Prediction Market Arbitrage

Systematic detection of pricing discrepancies across Polymarket, Kalshi, and other prediction markets. When identical events carry different implied probabilities on different platforms, risk-free profit becomes mathematically guaranteed.

Active Arb Opportunities

4.2%

Avg Spread

248

Markets Scanned

Platforms Tracked

How Prediction Market Arbitrage Works

Platform A

YES @ ¢58

Buy YES here

Same Event

Different Prices

Platform B

NO @ ¢35

Buy NO here

Total Cost

¢58 + ¢35 = ¢93

Guaranteed Profit

+7.5%

¢7 per $1 payout

The Core Principle

A binary event resolves to exactly one outcome — YES or NO. If you buy YES on one platform and NO on another, you're guaranteed exactly $1.00 in payout regardless of the result. If the combined cost is less than $1.00, the difference is risk-free profit.

Arb Profit = $1.00 − (P_A(YES) + P_B(NO))
where P_B(NO) = 1 − P_B(YES)

Equivalent: Spread = P_B(YES) − P_A(YES)
If Spread > 0, arb exists buying YES on A + NO on B

Why Spreads Exist

Liquidity fragmentation — capital pools are isolated across platforms, preventing efficient price discovery.

Regulatory asymmetry — Kalshi (CFTC-regulated) has different market structures than Polymarket (crypto-native), creating structural pricing differences.

Information latency — news events cause immediate price movement on one platform while the other lags, sometimes by minutes.

Fee structures — different fee models (maker/taker vs. flat) create effective price differences even at similar quoted prices.

Platform Comparison

Platform	Type	Regulation	Fee Model	Settlement	Markets	Liquidity
Polymarket	Crypto (USDC)	Offshore	~2% taker	UMA oracle	400+	Deep
Kalshi	USD (bank)	CFTC regulated	1¢ + 7% win	Internal oracle	300+	Deep
Metaculus	Reputation	N/A	Free	Community	5,000+	N/A
PredictIt	USD (limited)	CFTC no-action	5% withdrawal + 10% profit	Internal	50+	Thin

⚠️ Not Financial Advice. This tool is for educational and research purposes. Prediction market arbitrage involves real capital, counterparty risk, and regulatory considerations. Always do your own due diligence.

Arbitrage Scanner

Cross-platform price discrepancies ranked by spread magnitude. Opportunities are detected by matching identical or equivalent events across Polymarket, Kalshi, and other platforms.

—

Total Opportunities

—

Best Spread

—

Avg Spread

—

Total Volume

Arbitrage Calculator

Calculate guaranteed profit from cross-platform price discrepancies. Enter YES prices from two platforms for the same binary event to compute the arbitrage.

Cross-Platform Binary Arb Most Common

Platform A — YES Price (¢)

Platform B — YES Price (¢)

Position Size ($)

Days to Resolution

Platform A Fee (%)

Platform B Fee (%)

Strategy —

Buy YES on —

Buy NO on —

Total Cost —

Guaranteed Payout —

Gross Profit —

Net Profit (after fees) —

Return on Capital —

Annualized Return —

Multi-Outcome Arb Advanced

When an event has 3+ outcomes and the sum of all YES prices is less than $1.00, buying all outcomes guarantees profit.

Outcome Prices (¢, comma-separated)

Position Size ($)

Outcomes —

Sum of Prices —

Arb Exists? —

Guaranteed Profit —

Return —

Break-Even Analysis

Market Comparison

Side-by-side pricing comparison of identical events across prediction market platforms. Spreads highlighted where arbitrage may exist.

Event	Category	Resolution	Polymarket	Kalshi	Spread	Status

Spread Distribution

Spreads by Category

Arbitrage Strategies

Systematic approaches to capturing prediction market inefficiencies, from basic cross-platform arb to advanced multi-leg positions.

① Cross-Platform Binary Arbitrage

The simplest and most common strategy. Exploit price differences for the same binary event across two platforms.

1 Identify the same binary event listed on Polymarket and Kalshi with different YES prices.

2 Buy YES on the cheaper platform. Buy NO on the more expensive platform (where NO = 1 - YES is cheaper).

3 If total cost (YES_A + NO_B) < $1.00, the difference is locked-in profit at resolution.

4 Wait for event resolution. Collect $1.00 from the winning side. Net profit = $1.00 − total cost − fees.

Example: "Fed cuts rates in March 2026" — Polymarket YES ¢42, Kalshi YES ¢48. Buy YES on Poly (¢42), NO on Kalshi (¢52). Cost: ¢94. Profit: ¢6 = 6.4% return.

② Multi-Outcome Sum Arbitrage

When a multi-outcome event (e.g., election with 5 candidates) has prices that sum to less than $1.00, buying all outcomes guarantees profit.

1 Find multi-outcome markets where ΣP(outcome_i) < 1.00 (or > 1.00 for selling arb).

2 Buy 1 share of each outcome. Exactly one must win, paying $1.00.

3 If sum < $1.00, profit = $1.00 − Σprices. Scale position size proportionally.

Frequency: Rare on liquid markets. More common on newly listed multi-outcome events before market makers arrive. Typically 1-3% margins that close within hours.

③ Temporal/News Lag Arbitrage

Exploit the information propagation delay between platforms after a major news event.

1 Monitor news feeds for events that directly affect prediction markets (policy announcements, earnings, etc.).

2 When one platform prices in the news faster, buy/sell on the lagging platform at the stale price.

3 Hedge on the updated platform if desired, or hold a directional position if confident.

Risk: Requires speed. Latency is typically 30s–5min. Slippage and thin order books on the lagging platform can eat profits. This is not pure arb — it's statistical edge.

