Backtesting a Strategy — Trust, but Verify
7/6/2026
basics
guide
Backtesting replays your rules on history: would this have worked, and how badly does it hurt while working?
Level by level
Beginner
Rules must be mechanical enough for a stranger to execute; 30-50 historical setups marked by hand teaches more than any indicator purchase.
Intermediate
Include costs and slippage; judge expectancy, max drawdown and streaks — not just the win rate the course seller advertised.
Advanced
Overfitting is the silent killer: optimising parameters until history looks perfect guarantees the future won't. Out-of-sample splits and walk-forward testing are the honesty tools.
Common mistakes
- Testing on the same data you tuned on
- Ignoring costs — the strategy tax that eats small edges
Practise & tools
_Educational content only — not financial advice. Historical behaviour never guarantees future results._
Keep learning — free tools