The Quick Answer
Kalshi runs a set of contracts on how many major hurricanes the East Pacific produces in 2026, and on August 23 the board pays 78.5 cents for "more than three" and 13 cents for "more than eight." The basin has produced one. Our eight-model panel, which never sees the price before it answers, puts the season's likely finish at 3.55 majors and prices the top of that ladder near zero. The full rung-by-rung board, the 55-year storm ledger both sides are arguing over, and the one weather record that argues against us are below.
Free: The Weekly PM Market Brief — the 8-model panel's graded record, the week's biggest market-vs-model gaps, and what's spiking next. One email, Sundays.
New to event contracts on Kalshi? Our guide to how hurricane contracts are built and settled walks through landfall, intensity, and count markets before you read another line of this.
What These Contracts Actually Count
A "major hurricane" is a storm that reaches Category 3 on the Saffir-Simpson scale, which means a peak one-minute sustained wind of 96 knots, or about 110 mph. The East Pacific is the ocean basin the National Hurricane Center watches, running from the coast of the Americas west to 140°W. West of that line is the Central Pacific, and Kalshi trades that basin on a separate board.
The contract language is specific, and it is worth reading before anyone treats the ladder as a bet on a single storm:
If more than 8 hurricanes of category 3 or above occur in the Eastern Pacific between May 15, 2026 and December 01, 2026 as confirmed by the National Weather Service or equivalent national weather service, then the market resolves to Yes.
Every rung shares that sentence with a different number in it. That makes the whole board a season-total question with a fixed close of December 1, 2026, not a storm-by-storm sweat. The settlement authority is the National Weather Service, whose National Hurricane Center publishes the wind estimates the count is built from. Kalshi's weather-event contract terms leave the storm type and the ocean basin as blanks the exchange fills per market, so the sentence above is the one that binds this board.
Two clauses in those terms decide more than most readers would guess. First, if the National Weather Service issues preliminary, revised or corrected reports, "only the first published version that includes confirmation" is used to resolve the contract. Second, events reported after expiration and back-dated do not count unless the agency documented them in its official record before expiration. Hurricane winds get reanalyzed months after a season ends, and storms do get upgraded. None of that reaches this board. The count is whatever the National Hurricane Center had on the record by December 1, 2026, and a storm that gets promoted to Category 3 in next spring's reanalysis was never a major hurricane as far as this contract is concerned.
| Market card | |
|---|---|
| Board | Kalshi, KXHURRICANE-26DEC01EPACMAJ, six rungs |
| Question | How many East Pacific Category 3+ hurricanes in 2026 |
| Window | May 15, 2026 through December 1, 2026 |
| Settles on | National Weather Service confirmation, published through the National Hurricane Center |
| Type | CFTC-regulated event contract, 18+, not a sportsbook product |
| Availability | Kalshi is restricted or limited in several states as of August 2026. Check the platform's own eligibility page. |
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $189M traded
- 2027 Pro Football Champion · $69M traded
- 2028 U.S. Presidential Election winner? · $61M traded
- 2028 Republican presidential nominee · $59M traded
- Pro Baseball Champion · $58M traded
- Bitcoin price at the end of 2026 · $32M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub, and see how every settled call actually scored on the full graded scoreboard.
The Season So Far
Eight named storms have formed in the East Pacific in 2026. Two of them became hurricanes. One of those became a major hurricane.
That storm was Genevieve, which peaked at 140 knots on July 27, a Category 5 and the strongest thing the basin has produced all year. The other hurricane, Fausto, topped out at 90 knots on July 24, six knots short of Category 3. Everything else in the basin has been a tropical storm: Amanda, Boris, Cristina, Douglas and Hernan all peaked at 40 knots, and Elida reached 60.
So the ledger reads one major, banked, with about 100 days left on the clock. Against the 1971 to 2025 record, that is a slow year. Across those 55 seasons the East Pacific averaged 2.29 majors on the board by August 23 and 4.35 for the full season.
