Who this is for — Anyone comparing strategies with similar average return but different equity curves. Variance quantifies how much results «swing» around the mean.
Variance is the mean of squared deviations of returns from the average. Higher = more unstable capital path. Mathematical base for standard deviation and Sharpe ratio.
In plain terms — How far trades sit from the «typical» outcome — pure instability.
Formula and reading
Variance = Σ (Return − Mean)² / n
| Variance | Equity |
|---|---|
| Low | Smooth, predictable curve |
| High | Wide swings, stress ↑ |
In R: variance 0.25R² → std dev ≈ 0.5R.
Operational use
- Compare strategies at equal positive expectancy
- Alert: rolling variance ↑ → reduce size or frequency
- Not alone: negative expectancy + low variance = slow bleed
Typical mistake — Confusing low variance with low risk — mean reversion with fat tails can show moderate variance and black swans.
Example — Mean +0.3R/trade, variance 0.64R² (σ≈0.8R) vs variance 4R² (σ=2R) — same mean, very different drawdown potential.
Summary card
- Unit: R² or %² — use σ for readability.
- Pair: mean + return distribution.
- Action: persistent σ ↑ → regime review.
Gold path — Edge module. Index: Gold path.