Monitor & Adjust

The adjust phase of the workflow: watch performance, analyze data, adjust strategy or parameters, validate via backtest, then update; AI Agent can analyze or adjust on your behalf.

This page covers the Adjust phase of the workflow: after a backtest you can review candlesticks and entry / exit points and re-run with new parameters; while a bot is running you can watch its state, returns, and drawdown and adjust the strategy or execution config as needed. You can review performance via the system dashboard, Trade Statistics, and alerts—or hand things off to the AI Agent at any time to analyze performance, interpret data, suggest adjustments, or directly modify the strategy / execution config.

What to watch

DimensionFocus
Trading stateWhether things are running correctly, current positions and P&L, capital changes, fill frequency
PerformanceReturn, win rate, profit factor, max drawdown, etc.
RiskCurrent drawdown, position concentration, anomalies and alerts

The system provides a dashboard (overview and charts), Trade Statistics (multi-level stats, trade records, and K-line entry / exit playback), and alert notifications (outsized losses, failed trades, risk-limit hits, etc.). Alert channels and notification methods are configured in Notification Config.

How to optimize

  • After a backtest: based on the backtest result and the K-line / entry / exit playback, adjust buy / sell conditions or the per-pair execution config and re-run the backtest before going to paper or live.
  • While running: based on Trade Statistics and performance, adjust strategy logic or execution config; re-validate via backtest before updating the trading bot if needed. You can also ask the AI Agent to analyze performance, suggest adjustments, or directly modify the strategy / execution config.

The optimization loop is essentially: Observe → Analyze (including AI) → Adjust strategy / execution config → Re-validate via backtest → Update live / paper. See Backtest Optimization, Execution Config, Risk Management.

Flow overview

Daily observation

Use the dashboard and Trade Statistics to check trading state, return, and drawdown; watch out for alerts (outsized losses, failures, risk-limit hits, etc.).

Analyze and decide

When performance is off or the market regime shifts, diagnose the cause; you can ask the AI Agent to interpret data or suggest tweaks.

Adjust and validate

After changing strategy parameters or per-pair execution config, re-validate via backtest, then update the paper or live bot.

Iterate continuously

After updates, keep an eye on performance and repeat analyze / adjust / validate as needed to close the loop.

Tips

  • Data-driven: adjust only when you have enough data (e.g. at least a stretch of trades and backtests); avoid frequent or blind tweaks.
  • Risk first: when optimizing, prioritize keeping risk controls (stops, position, drawdown) intact—see Risk Management.
  • Avoid overfitting: during backtest and parameter search, watch out-of-sample behavior to avoid overfitting to historical data.

Next steps