The Board
Prices are the midpoint of the live bid and ask, fetched August 23, 2026. Every rung requires that many additional majors, because one is already banked. The last column is not a forecast. It is what actually happened in the six El Niño seasons since 1971, which is the case for buying this board.
| The contract | What it still needs | Kalshi | AI panel | El Niño seasons that added that many |
|---|---|---|---|---|
| More than 3 majors | 3 more by Dec 1 | 78.5¢ | 50% | 5 of 6 |
| More than 4 majors | 4 more | 62¢ | 25.5% | 3 of 6 |
| More than 5 majors | 5 more | 47¢ | 12% | 1 of 6 |
| More than 6 majors | 6 more | 35.5¢ | 4% | 1 of 6 |
| More than 7 majors | 7 more | 19.5¢ | 1.5% | 0 of 6 |
| More than 8 majors | 8 more | 13¢ | 0.5% | 0 of 6 |
Model estimates, not predictions of fact and not financial advice.
One row is not a disagreement at all. On "more than three majors" the panel landed on a coin flip and stayed there: after a round in which every model read the other seven, five of the eight sit between 48% and 50%. That convergence is the finding, not a shrug, and the seats name why the edges cancel. El Niño makes a quiet finish very unlikely, which lifts the rung; the season has banked one major where an El Niño year normally holds 2.67 by now, which caps it. Those two forces net out near even, and the market is paying 78.5 cents for the same outcome.
Read the bottom two rows again. To settle "more than eight," the East Pacific would have to produce eight major hurricanes between August 23 and December 1. In 55 satellite-era seasons the most any year ever added after August 23 was six, and it has happened twice, in 1983 and in 2015. Kalshi pays 13 cents on it. The rung above it, "more than seven," needs seven more and pays 19.5 cents. Neither number has a precedent in the record the contract settles against.
Every Seat's Number
Round-two figures, after each model read the other seven anonymized and could revise. The blend is the median of the eight.
| Contract | Blend | ChatGPT (GPT-5.5) | Claude Fable | Claude Opus | Claude Sonnet | Gemini 3.1 Pro | GLM 5.2 | Kimi K3 | DeepSeek V4 |
|---|---|---|---|---|---|---|---|---|---|
| More than 3 | 50% | 45% | 42% | 48% | 50% | 50% | 58% | 50% | 50% |
| More than 4 | 25.5% | 24% | 20% | 26% | 27% | 28% | 33% | 25% | 25% |
| More than 5 | 12% | 11% | 9% | 12% | 12% | 12% | 14% | 12% | 12% |
| More than 6 | 4% | 4% | 3.5% | 5% | 4% | 4% | 5% | 5% | 4% |
| More than 7 | 1.5% | 1.5% | 1% | 1.5% | 2% | 1% | 2% | 1.5% | 1% |
| More than 8 | 0.5% | 0.5% | 0.3% | 0.5% | 1% | 0.2% | 0.5% | 0.5% | 0% |
Every seat on this panel is graded against real market settlements — records to date: Claude Fable 82% on 616 graded calls · Claude Opus 83% on 686 graded calls · Claude Sonnet 81% on 664 graded calls · GPT 84% on 5,836 graded calls · Gemini 86% on 4,570 graded calls · Kimi 82% on 2,442 graded calls · GLM 81% on 2,465 graded calls · DeepSeek 80% on 2,490 graded calls. Recomputed daily; the full scoreboard is public.
Every number in this piece gets graded in public once the market settles, which is the only reason the records above exist.
More live boards from the same panel: Hurricane Fausto's peak-category ladder, the last Pacific storm this panel priced, has since settled and the panel lost it · Kalshi's live weather hub carries the other three ladders on this same event, including the Central Pacific major board at 67.5¢ for more than one · Harry Kane's Ballon d'Or contract sits at 61.5¢ with the award still two months out. Prices fetched August 23, 2026.
The Case For Buying This Board
The market is not being careless. It is trading a forecast, and the forecast is a good one.
On May 21, 2026, NOAA's Climate Prediction Center issued its East Pacific season outlook and gave a 70% chance of an above-normal year, with a 70%-confidence range of 15 to 22 named storms, 9 to 14 hurricanes, and 5 to 9 major hurricanes. Five to nine. The Kalshi ladder is priced almost exactly on that range.
