Polymarket Trading Strategies

This polymarket arbitrage trading bot utilizes the five arbitrage strategies implemented in the bot.

Overview

The polymarket arbitrage trading bot uses multiple strategies to identify arbitrage opportunities on Polymarket prediction markets. Each strategy targets different market inefficiencies and operates across multiple timeframes (5m, 15m, 1h).

For consulting with strategies and purchase, contact me at @ewindmer

Strategy 1: Intra-Market Arbitrage

In binary prediction markets, YES and NO tokens should always sum to $1.00. When the combined price of YES + NO is less than $1.00 (after fees), you can buy both outcomes and guarantee a profit.

Execution

Buy both YES and NO tokens simultaneously, then:

1. Buy YES and NO tokens separately through the CLOB API
2. Merge tokens into complete sets
3. Redeem complete sets immediately for $1.00 per share
4. Profit is realized immediately (no waiting for resolution)

Configuration

• min_spread_pct: Minimum profit % required (default: 1.5%)
• max_position_usd: Maximum position size (default: $500)
• fee_pct: Polymarket fee percentage (default: 2.0%)

---

Strategy 2: Combinatorial Arbitrage

For markets with multiple outcomes (e.g., price ranges), the sum of all outcome prices should equal $1.00. When the total is less than $1.00, buy all outcomes for guaranteed profit.

Execution

Buy all outcome tokens in the market, then:

1. Buy all outcome tokens separately through the CLOB API
2. Merge tokens into complete sets
3. Redeem complete sets immediately for $1.00 per share
4. Profit is realized immediately (no waiting for resolution)

Configuration

• min_deviation_pct: Minimum deviation from $1.00 (default: 2.0%)
• max_position_usd: Maximum position size (default: $300)
• min_outcomes: Minimum number of outcomes required (default: 3)

---

Strategy 3: Cross-Platform Arbitrage

Compares Polymarket prediction prices with actual spot prices from exchanges (Binance/CoinGecko). When there's a significant discrepancy, trade on the assumption that Polymarket will converge to the fair price.

Fair Probability Calculation

The strategy estimates fair probability based on:

• Distance of spot price from strike price
• Price momentum (recent price changes)
• Time to market expiry

Execution

Buy the underpriced outcome (YES or NO) and wait for market correction.

Configuration

• min_price_diff_pct: Minimum price difference to trigger (default: 3.0%)
• max_position_usd: Maximum position size (default: $1000)
• stale_threshold_sec: Maximum age of price data (default: 30s)

---

Strategy 4: Endgame Arbitrage

When a market is close to resolution and one outcome has very high probability (>93%), buy that outcome for a small but near-certain profit.

Execution

Buy the high-probability outcome and hold until resolution.

Configuration

• min_probability: Minimum outcome probability (default: 0.93)
• max_time_to_resolution_hrs: Maximum hours until resolution (default: 48)
• min_annualized_return_pct: Minimum annualized return (default: 100%)
• max_position_usd: Maximum position size (default: $2000)

---

Strategy 5: Momentum/Mean-Reversion

Tracks Polymarket YES/NO prices as a price series and applies technical indicators:

• Z-score: Measures how far current price is from mean
• RSI: Relative Strength Index (oversold/overbought)
• Rate of Change: Price momentum
• VWAP divergence: Price vs volume-weighted average

Entry Conditions

Mean Reversion Buy (Oversold):

• Z-score < -threshold
• RSI < 35
• Rate of change > -0.5%
• Action: Buy YES (expect price to revert up)

Mean Reversion Sell (Overbought):

• Z-score > +threshold
• RSI > 65
• Rate of change < 0.5%
• Action: Buy NO (expect YES price to revert down)

Timeframe Parameters

5-minute (Scalping):

• Lookback: 12 candles (1 hour)
• Entry Z-score: ±2.0σ
• Take profit: 1.5%
• Stop loss: 1.0%

15-minute (Swing):

• Lookback: 16 candles (4 hours)
• Entry Z-score: ±1.8σ
• Take profit: 3.0%
• Stop loss: 2.0%

1-hour (Position):

• Lookback: 24 candles (1 day)
• Entry Z-score: ±1.5σ
• Take profit: 5.0%
• Stop loss: 3.0%

Configuration

Each timeframe has its own parameters in MomentumConfig:

• tf_5m_lookback, tf_5m_entry_zscore, tf_5m_take_profit_pct, tf_5m_stop_loss_pct
• tf_15m_* (same pattern)
• tf_1h_* (same pattern)

---

Signal Ranking

The StrategyAggregator ranks signals using a composite score based on:

• Expected profit % (30% weight)
• Confidence (25% weight)
• Strategy priority (20% weight)
• Urgency (15% weight)
• Risk/reward ratio (10% weight)

Strategy priorities:

1. Intra-market: 1.0 (highest)
2. Combinatorial: 0.95
3. Endgame: 0.90
4. Cross-platform: 0.80
5. Momentum/Mean-reversion: 0.70

Risk Management

• Maximum position size per trade
• Maximum portfolio exposure
• Daily loss limits
• Stop-loss and take-profit levels
• Consecutive loss protection

See risk_manager.py for details.

