The standard cross-venue arbitrage pitch is simple: if the same event is priced differently on two platforms and the combination does not sum to what it logically should, buy the mispriced side and collect the difference regardless of outcome. That pitch has one assumption buried in it: that both platforms are actually pricing the same question. Often they are. Sometimes they are not, and the gap between them is exactly where an apparently riskless position stops being riskless.
What actually happened in November 2025
During the 43-day government shutdown that ended that month, Kalshi and Polymarket ran contracts on when the shutdown would end: the same real-world event, but with different resolution mechanics.
Kalshi’s rule read a specific government website, the Office of Personnel Management page, at a fixed snapshot time on the expiration date—a morning ET close, typically 10:00–11:00 AM. Whatever the page showed at that exact moment settled the contract.
Polymarket ran a ladder of end-date contracts instead. Each required OPM to announce that the government was no longer shut down by 11:59 PM ET that day, thirteen hours later than Kalshi’s snapshot.
On November 12, Congress signed the funding bill and news outlets declared the shutdown over. Polymarket’s “ends November 12” contract traded up to 97 cents. But OPM’s website still showed an active lapse at 11 PM. At midnight, November 12 could no longer be the announcement date under Polymarket’s rule text, so the contract collapsed to about a penny. OPM updated its page the next day, and the November 13 rung settled at a dollar instead.
Both platforms settled correctly according to their published rules. They were not answering the same question. One asked what a webpage showed at a specific morning moment; the other asked when an announcement was made by an end-of-day deadline.
This is not a one-off
It is the same mechanism we found when checking resolution sources directly across categories. Our infrastructure and resolution-source comparison found that Kalshi licenses AP election data directly, while Polymarket requires three outlets to agree before resolving or falls back to certification. The platforms follow structurally different paths for the same event category.
Weather markets diverge even more concretely. In our comparison of 102 resolved city-days, agreement ranged from 61.8% in Miami—where both platforms read the same station and source family—to 32.4% in New York City, where they read different stations. Kalshi has centralized, regulator-backed authority to declare an outcome; Polymarket’s disputed cases can escalate to a decentralized UMA vote. Different authority creates different edge cases.
What this means for treating a position as hedged
Fee drag and slippage are real costs that eat into a spread. Settlement-definition risk is different: it is a binary risk that a position believed to be hedged was actually two different bets wearing the same label. Fees can make a thin spread unprofitable. Settlement-definition risk can make the hedge itself nonexistent, usually near a deadline or resolution boundary when the exact meaning of “when” or “which source” suddenly matters.
What to check before calling a cross-venue position a hedge
Read both platforms’ exact rule text for the specific contract, not just the event or city name. Check whether the deadline is a point-in-time snapshot or a window with a ladder. Check whether both contracts name the same institution, physical station, webpage, and update cadence. None of that is visible in the market title. Rules also change over time, so a rule checked last month is not guaranteed to apply today.
Historical trade and settlement data across Kalshi, Polymarket, Predict.fun, Limitless, Deribit, and Binance options is available through the API docs if you want to check settlement divergence on a specific pair of markets.
FAQ
Did Kalshi and Polymarket actually resolve the same shutdown bet in opposite directions?
Not exactly. Both platforms settled correctly according to their own published rules. They measured different things: a fixed morning snapshot versus an end-of-day announcement deadline. For a trader treating them as equivalent, the practical effect was still a sharp divergence in value on the same day.
Is this the same kind of risk as fee drag or slippage?
No. Fees and slippage erode a real spread. Settlement-definition risk determines whether a position that looked hedged actually was, and it tends to surface near a deadline or resolution boundary.
How can I check whether two markets on different venues actually share a resolution mechanic?
Read each platform’s rules for the exact contract, not the market title. Compare the deadline structure and named source or station, and re-check before relying on it because rules can change between markets.
Does this affect all prediction-market categories equally?
No. It is most visible in fast-moving, deadline-sensitive events and categories where the named source differs by platform. Categories pointing to the same single official source carry less of this specific risk.
Where can I check a specific market’s current resolution rules?
On the market page itself, on both platforms. Do not infer the rules from a past example or a similar-sounding market; check the live rules for the exact contract being traded.