Generate & Backtest launches a one-shot search of 200 randomized strategies (default basket). Use it to get quick candidates that pass your filters.
Start Continuous keeps generating and testing strategies indefinitely until you click Stop Continuous. This is useful for background research while you fine-tune filters and settings.

Key Parameters

Entry Indicators

Choose which indicators/conditions the randomizer is allowed to use for entries. You can check/uncheck families like:

• Moving Averages (SMA/EMA cross, MA position, Price vs MA, MACD)
• Oscillators (RSI thresholds)
• Breakouts (Highest/Lowest N, Donchian-style)
• Price action (crosses over/crosses under of price vs MA)

The generator respects your selection and limits overall complexity (e.g., max indicators per strategy) to reduce overfitting.

Filters

Only strategies that meet all selected filters are kept:

• Min Score — composite score (Sharpe OOS, drawdown penalty, complexity, etc.)
• Min Profit Factor — wins ÷ losses, typically > 1.2
• Max Drawdown % — peak-to-trough limit over the test
• Max Ulcer Index — measures the depth and duration of drawdowns
• Min Stability — equity straightness (0 to 1); higher = smoother curve
• Min Trades — require enough trades for statistical meaning
• Min Profit $ — net profit threshold in currency
• Max Time in Market % — maximum percentage of time a strategy can be in the market

Viewing a Strategy

Each result includes a View button. Click to open a detailed panel showing:

• Human-readable logic of entries/exits (e.g., “SMA(20) crosses over SMA(50), Exit: SL 2×ATR/TP 3×ATR”).
• Equity curve chart with drawdown and floating PnL (open trade included).
• ProBuilder code you can copy/paste into ProRealTime to backtest or run.

Disclaimer

ProRealQuant is a research tool. You are solely responsible for any trading decisions and outcomes. The generator/backtester provided here does not account for brokerage fees, spreads, slippage, financing, or execution constraints unless explicitly configured. Past performance (including simulated results) is not indicative of future results.