Run your first backtest

A backtest replays a strategy against historical candles and gives you metrics + trades. It's the cheapest, fastest validation before real capital.

Steps

1. Open the strategy detail page.
2. Click Run backtest.
3. Pick a date range (the longer the better, within reason - 12-24 months is a good starting point).
4. Set starting capital (default 10,000 USD).
5. Submit.

The run goes to a queue worker. Most runs finish in seconds.

Reading results

The detail page has four blocks. Walk through them in order:

1. Headline metrics - return, drawdown, Sharpe.
2. Equity curve - shape > final number.
3. Trades table - exit reasons, MAE, MFE.
4. Configuration snapshot - proof of what ran.

Sanity checks

• trade_count >= 30. Smaller samples are noise.
• max drawdown you could tolerate. If not, the strategy is not deployable even with positive expected return.
• Even exit reason mix. Skewed reasons hint at miscalibrated exits.

What to do next

• Re-run with a different date range (out-of-sample test).
• Iterate parameters one at a time.
• Once you have a robust backtest, paper-trade for at least a few days before going live.