Re-balancing in Practice

When NOT to Rebalance

Pomegra Learn

When NOT to Rebalance

Rebalancing is mechanically valuable—it buys low and sells high. But it has costs: transaction costs, capital gains taxes, and the time to execute. Sometimes, those costs exceed the benefit of rebalancing. This article explores when to skip rebalancing.

Key takeaways

Very small accounts (under $10,000) often have transaction costs exceeding rebalancing benefits.
New portfolios in year 1 naturally gravitate toward allocation; don't force rebalancing.
Taxable accounts with large embedded gains suffer heavy capital gains tax from rebalancing; consider tax-loss harvesting first.
Rebalancing frequency above 12 months per year generates cost without benefit.
In very volatile markets, wait for stabilization to avoid whipsaw trading.

Exception 1: Very small accounts (under $10,000)

Rebalancing has fixed costs: brokerage spreads, fractional share rounding, and your time. In a small account, these costs can exceed the benefit.

Example: $5,000 portfolio, 60/40 stocks/bonds

VTI (stocks): $3,000
BND (bonds): $2,000

Markets rally 20%. Drift: VTI $3,600, BND $2,000. Now 64% stocks, 36% bonds.

Rebalancing means selling $200 of VTI and buying $200 of BND.

Costs:

Bid-ask spread on VTI: ~0.01%, so $2 on the sell.
Bid-ask spread on BND: ~0.01%, so $2 on the buy.
Your time: 15 minutes = $10 (at $40/hour).
Total cost: ~$14.

Benefit:

The benefit of rebalancing is the outperformance from buying low and selling high. Historical data shows rebalancing adds 0.2–0.5% annually. On a $5,000 portfolio, that's $10–$25/year.

Cost-benefit: $14 cost versus $15 benefit. Break-even or slightly negative.

Recommendation:

In small accounts, rebalance annually (once a year), not quarterly. This reduces the frequency of costs. Or wait until the account grows to $10,000+ before strict rebalancing.

Exception 2: First year of investing (natural drift is OK)

When you first build a portfolio, you're often adding money over time. Each contribution shifts allocation. Rebalancing while you're still building is redundant.

Example: Building a $100,000 portfolio over a year

Month 1: Invest $10,000 (60/40 split): $6,000 stocks, $4,000 bonds. Month 2: Add $10,000 (direct to stocks for rebalance): Now $16,000 stocks, $4,000 bonds. Month 3: Add $10,000 (60/40 split): $22,000 stocks, $8,000 bonds. ... Month 12: Final $10,000 added. Total: $100,000, approximately 60/40.

In this scenario, you're rebalancing through contributions. You don't need to sell and buy; the new contributions do the work.

Recommendation:

In year 1 of a portfolio, don't rebalance unless drift is extreme (say, 70% stocks when your target is 60%). Use new contributions to target the underweighted asset instead.

Exception 3: Taxable accounts with large embedded gains

If you've held a winning position for years, it has a large unrealized gain. Selling it triggers a capital gains tax. The tax cost can swamp the rebalancing benefit.

Example: Taxable account with embedded gains

VTI: $60,000 (cost basis $30,000, gain $30,000)
BND: $40,000 (cost basis $38,000, gain $2,000)
Total: $100,000 (60/40 stocks/bonds)

Target: 50/50. Rebalancing requires selling $10,000 of VTI and buying $10,000 of BND.

Tax cost:

VTI gain ratio: $30,000 gain on $60,000 holding = 50% gain.

Selling $10,000 of VTI triggers $5,000 of gain.

Long-term capital gains tax at 15%: $750.

Benefit:

Rebalancing benefit: 0.2–0.5% annually on $100,000 = $200–$500/year.

Cost-benefit: $750 tax cost versus $350 annual benefit. Payback period: ~2 years.

Recommendation:

Don't rebalance if the tax cost is >50% of the annual rebalancing benefit. Instead:

Use new contributions to buy BND and reduce the VTI % over time.
Use tax-loss harvesting: If you have losses elsewhere, harvest them to offset the VTI gain and reduce tax drag.
Hold the position until the gain cools (more shares bought, diluting the gain ratio).

