Volatility drag: why variance quietly eats your returns
Two assets with the same average return can end up with very different outcomes once volatility is in the mix. Here is the arithmetic of variance drag, with the formula and a worked example.
The short answer
Volatility drag is the gap between an asset's arithmetic (average) return and its compound (geometric) return. The larger the volatility, the wider that gap: a simple approximation is geometric ≈ arithmetic − volatility²/2. A high-volatility asset must post outsized gains just to match a calmer one with the same average return.
Reviewed Aug 24, 2026
The counter-intuitive part
Most people hear “average return 10%” and picture a straight line to 10% more money. Compounding does not work that way when the path is bumpy. A year of +25% followed by −20% does not net +5% — it nets −1% (1.25 × 0.80 = 1.00, minus fees). The order does not matter; the volatility does.
This is volatility drag, and it is the single most under-taught number in amateur portfolio talk.
The formula, in plain terms
For an asset with arithmetic mean return μ and volatility σ (both as decimals), the approximate compound (geometric) return is:
geometric ≈ μ − σ² / 2
Worked example:
| Asset | Avg return μ | Volatility σ | Drag (σ²/2) | Compound return |
|---|---|---|---|---|
| Calm equity | 8% | 12% | 0.72% | ~7.3% |
| Wild digital | 8% | 60% | 18.0% | ~−10% |
Same average return. The calm asset compounds at roughly 7%; the wild one, despite the identical headline number, loses money on a compound basis. That is the whole point: volatility is not free flavour on top of return, it is a tax on it.
Why this matters for allocation
The drag is why “just buy the highest-return thing” is a trap when that thing is also the most volatile. It is also the mathematical backbone of two real strategies:
- Diversification lowers portfolio variance without lowering expected return much, so it raises the compound return. This is the only free lunch in finance, and drag is why.
- Rebalancing harvests a small premium in volatile markets by systematically selling high and buying low — it directly attacks the drag.
You can see the effect with the cross-asset comparator: set two assets to the same average return and watch the median line separate as you raise volatility.
What we are not claiming
This is a mathematical property of compounding, not a prediction about any asset’s future. Real returns are not normally distributed and tails are fatter than the formula assumes. Treat the rule as a lens for comparing assumptions, not as a price target.
Frequently asked questions
Is volatility drag the same as risk?
Does volatility drag apply to a diversified portfolio?
Can I use the half-variance rule precisely?
Sources & further reading
- 1Investopedia — Volatility draginvestopedia.com
- 2SEC — Investor Bulletin: Understanding Investment Riskinvestor.gov
Written by
Tracy Fang
Founder & quantitative researcher
Systematic-strategy researcher focused on equity factor models, backtest robustness and overfitting diagnostics (parameter plateaus, deflated Sharpe, PBO). Writes the investing and PEMF desks.
- Builds and stress-tests multi-factor equity models
- Publishes reproducible backtests with out-of-sample splits
First published Jul 21, 2026. Last reviewed Aug 24, 2026. Corrections: contact the desk.