deviation is churn. Pick a threshold — 5% absolute drift is common — and act only when an asset crosses it. If your BTC target is 50% and it sits at 53%, you wait. At 56%, you trade.
Fees and taxes are the hidden cost
In most jurisdictions, every rebalance is a taxable disposal, and every trade pays a spread. On a $10,000 portfolio, moving $1,000 costs almost nothing. On a $500 portfolio, the same percentage move may not be worth the friction. Smaller accounts can often rebalance with new deposits instead — buy the underweight asset with fresh money and skip the sale entirely.
Mistakes that quietly wreck the process
The biggest one isn't math. It's emotional. People rebalance into whatever just crashed, then panic-sell two weeks later when it drops further. The mechanical version works precisely because you're not making a judgment call about direction.
Other errors show up regularly:
- Rebalancing on price targets instead of percentages. "BTC hit $80k so I'll trim" is market timing wearing a rebalancing costume.
- Forgetting stablecoins. If 20% is meant to sit in cash equivalents, that 20% counts toward the allocation.
- Ignoring correlation. BTC and ETH often move together, so a two-asset portfolio can still be one bet in disguise.
- Rebalancing in a taxable account without checking whether new contributions could do the job instead.
How the frequencies compare
| Frequency | Upside | Downside |
|---|---|---|
| Monthly | Tight tracking to targets | More fees, more taxable events |
| Quarterly | Balances drift control with low churn | Misses sharp intra-quarter swings |
| Annually | Minimal trading | Allocation can drift far from intent |
When quarterly doesn't fit
If you're dollar-cost averaging small amounts weekly, you can rebalance continuously by directing new buys toward whatever's underweight. No selling required. That's cleaner for anyone still building a position rather than managing a finished one.
Questions that come up every quarter
Do I have to rebalance exactly every three months?
No. The calendar is a prompt, not a rule. What matters is checking on a schedule and acting when drift crosses your threshold. Some quarters you'll trade nothing, and that's a fine outcome. The discipline is in looking, not trading.
What if one asset has grown so much I don't want to sell it?
That reluctance usually means the position has gotten too large for your comfort — exactly what rebalancing is designed to catch. You don't have to sell all at once. Trimming in smaller pieces over a few sessions softens the psychological sting. Just don't let "I'll do it next quarter" become permanent.
Does rebalancing hurt returns?
Sometimes. In a strongly trending market, the winning asset keeps winning and rebalancing costs you some upside. You're trading that upside for lower concentration risk. Whether the trade-off helps depends on your goals and how you'd behave in a deep drawdown. The point isn't maximizing returns — it's keeping risk where you decided it should be.
Can I rebalance with new money instead of selling?
Yes, and it's often better for smaller portfolios. If you add funds regularly, route each deposit into whatever sits below target. You drift back toward your allocation without triggering a sale or taxable event. Large gaps take longer to correct, but the friction is close to zero.
Pick a date — the first weekend of January, April, July, and October works well — and put it in your calendar now. Then set a reminder for the week before, so you have time to check your drift band before the date arrives.