Tier B — reviewed by Bobi on 2026-09-15. The machine proposed tier A; the human reviewer overrode it.
As of 2026-09-16 12:30 UTC — source: Polyscan verified-pair review, n=1 for this pair
B — Same intended FOMC outcome partition under normal 25bp increments, but not settlement-identical under all possible outcomes. Polymarket explicitly rounds non-standard rate changes to the next 25bp bracket and defines the rate as the upper bound of the federal-funds target range; Kalshi contains neither rule in the supplied criteria. Cancellation/postponement handling also differs. In addition, Kalshi C26/H26 primary rule text omits the ">" present in the market title, creating an internal rule/title inconsistency. Suitable for cross-venue comparison with caveat; not safe to represent as guaranteed identical-settlement arbitrage.
Tier B — 1 material divergence, 2 technical edge cases, 1 source-data anomaly
Translated from the reviewer's original-language analysis; awaiting his confirmation of this English wording.
Cancelled or postponed meeting
Kalshi: If a scheduled FOMC meeting is cancelled and does not occur on its scheduled date, "Fed maintains rate" resolves YES and all others NO.
Polymarket: Waits for the July meeting's statement; if none is released by the end of the next scheduled meeting, resolves to No change.
Scenario: The 28-29 July meeting is postponed and actually held on 31 July with a 25bp hike. Kalshi may already be contractually No change while Polymarket uses the statement from the meeting that actually happened.
Non-standard rate changes
Kalshi: No such rule. The 25bps contract says literally "does a Cut/Hike of 25bps"; >25bps is a separate bucket.
Polymarket: If the change doesn't match a displayed option, it rounds up to the nearest 25bps. Its own example: 12.5bps counts as 25bps.
Scenario: The Fed cuts 12.5bps. Polymarket resolves "25 bps decrease" YES. Kalshi's written rules do not support concluding "Cut 25bps" YES.
Which rate is being measured
Kalshi: Only "the Federal Reserve cuts/hikes rates by X bps".
Polymarket: Explicitly the upper bound of the target federal funds range.
Scenario: The FOMC changes the range asymmetrically -- upper bound stays at 3.75%, lower bound moves 3.50% to 3.25%. Polymarket resolves No change, because it tracks the upper bound. Kalshi's text does not determine it.
About the venue's own data, not a criteria difference
Kalshi rule/title inconsistency (data quality, not criteria)
KXFEDDECISION-26JUL-C26's title and label read "Cut >25bps", but rules_primary reads "Cut of 25bps" -- the ">" is absent, with whitespace in its place. Same for H26. After whitespace normalisation the C26 rule text is byte-identical to C25's. Two different ladder buckets, indistinguishable by rule text.
Independently confirmed as a known issue in Kalshi's published data.
Looks like a divergence — is not one, because of Polymarket's rounding rule. Shown for the same reason the divergences above are — a reviewer who only ever finds problems isn't being rigorous, they're being pessimistic.
| Actual move | Kalshi | Polymarket (after rounding) |
|---|---|---|
| 12.5 bps | No | No (rounds to 25) |
| 25 bps | No | No |
| 25.1 bps | Yes | Yes (rounds to 50) |
| 37.5 bps | Yes | Yes (rounds to 50) |
| 50 bps | Yes | Yes |
The machine reached the right practical mapping here, but not by rigorous reasoning.
Example: KXFEDDECISION-26SEP — this series recurs; showing the nearest archived instance in our data, not necessarily the exact one reviewed.
Fed decision in Sep 2026?
Cut 25bps
If the Federal Reserve does a Cut of 25bps on September 16, 2026, then the market resolves to Yes.
Cut >25bps
If the Federal Reserve does a Cut of 25bps on September 16, 2026, then the market resolves to Yes.
Hike 0bps
If the Federal Reserve does a Hike of 0bps on September 16, 2026, then the market resolves to Yes.
Hike 25bps
If the Federal Reserve does a Hike of 25bps on September 16, 2026, then the market resolves to Yes.
Hike >25bps
If the Federal Reserve does a Hike of >25bps on September 16, 2026, then the market resolves to Yes.
Not yet captured locally — Polymarket's backfill hasn't reached this series yet.
Prices for this series aren't captured yet.