Trading bot
What it does
The Trading bot lets you describe a rule-based strategy (entry signals, exit signals, stock universe), run it against historical data to see how it would have performed, and — optionally — let the bot operate live with hard safeguards in place.
When to use it
- You have a rule-based idea and want to see whether it would have worked historically before risking real capital.
- You want to take repetitive trade decisions out of your hands.
- You want consistent execution of a strategy you’d otherwise abandon during drawdowns.
How to use it
The page has four tabs.
Strategy setup
- Define the stock universe (one ticker, a list, an index).
- Choose entry signals (price-based, indicator-based, or both).
- Choose exit signals (target, stop, time-based, signal reversal).
- Save the strategy.
Backtest
- Pick a historical date range.
- Click Run backtest.
- Review the equity curve, trade log, win rate, average return per trade, max drawdown, and Sharpe-style metrics.
Dashboard
- See the bot’s current status (Running, Idle, Error).
- See recent trades the bot has executed (or simulated).
- See performance vs your benchmark.
Controls
- Start / pause / stop the bot.
- Adjust safeguards — maximum loss per day, max position size, max number of open positions, kill-switch threshold.
Tips & gotchas
- Always backtest before going live. A strategy that looks good on paper often unravels in real conditions.
- Set conservative safeguards. They exist to protect you from a runaway bot or a flash event — use them.
- Over-fit strategies don’t generalise. A 90% win rate on one year of data is a red flag, not a green one.
- Slippage and costs matter. Backtests can flatter results — assume real costs are higher.
- The bot can’t trade what you can’t trade. Brokerage limits, instrument availability, and market hours all apply.
- Live operation requires a connected broker with the right permissions enabled on the broker side.