Would Prediction Markets Settle the Same Way Today? Kalshi Settlements vs Revised Economic Data

By Arthur Angelo, PredictionScout Research · Updated Oct 23, 2026, 8:38 AM ET

Key findings

• 136 of 1,855 contracts (7.3%) would flip if they were settled on the current official value instead of the first print. They come from 89 of 218 releases.
• Nonfarm payrolls (monthly change): 58 of 365 contracts in 28 of 40 releases would flip; the typical revision is 67K (39 of 40 first prints revised).
• Temporary help services jobs: 4 of 26 contracts in 3 of 5 releases would flip; the typical revision is 7.5K (4 of 5 first prints revised).
• Real GDP growth (advance estimate): 18 of 123 contracts in 9 of 13 releases would flip; the typical revision is 0.4 pt (13 of 13 first prints revised).
• Most recent revised jobs report: Aug 2026 payrolls settled on the first print of +162K; BLS’s current figure is +133K, so 1 of 19 contracts would now pay the other way.
• This is not an error in settlement. Kalshi’s rules say these markets settle on the first published value, and the scorecards on this site verify that they did. The point is how much of a market’s outcome depends on which vintage of the data counts.

Series by series

“Typical revision” is the median absolute difference between the first print Kalshi settled on and the value in today’s official data, across every scored release (unrevised releases count as zero). Each link goes to that series’ full scorecard.

Every release that would settle differently

Newest first. “Yes → No” means the contract paid Yes on the first print but the revised number is at or below the threshold.

Download the data

Release-by-release revisions (CSV)

One row per scored release: first print, current value, revision and flipped thresholds. No prices are included. Free to reuse under CC BY 4.0 with a link back to this page.

Methodology

• Markets. Every settled Kalshi threshold ladder scored on our economic scorecards: KXPAYROLLS, KXADP, KXU3, KXJOBLESSCLAIMS, KXCONTCLAIMS, KXTEMPHELP, KXGDP, KXUSRETAIL, KXPCECORE, KXHOUSINGSTART, KXBUILDPERMS, KXUSHOME. Each event’s contracts and results come from Kalshi’s public API; the contracts counted are the ones that had a price 24 hours before close (the same set the scorecards use).
• First print. The value Kalshi settled on. The scorecards check it against the first-published value in ALFRED; the period is matched through ALFRED’s first-release dates.
• Current value. The same period in the latest FRED vintage (series A191RL1Q225SBEA, ADPMNUSNERSA, CCSA, HOUST, ICSA, MSPNHSUS, PAYEMS, PCEPILFE, PERMIT, RSAFS, TEMPHELPS, UNRATE), pulled on each run, converted to Kalshi’s units and rounded the way the agency publishes it (payrolls, ADP and claims to the nearest thousand, rates and percent changes to 0.1, housing to the nearest thousand units, home prices to $100). Monthly changes are recomputed from current levels.
• Would flip. A contract flips if the current value lands on the other side of its threshold (“above” or “at least”, as each series defines it).
• Limits. Covers only the ladders our scorecards price; FRED’s current vintage can differ from an agency’s website for a few hours after a release; rounding follows each agency’s headline convention, and a value exactly at a threshold depends on it. Nothing here is investment advice. Last run: Oct 23, 2026, 8:38 AM ET.

Data sources: prices and volumes come from the public APIs of Kalshi and Polymarket, with economic data from FRED and ALFRED. All trademarks belong to their owners. PredictionScout is independent and not affiliated with any exchange. Nothing here is investment, betting, or legal advice.

Related: jobs report scorecard · GDP scorecard · markets vs official numbers · All research