The Rebalancing Bonus: How to Create Returns Out of Thin Air (Shannon’s Demon)
The Rebalancing Bonus: How to Create Returns Out of Thin Air (Shannon’s Demon)
COACHING POINTS
- The Concept: Claude Shannon (the father of information theory) proved that you can generate positive returns from two assets that have zero long-term returns, simply by rebalancing them against each other. This phenomenon is called Shannon’s Demon.
- The Mechanism: It works by “Buying Low and Selling High” automatically. When Asset A crashes and Asset B spikes, rebalancing forces you to sell expensive B to buy cheap A. You harvest the volatility itself as profit.
- The Requirement: To capture this “Rebalancing Bonus,” you need assets with low correlation (e.g., Stocks vs. Bonds, or Stocks vs. Gold) and high volatility. A portfolio of 10 highly correlated tech stocks generates almost zero bonus.
Most investors view volatility as risk. Mathematical investors view volatility as a resource. The “Rebalancing Bonus” is the only “free lunch” in finance besides diversification. It mathematically proves that the geometric return of a portfolio can exceed the weighted average return of its componentsโif, and only if, you rebalance religiously. Source: Claude Shannon’s Papers / Rebalancing Alpha Studies
Scenario: You own 50% Asset A and 50% Asset B. Both assets are wildly volatile but return 0% over 2 years.
- Year 1:
Asset A: Doubles (+100%).
Asset B: Halves (-50%).
Action: You Rebalance back to 50/50. (Sell high A, Buy low B). - Year 2:
Asset A: Halves (-50%). (Back to start).
Asset B: Doubles (+100%). (Back to start). - The Result:
Buy & Hold Investor: $100 -> $100. (0% Gain).
Rebalancing Investor: $100 -> $125. (+25% Gain).
You created a 25% profit from two assets that went nowhere.
Rebalancing Bonus Potential (Basis Points)
| Asset Pair (Correlation) | Annual Bonus Potential (bps) |
|---|---|
| US Stocks / US Stocks (High Corr) | 5 |
| Stocks / Bonds (Low Corr) | 50 |
| Stocks / Crypto (Uncorrelated + Volatile) | 150 |
*The bonus explodes when you mix assets that move differently (Low Correlation) and move violently (High Volatility).
What-If Scenario: Timing Matters
Comparison: Calendar Rebalancing vs. Threshold (Bands) Rebalancing.
| Strategy | Efficiency Score (0-100) |
|---|---|
| Calendar (Annual) | 70 |
| Calendar (Monthly) | 60 |
| Threshold (5% Bands) | 95 |
Execution Protocol
You cannot harvest volatility from two S&P 500 ETFs. You need things that zig when others zag. Historically, Long-Term Treasuries (TLT), Gold (GLD), and Managed Futures have provided the best counter-movements to Equities.
Adopt the “5/25 Rule.” Rebalance an asset class if its weight drifts by an absolute 5% (e.g., 60% -> 65%) or a relative 25% (e.g., 10% -> 12.5%). Check your portfolio quarterly or simply set alerts in your brokerage app.
The most tax-efficient way to rebalance is by buying the underweight asset with new savings (contributions). This resets the allocation without selling winners and triggering capital gains tax.
COACHING DIRECTIVE
- Do This: Embrace volatility in small doses. Adding a 5% slice of a highly volatile asset (like Bitcoin or Emerging Markets) to a boring portfolio can boost returns if you rebalance aggressively.
- Avoid This: Rebalancing too often (e.g., daily). You will eat up your bonus in transaction costs and short-term capital gains taxes. Momentum needs time to run before you trim it.
Frequently Asked Questions
Does this work in a bear market?
Yes. In a bear market, rebalancing forces you to buy stocks as they fall (using stable bond capital). This lowers your cost basis faster than a buy-and-hold investor, accelerating your recovery when the market turns.
Is there a tax cost?
Yes. Selling winners triggers capital gains tax in a taxable account. The “Rebalancing Bonus” works best in tax-advantaged accounts (IRA/401k) where you can trade freely without tax drag.
What is Geometric Return?
It is the “real” compounded return you eat. Volatility drags down geometric return (Volatility Drag). Rebalancing reduces volatility, thereby mechanically lifting the geometric return closer to the arithmetic average.