The reason NOAA went there is El Niño, and El Niño has since arrived with force. The Oceanic Niño Index, the number the Climate Prediction Center uses to define the state of the tropical Pacific, ran from −0.39 last winter to +1.39 for the May-June-July window. On August 13 the agency wrote that there is a greater than 90% chance of a very strong event this fall and winter, and a 69% chance the October-December value exceeds +2.5°C, which would put it in the record books.
The one number the whole board turns on. NOAA's May outlook called for 5 to 9 major hurricanes. The panel's blend calls for 3.55. Everything below is an argument about which of those two the last 100 days of the season will look like.
That matters here because El Niño does opposite things to the two Northern Hemisphere basins. It shreds Atlantic hurricanes with wind shear and it lifts East Pacific ones by removing the same shear and warming the water underneath them. In the six seasons since 1971 with an August-October index at or above +1.0, the East Pacific averaged 6.33 major hurricanes against an all-years average of 4.35. The three strongest of them finished with 7 majors in 1997, 10 in 2015 and 8 in 2023.
The forward pipeline backs the same story. The National Hurricane Center's outlook issued at 5 PM PDT on August 22 gave Invest EP92, southwest of Mexico, a 90% chance of forming inside two days, with a second disturbance at 50% and a third at 50% over a week. The basin is not asleep.
Where The Panel Broke With The Forecast
The eight models were shown the storm ledger, the index, the outlook and the live disturbances. They were not shown a single price. All eight came back well under the market, and the reason they gave is the same one in every seat's answer: the El Niño boost is real, but it is being asked to do work in the wrong part of the calendar.
The six El Niño seasons averaged 6.33 majors, and they got there by banking 2.67 of them before August 23 and adding 3.67 after. The 2026 season has banked one. The late-season surge those years produced came out of an atmosphere that had already proven it could build major hurricanes by midsummer. This one has spent three months making tropical storms that stall out: eight named storms, two hurricanes, a conversion rate of 25% against a basin norm above half.
The arithmetic that follows is unforgiving. In the full 55-season record, the number of majors that arrived after August 23 breaks down like this: nine seasons added none, twelve added one, sixteen added two, nine added three, five added four, two added five, two added six. That is the entire distribution, and its ceiling is six. So "more than three" needs a finish in the top third of that record, "more than five" needs one matched four times in 55 years, and "more than seven" needs a number the basin has never produced in the back half of a season.
Filter the record down to seasons that started as slowly as this one and the ceiling drops again. Sixteen seasons since 1971 held one major or fewer on August 23. Their average final count was 2.19. The highest any of them reached was four, in 1972 and 1987. Not one finished above four, which is the exact outcome the 62-cent rung is selling.
The two seasons that did reach four from this kind of start were both El Niño years, which is the honest counterweight and the reason the panel's "more than three" sits at a coin flip rather than in the teens. The third slow-start El Niño season, 2004, finished with three.
Where The Panel Changed Its Mind
After the first round, each model read the other seven with the names stripped off and could revise. The spread on "more than three" narrowed from 38-to-68 down to 42-to-58, and the movement came from one specific argument.
GLM 5.2 started at 68% on "more than three," the highest number on the board, and cut it to 58%:
The full-season El Niño major mean of 6.33 conflates pre- and post-Aug-22 pace, and this season's pre-Aug-22 pace (2 hurricanes, 1 major) is far behind the ~5.2 hurricanes and ~2.2 majors that El Niño seasons typically banked by this date. I was pricing EP majors as if the full-season El Niño boost applied entirely to the remaining season, which overstated the tail. — GLM 5.2
[Editor's note: the hurricane figure checks out exactly. The six El Niño seasons since 1971 averaged 5.17 East Pacific hurricanes banked by this date. The major-hurricane figure is 2.67, not 2.2; 2.29 is the all-seasons average. The correction runs in favor of the argument, not against it.]
