# Prediction Markets Guide

How to use prediction market data for trading insights and decision-making.

---

## Overview

Agencio Predict aggregates data from four major prediction markets:

| Platform | Focus | Data Available |
|----------|-------|----------------|
| **Polymarket** | Crypto-native, wide coverage | Prices, volume, liquidity |
| **Kalshi** | US-regulated, events | Prices, volume, contracts |
| **Metaculus** | Scientific/long-term | Probabilities, forecasts |
| **PredictIt** | Political events | Prices, volume, shares |

Prediction markets provide crowd-sourced probabilities for future events — often more accurate than polls or expert forecasts.

---

## Why Use Prediction Markets?

### Wisdom of Crowds

- Markets aggregate information from thousands of participants
- Participants have "skin in the game" — real money at stake
- Prices reflect collective probability estimates
- Often more accurate than individual experts

### Leading Indicators

- Markets often move before news breaks
- Price changes signal new information entering the market
- Useful for anticipating market-moving events

### Quantified Uncertainty

- Get explicit probability estimates (e.g., "70% chance of rate hike")
- Compare your expectations to market consensus
- Identify opportunities where you disagree with the market

---

## Accessing Prediction Market Data

### Dashboard

1. Go to **Dashboard**
2. Find the **Prediction Markets** tile
3. View trending markets and recent moves

### Dedicated View

1. Go to **Markets > Predictions**
2. Browse all available markets
3. Filter by platform, category, or status

### In Trading Strategies

Use prediction market signals in your AI trading strategies:

```
Entry: polymarket_prob("fed_rate_hike") > 80
Exit: polymarket_prob("fed_rate_hike") < 60
```

---

## Platform Details

### Polymarket

**What it is**: Decentralized prediction market on Polygon blockchain

**Best for**:
- Crypto and DeFi events
- Political outcomes
- Current events
- Sports and entertainment

**Data available**:
- Current probability (price)
- 24h change
- Trading volume
- Liquidity depth
- Historical prices

**Example markets**:
- "Will Bitcoin reach $100k by December?"
- "Will the Fed raise rates in March?"
- "Who will win the presidential election?"

### Kalshi

**What it is**: CFTC-regulated US prediction exchange

**Best for**:
- Economic indicators (CPI, jobs report)
- Weather events
- Financial metrics
- Regulatory outcomes

**Data available**:
- Contract prices
- Volume
- Open interest
- Expiration dates

**Example markets**:
- "Will CPI exceed 3% this month?"
- "Will unemployment rise above 4%?"
- "Will the S&P 500 close above 5000?"

### Metaculus

**What it is**: Forecasting platform focused on science and long-term predictions

**Best for**:
- Technology milestones
- Scientific breakthroughs
- Long-term forecasts
- AI and climate predictions

**Data available**:
- Community median probability
- Prediction count
- Resolution criteria
- Historical accuracy

**Example markets**:
- "When will AGI be achieved?"
- "Will fusion power be commercially viable by 2035?"
- "What will global temperature rise be by 2050?"

### PredictIt

**What it is**: Political prediction market (US-focused)

**Best for**:
- US elections
- Political appointments
- Legislative outcomes
- Policy decisions

**Data available**:
- Share prices
- Volume
- Buy/sell spreads
- Contract limits

**Example markets**:
- "Who will win the Senate race in Georgia?"
- "Will Congress pass the infrastructure bill?"
- "Who will be the next Fed Chair?"

---

## Reading Prediction Market Data

### Understanding Prices

| Price | Meaning |
|-------|---------|
| $0.70 | Market estimates 70% probability |
| $0.50 | Coin flip — maximum uncertainty |
| $0.90 | High confidence (90%) in outcome |
| $0.10 | Low probability (10%) |

### Price Movements

| Movement | Interpretation |
|----------|----------------|
| Sharp rise | New information favoring "Yes" |
| Sharp drop | New information favoring "No" |
| Gradual drift | Slow consensus shift |
| High volatility | Uncertainty, conflicting signals |

### Volume Signals

| Volume Pattern | Meaning |
|----------------|---------|
| High volume + price move | Strong conviction |
| High volume + stable price | Disagreement, active debate |
| Low volume + price move | Thin market, less reliable |
| Volume spike | Breaking news or information |

---

## Using Predictions in Trading

### Event-Driven Strategies

Trade based on prediction market signals:

**Example: Fed Rate Decision**
```
If Kalshi "rate_hike" > 80%:
  - Expect higher rates
  - Consider: short bonds, long financials, short growth stocks

If Kalshi "rate_hike" < 30%:
  - Expect rates held or cut
  - Consider: long bonds, long growth stocks
```

