Skip to main content

Backtesting

Marketing/Content note: Initial scaffolding — expand with annotated screenshots of a real backtest report.

Backtesting runs your strategy against historical market data so you can evaluate it before risking capital.

Reading the metrics

MetricWhat it tells you
Sharpe ratioRisk-adjusted return. Higher is better; compare similar strategies.
Maximum drawdown (MDD)The largest peak-to-trough loss. Lower is better.
Win rateShare of trades that were profitable.
Profit factorGross profit ÷ gross loss. Above 1.0 is profitable in-sample.

Avoid common pitfalls

  • Overfitting — a strategy tuned perfectly to the past often fails live. Prefer simple rules that generalize.
  • Look-ahead bias — make sure rules only use information available at the time of each decision. Horizon's engine is event-driven to prevent this.
  • Ignoring costs — fees and slippage matter; factor them into expectations.
warning

Backtested performance is not a prediction. Past results never guarantee future performance.

Next steps

When results look promising, connect a broker and read Going live.