Research · 6 min read
Reading a backtest honestly: the five numbers that matter
Net return is the least useful headline number. Drawdown, profit factor, win rate and sample size tell the real story.
A backtest is a simulation of how a set of rules would have behaved on past data. It is useful for understanding a strategy’s character — not for predicting what it will earn next month.
The numbers
- Net return — the headline, but meaningless without the risk taken to get it.
- Max drawdown — the deepest peak-to-trough fall in equity. This is what you must be able to sit through.
- Profit factor — gross profit divided by gross loss. Above 1.0 is profitable; above 1.5 is robust in most contexts.
- Win rate — how often trades win. Trend-following systems often win less than half the time and still make money.
- Trade count — fewer than ~30 trades says very little; 100+ gives a more meaningful sample.
What a backtest cannot tell you
Backtests assume fills that may not happen, ignore some slippage, and are inevitably fitted to data the designer has seen. Market regimes change. Treat any backtest — including ours — as a description of the past under stated assumptions, and expect live results to differ.