### Contrarian Signals

When prediction markets show extreme probabilities:

| Market Shows | Contrarian View |
|--------------|-----------------|
| > 95% probability | Is the market overconfident? |
| < 5% probability | Is a surprise possible? |
| Rapid 20%+ move | Overreaction to news? |

### Correlation with Assets

Some prediction markets correlate with asset prices:

| Prediction Market | Correlated Assets |
|-------------------|-------------------|
| Fed rate decisions | Bonds, bank stocks, USD |
| Election outcomes | Sector ETFs, policy-sensitive stocks |
| Economic indicators | Broad market indices |
| Crypto regulations | Bitcoin, Ethereum, crypto stocks |

---

## DSL Signals for Strategies

### Available Signals

| Signal | Description | Example |
|--------|-------------|---------|
| `polymarket_prob` | Polymarket probability | `polymarket_prob("event") > 70` |
| `polymarket_change` | 24h probability change | `polymarket_change("event") > 10` |
| `kalshi_prob` | Kalshi probability | `kalshi_prob("cpi_above_3") > 60` |
| `pred_market_high` | Any market above threshold | `pred_market_high(80)` |
| `pred_market_momentum` | Probability trending | `pred_market_momentum("up")` |

### Strategy Examples

**Rate Hike Hedge**
```yaml
name: Fed Rate Hedge
entry:
  when: kalshi_prob("fed_rate_hike") > 75
  action: short TLT
exit:
  when: kalshi_prob("fed_rate_hike") < 50
  or: event_resolved
```

**Election Sector Rotation**
```yaml
name: Election Play
entry:
  when: polymarket_prob("candidate_a_wins") > 60
  action: long XLF  # financials
exit:
  when: polymarket_prob("candidate_a_wins") < 40
```

---

## Alerts and Notifications

### Setting Prediction Alerts

1. Go to **Markets > Predictions**
2. Find a market you want to track
3. Click **Set Alert**
4. Configure:
   - Probability threshold (e.g., > 70%)
   - Direction (rises above / falls below)
   - Notification method (push / email)

### Alert Examples

| Alert Type | Use Case |
|------------|----------|
| Probability > 80% | High-confidence signal to act |
| Probability < 20% | Contrarian opportunity |
| Change > 15% in 24h | Breaking development |
| Volume spike | New information entering |

---

## Historical Analysis

### Viewing History

1. Click on any prediction market
2. Select **History** tab
3. View:
   - Price chart over time
   - Volume history
   - Key events marked

### Backtesting with Predictions

Include prediction market data in backtests:

1. Create strategy with prediction signals
2. Run backtest
3. See how prediction-based entries would have performed

**Note**: Historical prediction data availability varies by platform.

---

## Market Categories

### Economic

- Inflation (CPI, PCE)
- Employment (jobs reports, unemployment)
- GDP growth
- Interest rates
- Trade data

### Political

- Elections (presidential, congressional, international)
- Policy decisions
- Appointments
- Legislation

### Technology

- Product launches
- Company milestones
- Regulatory decisions
- AI developments

### Crypto

- Price milestones
- ETF approvals
- Regulatory actions
- Protocol upgrades

### Other

- Weather events
- Sports outcomes
- Entertainment awards
- Scientific discoveries

---

## Best Practices

### Do

- Use predictions as one input among many
- Consider liquidity before trusting prices
- Track your accuracy vs market accuracy
- Set alerts for important thresholds
- Understand resolution criteria

### Don't

- Treat predictions as certainties
- Ignore thin/illiquid markets' limitations
- Overtrade on small probability changes
- Forget that markets can be wrong
- Miss resolution dates

---

## Limitations

### Market Inefficiencies

- Small markets may have limited participants
- Arbitrage opportunities can exist
- Some markets have position limits
- Liquidity varies significantly

### Data Delays

- Prices update periodically, not tick-by-tick
- Some platforms have API rate limits
- Resolution may take time after events

### Coverage Gaps

- Not all events have prediction markets
- Some regions/topics underrepresented
- Historical data may be incomplete

---

## Troubleshooting

### Data Not Loading

- Check if market is still active
- Some markets close before resolution
- Platform may be experiencing issues

### Prices Seem Wrong

- Check for recent news
- Verify you're looking at the right contract
- Some markets have multiple related contracts

### Can't Find a Market

- Try different search terms
- Check multiple platforms
- Market may have resolved or closed

---

## Related Guides

- [All Seeing Eye](./12-all-seeing-eye.md) — AI market intelligence
- [AI Trader](./18-ai-trader.md) — Strategy creation
- [Algorithm Reference](./14-algorithm-reference.md) — Signal syntax
- [Economic Calendar](./08-economic-calendar.md) — Event tracking

---

*Prediction market data is for informational purposes. Past prediction accuracy does not guarantee future accuracy.*