Configuring ClawdBot Strategies

The Strategy Engine

ClawdBot doesn't just execute random trades; it follows a strict set of rules defined in your strategy.json file. This allows for deterministic behavior even in volatile market conditions.

Basic Structure

A strategy configuration looks like this:

{
  "name": "election-arbitrage",
  "markets": ["polymarket", "kalshi"],
  "triggers": [
    {
      "condition": "price_divergence > 0.05",
      "action": "buy_lower",
      "max_exposure": 1000
    }
  ]
}

Core Concepts

1. Signals

Signals are data inputs. OpenClaw supports:

• Price: Real-time odds from the order book.
• Volume: Sudden spikes in activity.
• Sentiment: (Requires LLM Plugin) News sentiment analysis.

2. Triggers

Triggers are logic gates. IF a signal crosses a threshold, THEN an action is taken.

• Example: If "Trump 2024" odds on Polymarket are > 5% higher than Kalshi, execute arbitrage.

3. Execution

ClawdBot supports various execution algorithms to minimize slippage:

• TWAP: Time-Weighted Average Price (slow buying).
• Snipe: Immediate execution for arbitrage.
• Maker: Placing limit orders to earn spread.

Risk Management

Always define your global_stop_loss in the config.

"risk": {
  "max_drawdown_daily": "2%",
  "kill_switch_enabled": true
}

This ensures that if your bot malfunctions or the market crashes, the agent will shut itself down to preserve capital.