④ Correlated Event Pair Trading

Trade implied correlations between related events when platforms misprice the joint probability.

1 Identify events with logical dependencies (e.g., "Fed raises rates" + "Inflation above 3%" have positive correlation).

2 If platforms price them as independent when they're correlated, construct a position that profits from the correlation.

3 Example: If P(A) = 40%, P(B) = 50%, but P(A∩B) should be 30% (not 20% if independent), exploit the mispricing.

Advanced: Requires statistical modeling. Not pure arb — involves modeling risk. Use conditional probability analysis and historical base rates.

Execution Checklist

Step	Action	Critical?
1	Verify events are truly identical (same resolution criteria, same date)	YES
2	Check order book depth — can you fill at the quoted price?	YES
3	Account for ALL fees (entry, exit, withdrawal, currency conversion)	YES
4	Execute both legs simultaneously (or as close as possible)	HIGH
5	Confirm both positions are filled at expected prices	YES
6	Document the trade: entry prices, sizes, expected resolution date	GOOD PRACTICE
7	Monitor for resolution disputes or rule changes	HIGH
8	Track capital lockup duration vs. annualized return	GOOD PRACTICE

Risk Analysis

Prediction market arbitrage is often called "risk-free," but real-world execution introduces several risk vectors that can erode or eliminate theoretical profits.

Execution Risks Critical

Slippage Risk

Order book depth may not support your position size at the quoted price. A 2¢ spread can vanish if you need to walk up the book.

Timing Risk

Between executing leg 1 and leg 2, prices can move. In volatile markets, the spread may close before you complete both legs.

Fill Uncertainty

Limit orders may not fill. Market orders incur slippage. Partial fills leave you with an unhedged directional position.

Platform Risks Significant

Counterparty Risk

Platform insolvency, hack, or regulatory shutdown can lock up funds. Crypto-based platforms (Polymarket) carry smart contract risk. US-regulated platforms (Kalshi) have SIPC-like protections but are not FDIC insured.

Resolution Dispute

Platforms may interpret event outcomes differently. If Platform A resolves YES but Platform B resolves "ambiguous," your arb collapses.

Rule Changes

Platform fee changes, withdrawal restrictions, or market delistings can affect open positions.

Capital Efficiency Risks

Capital Lockup

Funds are locked until resolution. A 3% arb that takes 6 months to resolve is only 6% annualized — barely above risk-free rates. Opportunity cost matters.

Cross-Platform Capital

You need funded accounts on multiple platforms simultaneously, fragmenting your capital. Transfer times (especially crypto → fiat) add friction.

Fee Erosion

Entry fees, exit fees, withdrawal fees, gas fees (crypto), and currency conversion can collectively eat 2-5% of the position. A 4% gross arb may be 0% net.

Risk-Adjusted Return Framework

Net Return = Gross Spread
  − Entry Fees (both platforms)
  − Exit/Settlement Fees
  − Withdrawal Fees
  − Slippage Estimate
  − Gas/Transfer Costs
  × (1 − Dispute Probability)

Annualized = Net Return × (365 / Days to Resolution)

Minimum threshold: Net Annualized > 15%
Below 10% annualized: not worth the risk

References & Resources

Academic papers, platform documentation, and tools for prediction market arbitrage research.

Academic Literature

[1] Wolfers, J. & Zitzewitz, E. (2004). "Prediction Markets." Journal of Economic Perspectives, 18(2), 107-126.

[2] Arrow, K.J., et al. (2008). "The Promise of Prediction Markets." Science, 320(5878), 877-878.

[3] Manski, C.F. (2006). "Interpreting the Predictions of Prediction Markets." Economics Letters, 91(3), 425-429.

[4] Rothschild, D. (2009). "Forecasting Elections: Comparing Prediction Markets, Polls, and Their Biases." Public Opinion Quarterly, 73(5), 895-916.

[5] Berg, J., Nelson, F. & Rietz, T. (2008). "Prediction Market Accuracy in the Long Run." International Journal of Forecasting, 24(2), 285-300.

[6] Hanson, R. (2003). "Combinatorial Information Market Design." Information Systems Frontiers, 5(1), 107-119.

[7] Page, L. & Clemen, R.T. (2013). "Do Prediction Markets Produce Well-Calibrated Probability Forecasts?" The Economic Journal, 123(568), 491-513.

Platform APIs & Docs

Polymarket CLOB API
https://docs.polymarket.com
REST + WebSocket. Public orderbook, trades, market data. USDC settlement on Polygon.

Kalshi Trade API v2
https://trading-api.readme.io/reference
REST API. Authenticated trading. Public market data endpoints. CFTC-regulated.

Gamma Markets API
https://gamma-api.polymarket.com
Polymarket's market aggregation layer. Simplified access to events and outcomes.

Metaculus API
https://www.metaculus.com/api/
Community forecasting data. Useful as a calibration reference, not for trading.

Tools & Resources

Data Sources

• Polymarket Subgraph — on-chain trade data via The Graph
• Dune Analytics — Polymarket volume dashboards
• Kalshi Market Data — public REST endpoints
• FRED — economic data for resolution verification

Execution Tools

• py-clob-client — Polymarket Python SDK
• Kalshi Python SDK — Official Kalshi trading client
• ccxt — Unified crypto exchange library
• web3.py — Direct contract interaction

Monitoring

• TradingView — Some prediction markets have charts
• Arkham Intelligence — Whale wallet tracking
• Nansen — On-chain analytics
• Custom alerts — Telegram bots for spread notifications

Citation: PredArb is a research tool built by Q for educational analysis of prediction market microstructure. Data is illustrative and should be independently verified before any trading decisions. For the latest market data, consult platform APIs directly.