PairWatch matches events across Polymarket and Kalshi, verifies the settlement rules line by line, and sends a free alert only when a real, fillable net spread of 0.5% or more survives both venues' fees.
Most cross-venue "arbitrage" is a mirage: different settlement rules, different granularity, or books too thin to fill. PairWatch is built to reject all three.
Every pair is matched and then verified by hand. Near-identical markets that differ in payoff structure, measurement, or timing are rejected — and the rejection is documented, not hidden.
Rounding clauses, cancellation rules, resolution sources, close times. If the two venues can settle the "same" event differently, the alert says so — before you trade, not after.
Books on both venues are walked at 500 shares. An alert fires only if the net spread is ≥ 0.5% after both venues' fees and both legs fully fill. Checks run about every 15 minutes; maximum 5 alerts a day.
Every alert is timestamped in the Telegram channel at the moment it fires — nothing is back-posted or edited. First two days, complete:
| Date | Time (CT) | Market | Net spread |
|---|---|---|---|
| Nov 13, 2026 | 08:33 | Fed decision Mar 2027 — hike >25bp | +1.0% |
| Nov 13, 2026 | 12:20 | Fed decision Dec 2026 — no change | +0.7% |
| Nov 13, 2026 | 14:05 | Fed decision Dec 2026 — hike >25bp | +0.5% |
| Nov 13, 2026 | 18:20 | Fed decision Mar 2027 — hike >25bp | +0.6% |
| Nov 13, 2026 | 20:20 | Fed decision Dec 2026 — hike >25bp | +0.6% |
| Nov 14, 2026 | 00:35 | Fed decision Mar 2027 — hike >25bp | +0.6% |
| Nov 14, 2026 | 04:20 | Fed decision Dec 2026 — hike >25bp | +0.7% |
| Nov 14, 2026 | 06:50 | Fed decision Dec 2026 — hike >25bp | +0.9% |
| Nov 14, 2026 | 06:50 | Fed decision Mar 2027 — hike >25bp | +0.6% |
| Nov 14, 2026 | 12:50 | Fed decision Mar 2027 — hike >25bp | +0.6% |
Looks like the same question. It isn't. From the rejected-pairs library (11 cases and counting):
Polymarket lists daily ranges $2,000 wide. Kalshi lists hourly buckets $100 wide. Same underlying, completely different bets — no pair, no alert.
Polymarket rounds a non-standard rate change up to the nearest 25bp bracket. Kalshi has no equivalent clause. A 20bp hike can settle as different outcomes on the two venues — every Fed alert carries this warning.
One venue's "CPI" market tracks core inflation, the other's tracks headline. Similar titles, different numbers, different settlements. Rejected.
Nothing. Alerts are free on Telegram.
No. It is an intelligence tool only. You verify both books and trade — or don't — yourself.
Verified pairs today concentrate in Fed decisions and NFL moneylines, where both venues genuinely list the same question. Coverage grows pair by pair, only after rule verification — never by bulk matching.
By design. Most apparent spreads fail the fee, depth, or rule-difference test. A tool that alerts on everything alerts on nothing.