How Strategies Are Combined

The bot runs all applicable strategies on each market simultaneously, then combines and ranks the results.

Scanning Process

1. Market Discovery: Bot discovers all crypto markets for BTC, ETH, XRP, SOL

2. Parallel Scanning: For each market, all applicable strategies run in parallel:

   • Intra-Market (if binary market)
   • Combinatorial (if 3+ outcomes)
   • Cross-Platform (if strike price can be extracted)
   • Endgame (all markets)
   • Momentum/Mean-Reversion (all binary markets, per timeframe)

3. Signal Collection: All signals from all strategies are collected into a single list

4. Filtering: Low-quality signals are filtered out:

   • Minimum confidence: 0.35 (35%)
   • Minimum profit: 0.5%

5. Ranking: Remaining signals are ranked by composite score

6. Execution: Top-ranked signals are executed (max 3 per scan cycle)

Composite Scoring Formula

Each signal gets a composite score (0.0 to 1.0) calculated as:

score = (profit_score × 0.30) +
        (confidence_score × 0.25) +
        (strategy_priority × 0.20) +
        (urgency_score × 0.15) +
        (risk_reward_score × 0.10)

Where:

• profit_score: Expected profit % / 10 (capped at 1.0)
• confidence_score: Signal confidence (0.0 to 1.0)
• strategy_priority: Strategy type priority (see below)
• urgency_score: HIGH=1.0, MEDIUM=0.67, LOW=0.33
• risk_reward_score: Risk/reward ratio × 5 (capped at 1.0)

Strategy Priority Weights

Strategies are weighted by their inherent reliability:

1. Intra-market: 1.0 (risk-free arbitrage)
2. Combinatorial: 0.95 (risk-free arbitrage)
3. Endgame: 0.90 (high probability)
4. Cross-platform: 0.80 (directional, requires convergence)
5. Momentum/Mean-reversion: 0.70 (technical analysis, less certain)

Example: Multiple Signals on One Market

A single market might generate multiple signals:

Market: "Will BTC be above $100k by Friday?"

1. Intra-Market Signal: YES=$0.45, NO=$0.50 → Combined=$0.95 → Arbitrage opportunity

   • Score: 0.85 (high priority, good profit)

2. Cross-Platform Signal: Spot price suggests 70% probability, but YES is $0.45

   • Score: 0.72 (medium priority, good mispricing)

3. Momentum Signal (5m): Oversold condition detected

   • Score: 0.58 (lower priority, technical signal)

4. Endgame Signal: Market resolves in 2 hours, YES at $0.96

   • Score: 0.91 (high priority, near-certain profit)

Result: Signals ranked: Endgame (0.91) → Intra-Market (0.85) → Cross-Platform (0.72) → Momentum (0.58)

The bot would execute the Endgame signal first, then Intra-Market if capital allows.

Signal Deduplication

The bot can generate multiple signals for the same market from different strategies. The risk manager prevents duplicate positions in the same market, so only the highest-ranked signal will execute.

Timeframe Handling

Some strategies (Cross-Platform, Momentum) run across multiple timeframes (5m, 15m, 1h). Each timeframe generates a separate signal, allowing the bot to:

• Capture short-term scalping opportunities (5m)
• Identify swing trades (15m)
• Take position trades (1h)

Configuration

Strategy parameters can be adjusted in config.py. Each strategy has its own configuration class:

• IntraMarketConfig
• CombinatorialConfig
• CrossPlatformConfig
• EndgameConfig
• MomentumConfig

Strategies can be enabled/disabled individually via the enabled flag in each config.

Adjusting Strategy Weights

To change how strategies are prioritized, modify STRATEGY_PRIORITY in StrategyAggregator class:

STRATEGY_PRIORITY = {
    "intra_market": 1.0,        # Increase for more risk-free arb focus
    "combinatorial": 0.95,
    "endgame": 0.90,
    "cross_platform": 0.80,     # Increase for more directional trades
    "momentum_mean_reversion": 0.70,  # Increase for more technical trades
}

Adjusting Composite Score Weights

To change what factors matter most, modify the weights in _composite_score():

composite = (
    profit_score * 0.30 +      # Increase for profit-focused
    confidence_score * 0.25 +   # Increase for confidence-focused
    strategy_score * 0.20 +    # Increase for strategy-type preference
    urgency_score * 0.15 +     # Increase for time-sensitive trades
    rr_score * 0.10            # Increase for risk/reward focus
)