Nobel Peace Prize 2026: Navi Pillay wins, and our frozen forecast is scored

The 2025 Nobel Peace Prize market moved further before its announcement than any other award in the data. In the 12 hours before the Norwegian Nobel Committee named María Corina Machado on 10 October 2025, her Polymarket price rose 65.8 points, from 1.5% to 67.3%. Across 50 award markets measured the same way, decided at 11 announcements between January 2025 and April 2026, the median change was +0.35 points and the median size of a move, in either direction, was 2.0 points. One other winner rose by more than 20.

Navi Pillay won the 2026 Nobel Peace Prize, according to the official Nobel result, verified on 9 October. Our separate experimental forecast, frozen before the announcement, has been scored. Pillay was outside its seven named candidates; the model assigned 69.2% to that outside group. This was not a prediction of Pillay by name.

The measurement is a timing statistic: how far a price travelled before a scheduled announcement, set against that market's own quiet period. It says nothing about who traded or why.

The table below is a snapshot from 19:00 UTC on 29 September 2026. It is not a live ranking and not a Vultax forecast. The favourite's price was more than twice that of the next candidate, Save the Children. A year earlier the favourite a day out was the same organisation, Sudan's Emergency Response Rooms, at 27%, and it did not win; Machado traded at 0.85%.

The committee can name joint recipients, and the market still pays on one outcome. Polymarket's rules for this market set an order of precedence for that case, so the official list of recipients and the market's winning outcome are recorded separately.

Download table figures:CSVJSON with sources

The scheduled announcement was Friday 9 October at 09:00 UTC, 11:00 in Oslo and 12:00 in Tallinn. Saved records exist for all four planned Polymarket checkpoints: seven days, 24 hours, one hour and ten minutes before the announcement. The record times below identify saved captures, not a guarantee that every candidate had a usable two-sided quote.

A separate order-book archive covers a selected 18 of the 32 listed candidates; it is not an uninterrupted record of every candidate. The historical 2025 market-movement study remains below. A comparable final 12-hour movement result for the 2026 winner has not been added to this page.

Polymarket's own price history reproduces the widely quoted figures exactly: 3.75% at 21:29 UTC on 9 October and 72.8% at 00:00 UTC on 10 October. The move began with a $14,250 fill at 5.0% at 22:42 UTC on 9 October; from the next minute the price never again printed below 5%.

In the first hour of the move 215 fills worth $74,194 came from 71 taker wallets. Over the full 12 hours before the announcement the market traded $946,165 in 4,571 fills, against a quiet-period average of $491 per 12 hours.

Download table figures:CSVJSON with sources

Download chart figures:CSVJSON with sources

Every hour of the 30 days before the window gives one 12-hour price change, 709 in all. The median absolute change was 0.05 points, 99% were within 0.9 points and the largest was 1.15 points.

The final 12 hours moved 57 times that largest change. Over 24 hours the multiple is 74. The 709 windows overlap, so they describe the market's usual range rather than 709 separate trials.

The same statistic was computed for every award in the tape whose result is fixed before a scheduled reveal: Oscars and Grammys categories, NFL and MLB MVPs, the Heisman, the Game Awards, the Ballon d'Or and the Nobel Peace Prize. That is 50 winner markets decided at 11 announcements; 33 of them belong to two Oscar nights, so they are not 50 independent events. The window closes at the start of the ceremony or broadcast.

With its sign kept, the median change was +0.35 points: 26 winners rose, 22 fell and two did not move. The median size of a move, ignoring direction, was 2.0 points. Two winners rose by more than 20 points.

The second, NFL MVP 2024-25 (Josh Allen), rose 48.0 points, which was 0.74 times that market's own largest 12-hour move in the preceding month: it had already moved 64.5 points within 12 hours during those 30 days. Three winners moved outside their own 30-day range, and the favourite 24 hours out won 34 times.

In 48 of the 50 markets, the dollars traded in the final 12 hours exceeded that market's quiet-period average. The median was 13.9 times the average; the next-largest surge after the Nobel was 112 times.

Download table figures:CSVJSON with sources

Download chart figures:CSVJSON with sources

Three days after the 2025 announcement, Kristian Berg Harpviken, director of the Norwegian Nobel Institute and secretary of the committee, told Norway's TV2 that the possible leak was "highly likely" espionage and that the institute would investigate. This study measures the market, not the leak.

A steep line the night before invites the conclusion that somebody knew. Four details decide what such a chart can support, and they belong in its caption.

• The clock - "before the announcement" can mean the night before, the scheduled start or the first news report. This page uses one second before the scheduled 09:00 UTC start, fixed before the result.
• The sampling - prices are 10-minute points, the last one at or before each mark. A point does not say when information reached the market, and a gap is not evidence that nothing happened inside it.
• The quantity - a small trade can move a displayed price, so the 2025 move is reported with its dollars, fills and wallet count beside it.
• The coverage - a clean line for one outcome says nothing about outcomes that were not captured. The record above shows what is stored for 2026 and what is still to come.

Prices are Polymarket's CLOB price history at 10-minute resolution; a price is the last point at or before the instant. The pre-announcement change is the price one second before the scheduled start minus the price 12 hours earlier. The baseline is every 12-hour change on an hourly grid over the 30 days before that window.

Dollars, fills and taker wallets come from the Polymarket-v1 fill tape, where one row is one maker fill. Wallets are counted, never listed. Awards decided live on the night (Eurovision) are kept out of the base rate, as are markets without 30 days of price history.

The 50 markets are the ones that met that rule, described as found. Markets from one ceremony move together and the baseline windows overlap, so nothing here is a probability, a calibration of award markets in general or a forecast of a later announcement.

Vultax froze a first-party baseline at 14:47:58 UTC on 8 October 2026. It assigns 30.8% to the selected recipient being among seven PRIO headline names and 69.2% outside that group. This model is separate from the historical market-movement study and the original market checkpoints.

The announcement and scorecard show the official winner and the recorded v1 categorical Brier score of 0.1082 (lower is better). The separate binary question, whether any official laureate would be among the listed names, scored 0.0947; these are different measures. Download the unchanged forecast, probabilities and source-linked inputs. A score on one prize does not establish persistent skill or a market-beating advantage.