Disclaimer: Not financial advice. Educational purposes only.

Prediction Market Arbitrage Bot

An educational bot that detects and executes synthetic arbitrage strategies between Polymarket and Kalshi. Built with pmxt - the number 1 unified API for prediction markets.

What is Synthetic Arbitrage?

Traditional arbitrage guarantees profit by exploiting price differences. This bot implements synthetic arbitrage: buying YES on one platform and No on another for the same outcome. If executed simultaneously at favorable prices, you lock in profit when prices converge.

Example:

• Polymarket: Kevin Warsh YES = 41¢
• Kalshi: Kevin Warsh NO = 57¢
• Total cost: 98¢
• Payout if executed: 100¢
• Potential profit: 2¢

Execution Strategy: We aggressively take liquidity using Market Orders on both platforms simultaneously. We do not place limit orders or wait for fills—we instantly seize the best available price to lock in the spread.

Quick Start

1. Install Dependencies

npm install

2. Configure Markets

Edit config.js to set your target markets:

export const config = {
  polymarketUrl: 'https://polymarket.com/event/who-will-trump-nominate-as-fed-chair',
  kalshiUrl: 'https://kalshi.com/markets/kxfedchairnom/fed-chair-nominee/kxfedchairnom-29',
  pollIntervalSeconds: 30,
  minProfitCents: 1,
  tradingMode: 'YOLO', // or 'CONSERVATIVE'
  dryRun: true, // Set to false for live trading
};

3. Set Up Credentials (Optional for Dry Run)

Copy .env.example to .env and add your API keys:

cp .env.example .env

Edit .env:

POLYMARKET_PRIVATE_KEY=your_private_key_here
KALSHI_API_KEY=your_api_key_here
KALSHI_API_SECRET=your_api_secret_here

4. Run the Bot

npm start

How It Works

1. Market Matching

The bot uses fuzzy matching (Jaccard + Levenshtein distance) to pair outcomes:

• "Kevin Warsh" on Polymarket -> "Kevin Warsh" on Kalshi
• Handles slight naming variations automatically

2. Arbitrage Detection

For each matched outcome, calculates both strategies:

• Strategy 1: Buy YES on Polymarket + Buy NO on Kalshi
• Strategy 2: Buy YES on Kalshi + Buy NO on Polymarket

Picks the strategy with maximum profit.

3. Execution (YOLO Mode)

• Finds the best arbitrage opportunity (highest profit)
• Goes ALL IN with available capital
• Places market orders on both platforms
• Waits for fills

4. Exit & Rotation Strategy

• Profit Taking: When markets converge (profit < 1¢), we sell to realize gains.
• Opportunity Rotation: If a better arbitrage opportunity appears (higher spread), we immediately exit the current position to rotate capital into the more profitable trade.

Configuration Options

Option	Description	Default
polymarketUrl	Polymarket event URL	Required
kalshiUrl	Kalshi market URL	Required
pollIntervalSeconds	How often to check prices	30
minProfitCents	Minimum profit to execute trade	1
tradingMode	'YOLO' (all-in) or 'CONSERVATIVE'	'YOLO'
tradeAmountCents	Fixed amount if CONSERVATIVE	100
matchingThreshold	Fuzzy match threshold (0-1)	0.7
dryRun	Test mode (no real trades)	true

Example Output

================================================================================
  PREDICTION MARKET ARBITRAGE BOT
  Educational Demo - Built with https://pmxt.dev
================================================================================

Initializing Arbitrage Bot...
API clients initialized

Bot started! Polling every 30s
Min profit: 1¢
Trading mode: YOLO
Dry run: YES

--------------------------------------------------------------------------------

Fetching markets...
   Polymarket: who-will-trump-nominate-as-fed-chair
   Kalshi: kxfedchairnom

Found 6 Polymarket outcomes, 3 Kalshi outcomes

Matched 3 outcome pairs

Found 1 arbitrage opportunity:

   1. Kevin Warsh
      Buy YES on Polymarket (42¢), Buy NO on Kalshi (57¢)
      Profit: 1.00¢ (0.95 match)

EXECUTING ARBITRAGE OPPORTUNITY!
   Outcome: Kevin Warsh
   Strategy: Buy YES on Polymarket (42¢), Buy NO on Kalshi (57¢)
   Expected Profit: 1.00¢

   [DRY RUN] Would execute: polymarket YES 1000¢ on market poly-123
   [DRY RUN] Would execute: kalshi NO 1000¢ on market kalshi-456

Both trades executed successfully!

Project Structure

prediction-market-arbitrage-bot/
├── config.js              # User configuration
├── .env.example           # Environment variables template
├── package.json           # Dependencies
├── src/
│   ├── index.js          # Entry point
│   ├── bot.js            # Main bot logic
│   ├── matcher.js        # Fuzzy matching algorithms
│   └── arbitrage.js      # Arbitrage calculations
└── README.md             # This file

Disclaimer

This is an educational project to demonstrate algorithmic trading concepts.

• Ignores gas fees, trading fees, and slippage
• Uses simplified market order execution
• Not optimized for real-world profitability
• Use at your own risk

Built With

• pmxt.dev - Unified prediction market API
• Node.js - Runtime
• Pure JavaScript - No frameworks needed

Learn More

• pmxt.dev Documentation
• Polymarket API
• Kalshi API

Disclaimer: Not financial advice. Educational purposes only.

License

MIT