Backtest
Run your strategy on historical data; cost assumptions and what the metrics mean.
A backtest runs your strategy bar by bar over the historical candles of the symbols you choose.
Settings
- Market: spot or USDⓈ-M futures. On spot, only long strategies and leverage 1.
- Symbols: each symbol runs separately with its own capital; the report shows a per-symbol table and a total row.
- Date range: as many years back as your plan allows.
- Starting capital and leverage (futures).
- Advanced: maker/taker fees and slippage.
Cost assumptions
The defaults are deliberately conservative, so a backtest does not look better than reality would.
| Futures | Spot | |
|---|---|---|
| Maker / taker fee | 0.02% / 0.05% | 0.1% / 0.1% |
| Slippage | 5 bps | 5 bps |
| Fill limit per bar | 10% of bar volume | 10% of bar volume |
- Because of the volume limit, large orders may fill partially.
- Stop and target prices are rounded to the tick away from the entry.
- Liquidation on futures: if equity falls below maintenance margin at the bar's adverse extreme, the position is closed at that price. A single conservative maintenance rate is used; liquidation may happen earlier than in reality, never later.
- Funding fees: on futures, historical funding rates are applied (longs pay and shorts receive when the rate is positive), using the close price at the funding time. Where funding data is missing, no funding is applied and the report shows a warning.
Report
- Equity curve and trade markers on the chart.
- Trades: entry/exit time and price, quantity, fees, funding (futures), net result, R and exit reason (signal, stop-loss, trailing, target, time, opposite signal, liquidation, end of data).
- Events: skipped entries (e.g. the amount is below the exchange minimum), conflicting signals and missing funding data.
Metrics
| Metric | Meaning |
|---|---|
| Net profit, return % | Result including fees and funding |
| CAGR | Compound annual growth rate |
| Max drawdown | Deepest drop from a peak (%, amount, duration) |
| Sharpe / Sortino | Risk-adjusted return from bar returns (risk-free rate 0); Sortino counts only downside volatility |
| Calmar | CAGR / max drawdown |
| Win rate | Share of winning trades |
| Profit factor | Gross profit / gross loss |
| Average trade | Expected net result per trade (expectancy) |
| Average R | Result relative to risk, for trades with a stop-loss |
Annualization is 24/7 for crypto (a year is 365.25 days). The stock-market assumption of 252 trading days is not used.
Limits and retention
- Your daily backtest + optimization allowance depends on your plan and resets at 00:00 UTC; deleting jobs does not give it back.
- The last 50 backtests or 90 days (whichever comes first) are kept.