Shannon's Demon and the Rebalancing Bonus

Imagine two assets, A and B, that have identical long-term returns but different volatility. Asset A oscillates 20%, 20%, 20% over three years. Asset B oscillates 50%, -50%, 50% over the same span. Both end at the same price, yet a portfolio that regularly rebalances between them captures excess returns from the volatility of B.

This counterintuitive phenomenon is called the "rebalancing bonus" or "Shannon's Demon," named after information theorist Claude Shannon. It's a mathematical proof that rebalancing—purely mechanical buying low and selling high—can generate returns without any predictive insight, just volatility and discipline.

This article explores the mathematics, the real-world constraints, and the practical implications of the rebalancing bonus.

Quick definition: The rebalancing bonus is the excess return generated by mechanically rebalancing between volatile assets with equal long-term returns, due to the mathematical asymmetry between arithmetic and geometric growth.

Key Takeaways

• The rebalancing bonus is real and measurable: a 50/50 portfolio of two identical-return, different-volatility assets outperforms through rebalancing alone
• The bonus emerges from volatility drag: when an asset falls, a dollar buys more shares; when it rises, you own fewer shares (via rebalancing), locking in the gain
• The larger the volatility difference between assets, the larger the potential bonus; a 60/40 bonds/stocks mix captures less bonus than a 50/50 mix of highly uncorrelated assets
• The bonus is reduced by transaction costs, taxes, and bid-ask spreads; the net benefit depends on portfolio size and rebalancing frequency
• The bonus is strongest in high-volatility, oscillating markets and weakest in trending markets (where you're selling winners continuously)
• Real-world rebalancing bonuses are 0.5–2% per year on typical portfolios; meaningful but not transformative

The Mathematics: Why Two Identical Assets Diverge

Start with a simple example: Asset A and Asset B, each costing $100 initially.

Year 1:

• Asset A: up 20% → $120
• Asset B: down 20% → $80
• Portfolio balance (equal-weighted): A = $120, B = $80, total = $200

If you rebalance to 50/50, you sell $20 of A and buy $20 of B:

• A: $100
• B: $100

Year 2:

• Asset A: up 20% → $120
• Asset B: up 150% (recovery from crash) → $250
• Portfolio balance: A = $120, B = $250, total = $370

Rebalance to 50/50:

• A: $185
• B: $185

Year 3:

• Asset A: up 20% → $222
• Asset B: down 26% (reverting) → $137
• Portfolio balance: A = $222, B = $137, total = $359

What's happened? Both assets oscillated through the same returns (+20%, -20%, then +150%, -26%, etc.), yet the rebalanced portfolio ($359) is higher than a static 50/50 buy-and-hold portfolio ($320).

The rebalancing bonus in this example is roughly $40, or 12% extra return.

How? Because every time Asset B fell (and A rose), you rebalanced by selling the winner (A) and buying the loser (B). When B recovered, you owned more of it (due to the previous rebalancing). This "buy low, sell high" mechanic is pure leverage from volatility and discipline.

The Volatility Drag Concept

The mathematical intuition is volatility drag: when a portfolio component is volatile, the geometric return (what you actually earn) lags the arithmetic return (the average).

Example: An asset that goes from $100 → $50 (–50%) → $75 (+50%) has an arithmetic average return of 0% but a geometric return of –13% (you ended at $75).

Rebalancing between two volatile assets exploits this drag. You're forced to sell the asset that's recovered from a crash (high after the recovery) and buy the asset that crashed (low). This continuous mechanical switching between assets generates returns purely from the volatility pattern, not from any skill.

Measuring the Rebalancing Bonus in Real Portfolios

Academic research has quantified the rebalancing bonus on realistic portfolios:

• A 50/50 stocks/bonds portfolio rebalanced annually captures roughly 0.5–1% annual bonus from rebalancing
• A 70/30 stocks/bonds portfolio captures 0.2–0.5% bonus (less volatility mismatch)
• A 60/40 stocks/bonds portfolio captures 0.3–0.7% bonus
• A 25/25/25/25 portfolio (four uncorrelated assets) captures 0.8–1.5% bonus

The bonus varies with volatility, correlation, and rebalancing frequency. Higher volatility = higher bonus. Lower correlation = higher bonus. More frequent rebalancing = higher bonus (but also higher costs).

Research by Vanguard and others shows that on a typical $1 million portfolio, the annual rebalancing bonus, after transaction costs and taxes, is roughly $3,000–$7,000 per year for a 60/40 allocation. This compounds to meaningful wealth over decades.

The Costs That Reduce the Bonus

The theoretical rebalancing bonus assumes zero transaction costs, zero taxes, and instant execution. Reality is costlier:

• Transaction costs: Bid-ask spreads, commissions (now rare), and market impact reduce realized bonus by 0.1–0.5% annually
• Taxes: Rebalancing in taxable accounts triggers capital gains, which can be 15–37% of the realized gain, reducing the after-tax bonus
• Timing friction: Rebalancing can take days to execute, during which markets move

On a $1 million portfolio in a taxable account rebalancing annually, net bonus after taxes might be 0.2–0.4% annually instead of 0.5–1% theoretical.

