Cross-venue price discovery: Polymarket, Kalshi, and the law of one price

When the same event trades in two places, prices can diverge. Why that happens, what the literature implies, and how Algomarket’s arb layer should be read.

Fragmented venues are not new. Equity and futures markets have long confronted the “law of one price”: economically identical claims should trade at similar prices after costs, or arbitrageurs compress the gap. Prediction markets now face a version of that problem as related event contracts appear on multiple platforms - notably crypto-native books like Polymarket and regulated U.S. venues such as Kalshi.

Algomarket’s arbitrage scanner (Pro) surfaces those gaps after normalizing contract identity. This note explains why raw spreads are not free money, and how price-discovery research should frame what you are looking at.

Identical events are rarer than they look

Before microstructure matters, contract design matters. Two markets can share a headline question and still differ in resolution source, cutoff time, early close rules, or payout currency. Prediction-market scholarship repeatedly stresses that informational efficiency depends on well-specified claims (Wolfers & Zitzewitz, 2004; Arrow et al., 2008). A scanner that links venues without checking definitions will invent “arbs” that are really basis risk.

“A quoted cross-venue spread is a research prompt: Are the contracts equivalent? Is one book thinner? Has news hit one venue first?”Algomarket methodology

What can keep prices apart

Fees and spreads - retail fee schedules and bid–ask width can exceed a seemingly large mid-to-mid gap.
Latency and attention - news can hit one venue’s order book before the other; flicker is not a locked trade.
Capital and access - different participant pools (crypto wallets vs. KYC’d exchange accounts) change who can close the gap.
Inventory and hedging - market makers may quote asymmetrically without expressing a directional view.

Identity

Match event definitions first

Costs

Fees + spread eat the gap

Stability

Prefer persistent mispricings

Source: Operational checklist derived from law-of-one-price reasoning; not empirical win rates

How Algomarket uses the research

We treat cross-venue discrepancies as comparative research objects: show both prices, link both venues, and leave execution judgment to the user. That stance mirrors academic caution about reading market prices as clean probabilities (Manski, 2006; Wolfers & Zitzewitz, 2006). The Pro scanner is a desk tool for discovery, not a guarantee of risk-free return after costs.

Free users still benefit from the conceptual frame via the dashboard and screener: understand which books are liquid and how odds sit before chasing venue gaps. Pro unlocks the continuous comparison layer when you are actively hunting fragmentation.

References

Arrow, K. J., et al. (2008). “The Promise of Prediction Markets.” Science, 320(5878), 877–878.
Manski, C. F. (2006). “Interpreting the Predictions of Prediction Markets.” Economics Letters, 91(3), 425–429.
Wolfers, J., & Zitzewitz, E. (2004). “Prediction Markets.” Journal of Economic Perspectives, 18(2), 107–126.
Wolfers, J., & Zitzewitz, E. (2006). “Interpreting Prediction Market Prices as Probabilities.” NBER Working Paper No. 12200.
Hasbrouck, J. (1995). “One Security, Many Markets: Determining the Contributions to Price Discovery.” Journal of Finance, 50(4), 1175–1199.

Start free trial