Backtests
Simulate a strategy or bot against stored candles, then review metrics when the run finishes.
Draft, then run
Create a backtest draft, review settings, then start the run. Targeting a strategy requires an explicit asset (ticker, exchange, asset type). Targeting a bot inherits assets from its instances.
Settings that matter
- Timeframe and date range must overlap stored candle coverage.
- Optional budget, commission, slippage, and sizing assumptions shape modeled results.
- Empty or thin history produces empty or misleading metrics.
Results
Runs compute asynchronously. Open the backtest detail page for progress and results. You can optionally let the in-app assistant watch for completion from AI Agent settings.