On a $100,000 portfolio, transaction costs and slippage might eliminate 50% of the theoretical bonus, reducing it to 0.25–0.5%.

This is why rebalancing makes sense for large portfolios ($500K+) but is less attractive for small portfolios where costs dominate returns.

When the Bonus Disappears: Trending Markets

The rebalancing bonus is strongest in oscillating, mean-reverting markets and weakest in trending markets.

In a market where stocks rise 20% per year and bonds rise 2% per year consistently, rebalancing is a drag. You're continuously selling rising stocks (winners) and buying flat bonds (losers). Over 10 years of consistent outperformance, the rebalancing bonus turns negative—you'd have been better off just holding the 60/40 and letting stocks drift to 80% or higher.

This is a key insight: the rebalancing bonus exists only when assets oscillate around their trend. In a steadily uptrending or downtrending market, rebalancing is suboptimal. In an oscillating market (up, down, up, down), rebalancing is superior.

From 2009–2021, stocks consistently outperformed bonds. The rebalancing bonus was near zero (or negative after costs). From 2022–2024, with rising rates and stock volatility, the rebalancing bonus likely re-emerged.

Real-World Bonus Measurement: 60/40 Historical

Using 100 years of data (1924–2024), researchers have measured the annual rebalancing bonus on a 60/40 stocks/bonds portfolio with annual rebalancing:

• 1926–1950: +0.8% per year (high volatility, oscillations)
• 1950–1980: +0.5% per year (steady bull market, then steady bear market; moderate bonus)
• 1980–2000: +0.3% per year (bull market, lower volatility, lower bonus)
• 2000–2010: +0.9% per year (two major crashes, high volatility, high bonus)
• 2010–2020: +0.1% per year (steady bull market, low bonus, negative in early years)
• 2020–2024: +0.6% per year (volatility recovery, bonus recovery)

The pattern is clear: bonus emerges in volatile, oscillating periods. Bonus disappears in calm trending periods.

This explains why rebalancing was mocked from 2015–2019 (stocks up steadily, bonds down steadily) but praised again in 2020–2022 (extreme volatility, reversals).

Shannon's Demon Explained: The Thought Experiment

Claude Shannon proposed a thought experiment: a coin-flip that alternates assets A and B every period. A goes up 50% if the flip is heads, down 50% if tails. B does the opposite.

With equal starting amounts and a rebalancing rule ("keep 50/50 ratio"), your wealth compounds at a steady 12.5% per year, despite the underlying assets having zero expected return (50% up, 50% down = 0%).

The "demon" (the rebalancing rule) extracts wealth purely from volatility and discipline. This is the purest form of the rebalancing bonus.

Real markets are not this extreme (volatility is lower, outcomes are less binary), but the principle holds: rebalancing between volatile assets with equal long-term returns generates measurable gains from the volatility pattern.

The Volatility Smile: When Bonus Peaks

The rebalancing bonus is highest when:

1. Volatility is high: Wide swings between assets create more "buy low" and "sell high" opportunities
2. Volatility is asymmetric: Stocks crash 30% while bonds fall 5%; the difference creates the arbitrage
3. Mean reversion is strong: Asset crashes are followed by recoveries; this requires a recovery to lock in the rebalancing gain
4. Rebalancing is frequent: Quarterly rebalancing captures more oscillations than annual
5. Assets are uncorrelated: Negative correlation between stocks and bonds maximizes the benefit (when stocks fall, bonds rise, creating rebalancing opportunity)

The bonus is minimized when:

1. Volatility is low: Steady, smooth markets reduce opportunities
2. Trends are persistent: Stocks consistently outperform; rebalancing is a drag
3. Mean reversion is weak: Crashes are not followed by recoveries; rebalancing leaves you overweighting the crashed asset
4. Assets are correlated: All assets rise and fall together; rebalancing doesn't help

Practical Implications for Portfolio Design

For investors seeking to maximize the rebalancing bonus:

• Increase diversification: More uncorrelated assets = more bonus potential. A 60/40 portfolio has lower bonus potential than a four-asset barbell (25% stocks, 25% bonds, 25% alternatives, 25% cash).
• Embrace volatility (within tolerance): Higher volatility assets = higher bonus. But volatility must be within your risk tolerance; a bonus means nothing if you panic-sell in a crash.
• Rebalance frequently: Quarterly rebalancing captures more oscillations than annual. But transaction costs must be factored.
• Use tax-advantaged accounts: Rebalancing generates no taxable gain in an IRA; maximize the bonus by rebalancing frequently in these accounts.
• Keep portfolio size reasonable: Bonuses as a percentage are meaningful for large portfolios ($1M+); for small portfolios ($50K), bonuses are easily overwhelmed by costs.

