Every crypto portfolio drifts. What starts as 40% BTC, 30% ETH, 20% alts, 10% stables becomes something different within months as some assets moon and others crash. Rebalancing is the discipline of restoring your target allocations by selling what's grown and buying what's lagged. It sounds boring, but it's one of the most reliably profitable habits in investing — buying low and selling high, systematically.
Why rebalance at all
Two reasons rebalancing is worth the effort:
**Risk control**: as a position grows, so does its impact on your portfolio. If ETH doubles and everything else stays flat, ETH is now a much larger percentage of your exposure than you intended. Rebalancing brings it back to your intended risk level.
**Systematic buying low / selling high**: rebalancing forces you to sell what's appreciated and buy what's lagged. This is contrarian to the natural instinct of "let winners run" but historically produces better returns because it captures mean reversion.
The empirical evidence: rebalanced portfolios have historically produced 0.5-1.5% additional annualized returns vs never-rebalanced portfolios, with lower drawdowns. Over long horizons, this compounds meaningfully.
Time-based rebalancing
The simplest approach: rebalance on a schedule regardless of what's happened.
**Quarterly**: check allocations every 3 months, restore targets. Most common cadence.
**Semi-annually**: every 6 months. Less trading, more drift.
**Annually**: once a year. Minimal trading, more drift, works for very long-term investors.
Time-based is easy to execute and mechanical (no judgment calls). Downside: you rebalance regardless of whether it's needed, which can create unnecessary tax events.
Threshold-based rebalancing
Alternative: rebalance whenever any position deviates X% from target.
Example rule: "rebalance if any position is more than 20% above or below its target allocation."
If your BTC target is 40% and BTC drifts to 50% (25% above target: (50-40)/40 = 25%), rebalance is triggered.
Threshold-based rebalancing:
- Requires less frequent action when markets are calm
- Triggers more actively during volatile periods
- Captures larger deviations for larger tax events / trading costs
- Requires more monitoring than time-based
For most active crypto investors, threshold-based works well because crypto volatility makes triggers happen naturally.
The hybrid: threshold with time backstop
The best of both: use threshold triggers but also check quarterly regardless. This ensures you catch small drifts that never reach a threshold.
Concrete rule: "check portfolio quarterly; rebalance any position that's more than 15% off target from any check; also perform a full annual review regardless."
What "rebalancing" actually means
For a crypto portfolio with a 40/30/20/10 target:
Current state: 55% BTC, 25% ETH, 15% alts, 5% stables (BTC ran, others didn't).
To rebalance to targets on a $100K portfolio:
- BTC: sell $15K worth (55% → 40%)
- ETH: buy $5K worth (25% → 30%)
- Alts: buy $5K worth (15% → 20%)
- Stables: buy $5K worth (5% → 10%)
This means selling BTC (up ~35% since last rebalance perhaps) and buying the laggards. Psychologically hard — you're "selling your winners." But that's exactly the point.
Tax-aware rebalancing
Every sale creates a taxable event (in most jurisdictions). If you've held BTC for 2 years and it's tripled, selling triggers significant capital gains tax.
Ways to reduce tax impact:
**Use new capital first**: if you're contributing to the portfolio monthly, use new contributions to buy underweight positions rather than selling overweight ones.
**Long-term vs short-term gains**: in the US, positions held over 1 year get lower long-term capital gains rates. Prefer selling long-term positions when possible.
**Loss harvesting**: if any positions are down, sell them alongside your rebalancing to realize losses that offset gains.
**Rebalance in tax-advantaged accounts first**: for those with crypto in IRAs or similar, rebalancing there is tax-free.
**Higher rebalance thresholds**: setting a 30% threshold instead of 15% reduces the frequency of tax events at the cost of more drift.
The crypto-specific complication
Crypto rebalancing has some unique considerations:
**Bear market bottom-timing**: rebalancing during a bear market often means selling remaining stablecoins to buy more BTC/ETH as they crash. This is uncomfortable but often correct — you're buying at bear market discounts.
**Bull market top-timing**: rebalancing in a bull run means selling BTC/ETH as they moon. Feels like leaving money on the table but is the discipline that captures gains.
**Cycle-aware allocations**: some investors adjust targets based on cycle position (higher stables in bull tops, lower stables in bear bottoms). This works when done systematically; fails when done emotionally.
**Alt rotation**: alts frequently 5-10x within a cycle. If your alt position was 10% at cycle bottom and is now 40% at cycle top, aggressive trimming is defensible even beyond your normal threshold.
When not to rebalance
Cases where holding drift makes sense:
**Very long-term convictions**: if you specifically believe BTC will outperform everything for the next decade and you're building generational wealth, letting BTC run to 70% might be defensible (though risky).
**Illiquid positions**: some altcoins or NFTs may have poor liquidity. Forcing a sale in low-liquidity conditions can cost more in slippage than rebalancing gains.
**Near a decision point**: if you're close to needing the money (major purchase, retirement, etc.), the rebalancing framework changes — you're transitioning to distribution mode.
**Immediate post-rebalance drift**: rebalancing back to targets and then having one position spike 5% doesn't require immediate re-rebalancing. Small deviations are noise.
The behavioral discipline
Rebalancing works because it forces action against your instincts. When BTC has doubled and altcoins are lagging, everyone wants to sell alts and pile into more BTC. The rebalancing rule says the opposite: trim BTC, add to alts.
This is uncomfortable. Every fiber of your being will resist selling winners in an uptrend. But historically, mean reversion is a real thing — winners tend to underperform after outperformance, and vice versa.
The discipline: pre-commit to the rules, execute mechanically, don't second-guess.
For different portfolio sizes
**Small portfolios (<$25K)**: rebalance more crudely (annually). Trading costs and complexity matter more; don't overthink small deviations.
**Medium portfolios ($25K-500K)**: quarterly threshold-based rebalancing. Enough scale to justify the effort; not so large that tax optimization dominates.
**Large portfolios ($500K+)**: threshold-based with tax awareness. Consider professional help for tax planning around rebalancing.
**Institutional**: continuous rebalancing via automated systems. Different game entirely.
A concrete rebalancing template
Quarterly review process:
1. Pull current portfolio value and per-asset percentages. 2. Compare each asset to target allocation. 3. Identify assets more than 15% off target (up or down). 4. For each deviant position, calculate rebalance size in USD. 5. Consider tax impact: prefer long-term positions for sales, use losses to offset gains where possible. 6. Consider using new cash contributions first for underweight positions. 7. Execute the rebalance in one session. 8. Record in a portfolio tracker for future reference.
For most retail investors, this takes 30-60 minutes per quarter. Small time investment, meaningful long-term impact.
The takeaway
Rebalancing is not exciting. It doesn't make headlines. It doesn't win Twitter arguments. It just quietly captures the mean-reverting nature of markets over long horizons.
Investors who rebalance systematically compound better than investors who let portfolios drift. The evidence is strong and long-standing. Whether you rebalance quarterly, threshold-based, or hybrid — pick a rule and follow it.
The undisciplined alternative is watching your allocation drift into concentrations you never intended, and then bearing the risk of major single-position losses. Boring but effective usually beats exciting but random over time.