Exception 4: Too-frequent rebalancing (more than monthly)

Some investors rebalance weekly or even daily. This generates excessive trading costs and opportunity for whipsaw.

Example: Weekly rebalancing in a volatile market

Week 1: VTI rallies 5%, you trim. Cost: $50 spread.
Week 2: VTI falls 4%, you buy back. Cost: $50 spread.
Week 3: VTI rallies 6%, you trim again. Cost: $50 spread.

In five weeks, you've paid $250 in spreads to make trades that partially offset each other. The market is in the band's range naturally; frequent rebalancing is whipsaw trading, not disciplined rebalancing.

Recommendation:

Rebalance quarterly at minimum, annually at most. Monthly rebalancing is reasonable for large, automated accounts but excessive for small DIY portfolios.

Exception 5: Volatile markets (wait for stabilization)

In extremely volatile markets, allocations swing wildly. Rebalancing in the middle of a panic might trigger early, and then the market recovers and you've sold the dip.

Example: March 2020 COVID crash

On March 19, 2020, the stock market fell 12% in one day. Portfolios lurched. A 60/40 portfolio might have briefly been 50/50 or worse.

A mechanical rebalancer would buy stocks at the panic low (good). But if you manually checked your portfolio on March 19 and saw red everywhere, you might have waited two weeks for "stability" and missed the best buying opportunity.

Recommendation:

If you have automated rebalancing, let it execute during panics—that's exactly when you want to buy low.

If you're manually rebalancing:

Check quarterly, not daily.
During volatility, don't try to time the exact bottom.
Rebalance on schedule (quarterly, annual), not in panic response.

Exception 6: Extremely concentrated positions with tax loss potential

If you have a large position with an unrealized loss (and it's in a taxable account), holding it gives you the opportunity to harvest that loss in the future. Selling now to rebalance eliminates the tax-loss opportunity.

Example:

Concentrated position in ABC Corp: $40,000 (cost basis $50,000, loss $10,000)
Bond fund: $60,000
Total: $100,000

If you sell ABC to rebalance, you lose the $10,000 loss that you could harvest later.

Recommendation:

In this case, hold ABC for now. Use new contributions to buy bonds. When ABC recovers, harvest the loss if it becomes profitable. Rebalance using contributions and loss harvesting rather than selling the loss position immediately.

Exception 7: Imminent withdrawal (prepare for the spend)

If you're retiring or about to withdraw money, don't rebalance into more stocks. Instead, start building cash/bond buckets in advance.

Example: Retiring in 6 months

Current allocation: 70/30 stocks/bonds. Target allocation at retirement: 50/50 stocks/bonds.

Instead of rebalancing now (selling 20% of stocks), gradually shift using new contributions and dividend reinvestment over the next six months. By retirement, you're already at 50/50 and you have a cash bucket built up for year 1 spending.

Recommendation:

As you approach a major withdrawal (retirement, home purchase, etc.), shift toward that portfolio gradually over 1–2 years, not suddenly. Use contributions to do the work.

Exception 8: Micro-drifts within a tight band

Your band is 55–65%. Stocks are at 56%. Technically within the band, but barely. The cost of rebalancing (spreads, commissions, time) exceeds the benefit of a 1% move.

Recommendation:

Set a tighter internal trigger. Only rebalance if drift exceeds 3–5% from target, even if the band is wider. This avoids the friction of constant micro-rebalances.

The cost-benefit decision tree

For any potential rebalance, ask:

Is the cost of rebalancing ≤25% of the expected annual benefit?
- Benefit = expected rebalancing outperformance (typically 0.2–0.5% annually).
- Cost = taxes + transaction costs + spreads.
- If cost > 0.25 × benefit, don't rebalance.
Is the portfolio in year 1 of contributions?
- If yes, use contributions to rebalance. Don't sell and buy.
Is this a taxable account with large embedded gains and no tax-loss offsets?
- If yes, delay rebalancing. Use contributions. Wait for tax-loss harvesting opportunities.
Will this rebalance happen more than once per quarter?
- If yes, reduce frequency. Quarterly is the optimal minimum.
Are markets extremely volatile?
- If yes and you're manual, wait for next scheduled rebalance. If you're automated, execute (machines don't panic).

Exceptions expressed as a flowchart