Real-World Examples

The Small Bonus in a Bull Market: An investor rebalanced a 60/40 from 2015–2019 (steady stock bull run). His rebalancing bonus was near zero because stocks consistently outperformed. He'd have been better off just holding 60/40 and letting stocks drift to 70% or higher. But because he didn't know a bull market was coming, the rebalancing rule was still appropriate ex-ante.

The Large Bonus in a Volatile Period: An investor rebalanced a 60/40 from 2007–2010 (crash, recovery, recovery, crash). His rebalancing bonus was roughly 2% per year (gross) or 1.2% per year (net of taxes and costs). This added roughly $60,000 to his portfolio over the period, purely from the mechanical discipline of rebalancing into crashes and rallies.

The Uncorrelated Assets Bonus: An investor with a 25/25/25/25 portfolio (stocks, bonds, alternatives, commodities) rebalancing quarterly from 2000–2024 likely captured bonuses of 0.8–1.5% annually, higher than a simple 60/40, because the four assets are less correlated and each oscillates differently.

Common Mistakes

1. Overestimating the bonus: Theoretical bonuses are often 1–2% annually, but after taxes, costs, and timing slippage, net bonuses are 0.3–0.8%. Don't expect the bonus alone to drive returns.

2. Rebalancing too frequently: Daily or weekly rebalancing might capture the bonus faster, but transaction costs and slippage overwhelm gains. Quarterly or annual rebalancing is typically optimal.

3. Ignoring volatility requirements: A portfolio designed to maximize the rebalancing bonus might have higher volatility than you can tolerate. Comfort with your allocation matters more than a 0.5% annual bonus.

4. Assuming bonus exists in trends: If stocks are rising steadily, the bonus disappears, and rebalancing is a drag. Don't rebalance more aggressively based on past bonuses that emerged in oscillating markets.

5. Forgetting the bonus is mathematical, not predictive: The bonus doesn't require you to predict which asset will outperform. It's pure mechanics. Treating it as a market-timing edge is an error.

FAQ

Q: Is the rebalancing bonus the same as the "rebalancing benefit"? A: No. The rebalancing benefit is the risk-control value (reduced volatility, better crashes). The rebalancing bonus is purely mathematical gain from oscillating assets. Both matter; they're separate.

Q: Should I increase rebalancing frequency to capture more bonus? A: Up to a point. Quarterly rebalancing typically captures most of the available bonus. Monthly rebalancing adds little and increases costs. Daily rebalancing is excessive.

Q: Does the bonus apply to 401(k) rebalancing as much as taxable? A: Yes, more so. Tax-advantaged accounts avoid the tax drag, so the gross bonus is realized as net bonus. Taxable accounts lose 15–37% of the bonus to taxes. Maximize rebalancing frequency in tax-advantaged accounts.

Q: Can I use the rebalancing bonus as a return forecast? A: No. The bonus is backward-looking (based on observed volatility) and future-dependent (it requires mean reversion to persist). Don't assume the past 5 years' bonus will repeat.

Q: What if the rebalancing bonus is negative (rebalancing hurts)? A: In a strongly trending market (stocks up every year, bonds flat), rebalancing is a drag. This happened 2015–2019. Accept it as the cost of discipline. The bonus will likely re-emerge when volatility increases.

Related Concepts

• Volatility Drag: The mathematical asymmetry between arithmetic and geometric returns; the principle underlying the rebalancing bonus.
• Mean Reversion: The tendency of assets to revert to historical averages; rebalancing profit depends on mean reversion occurring after rebalancing buys the dip.
• Correlation: The degree to which assets move together; lower correlation enables higher rebalancing bonuses.
• Risk Parity: A portfolio construction technique that aims to maximize diversification (and rebalancing bonus) by allocating equally to volatility, not equal dollars.
• Compound Annual Growth Rate (CAGR): The actual growth rate achieved; rebalancing bonus flows here, not into higher volatility.

Summary

The rebalancing bonus is a mathematically real phenomenon where pure mechanical discipline—buying low and selling high through rebalancing—generates excess returns from portfolio oscillations. The bonus is strongest in volatile, mean-reverting markets and weakest in calm trending markets.

On a typical 60/40 portfolio, the bonus is 0.3–0.8% annually after costs and taxes. For a $1 million portfolio, this compounds to meaningful wealth over 30 years. But the bonus is not guaranteed; it depends on market regime and volatility patterns you cannot predict.

The key insight: rebalancing is not just about risk control. It's also a source of return, derived from mechanical discipline and volatility capture. Understanding this is motivating when discipline is hardest—when you're rebalancing into a crash or out of a bull run.

The bonus is the mathematical argument for rebalancing. The risk-control benefits are the practical argument. Together, they make a compelling case.

Next

The rebalancing bonus is a mathematical phenomenon that applies to all rebalancers—individuals and institutions alike. But some investors delegate rebalancing to intermediaries: target-date funds. The next article explores how these funds mechanically rebalance for you and whether the approach is worthwhile.