Published 2026-10-14
Backtest
A backtest runs a trading rule over historical data to see what it would have done. It is the standard way to evaluate a strategy — and the standard way to fool yourself, if costs, look-ahead bias and survivorship are left out.
What it is
A backtest runs a trading rule over historical data to see what it would have done.
It is the standard way to evaluate a strategy — and, done carelessly, the standard way to fool yourself.
The three things that must be in it
A backtest without all three of these is not a result:
| Must include | Otherwise |
|---|---|
| Look-ahead protection | The rule uses information it could not have had (look-ahead bias) |
| Survivorship handling | The dataset excludes the failures (survivorship bias) |
| Costs and slippage | A rule that trades often looks free when it is not |
This is rule 3 of the standards stated concretely.
How to check
Ask for the parameters: the window, the universe, the cost assumptions.
If none are stated, the curve is a screenshot, not a result.