Gemini 3.1 Pro read the same fact and moved the other way, from 45% up to 50%, on the ground that the El Niño lift is proportionally bigger on majors than on hurricanes:
I also moved my EPAC major >3 up from 45% to 50% because [another seat] pointed out the El Niño multiplier on majors (mean 6.33 vs 4.35 climatology) is significantly stronger than on total hurricanes, which I underweighted in Round 1. — Gemini 3.1 Pro
Kimi K3 held its number through both arguments and named why:
I held EP majors >3 at 50 despite [the highest seat's] 68 [...] only 1 major is banked against a typical 2.2 by Aug 22, and the post-Aug-22 major-add distribution (P(+>=3) = 20/55 unconditional) caps even an El-Nino-shifted run near a coin flip. — Kimi K3
Claude Opus made the sharpest version of the point, and it is the one that would move us if the season turns:
El Niño lifts the FLOOR of the remaining-add distribution — no El Niño season added fewer than 3 — but has never produced the record-class tail. — Claude Opus
That is the whole disagreement in one sentence. El Niño makes a quiet finish very unlikely. It has never made a historic one from a start like this.
The Argument Against Us
There is a record on the other side of this, and it is ours.
On July 25 this same panel priced Hurricane Fausto's peak-intensity ladder while the storm was live at 85 knots. The blend said Fausto had a 59% chance of reaching Category 3. Kalshi said two cents. Fausto peaked at 90 knots and the contract settled No on August 10. The market was right by a mile and the panel was not close, on a Pacific hurricane, five weeks ago.
That was not a one-off. Across the panel's graded weather calls, the seats' record when they disagree with the price by ten cents or more runs from 20% to 50% depending on the seat, and the largest sample of those disagreements, 127 calls from the GPT seat, came back at 26.8%. On weather markets specifically, when this panel has fought the price, the price has usually won.
The difference the seats would point to is what they are arguing with. Fausto was a live storm, and the market was reading satellite imagery and hurricane-hunter data that updates every six hours. A season-total board is a different animal: the input is a 55-year count, the horizon is 14 weeks, and the number that decides it is arithmetic rather than a forecast. Whether that distinction holds is the experiment, and it settles on December 1.
One more caution belongs here. The book is thin. All six rungs carry both a bid and an ask, but the spreads run 10 to 18 cents wide, and the whole ladder has traded about 2,188 contracts in its life against roughly 1,590 still open. On a spread that wide the round trip costs a large share of the position, and the midpoints in the board above are a fair-value construction rather than a price anyone actually paid.
What Would Change The Panel's Mind
Four things, each checkable, each with the direction it pushes.
- Two more majors by September 30. That puts the season at three with two months still to run, and "more than four" stops being a stretch. Every rung from "more than three" up moves toward the market.
- The August-October index verifies below +1.0. The Climate Prediction Center publishes the Oceanic Niño Index monthly. If this event tops out weaker than advertised, the panel's floor argument weakens and its whole ladder drops back toward the unconditional base rates, which are lower still.
- A named storm reaches Category 3 before the end of August. EP92 was at a 90% formation chance on August 22. The clock argument the panel leaned on is a statement about intensification failure, and one fast intensifier inside the next eight days is the direct falsification of it.
- A storm reaches Category 3 only after crossing 140°W. The contract says "in the Eastern Pacific" and never defines where that ends. Fausto has already made the question live: it peaked at 90 knots at 130.6°W, then crossed 140°W on July 26 still a hurricane at 75 knots. A storm that forms in the east and only reaches major strength in the west would land on a board whose rules do not say whether it belongs there. That risk sits on every rung, and it points in whichever direction the exchange rules.
The Settlement Timeline
| Date | What happens |
|---|---|
| Now through late September | Climatological peak of the East Pacific season. The 55-year record puts most late-season majors in this window. |
| Early September | Climate Prediction Center publishes the next Oceanic Niño Index value and its monthly ENSO discussion. |
| Mid-October onward | Activity thins out in most years, and not in the big ones. 2015 made three majors after October 15, including Patricia at 185 knots on October 22 and Sandra on November 26. |
| December 1, 2026 | Contract close and settlement, on National Weather Service confirmation. |
This page is re-scored when the story moves, which on this board means a new major hurricane or a new index reading. Prices and storm counts are as of August 23, 2026.
The Bottom Line
The market and the panel are not disagreeing about El Niño. They agree it is here, they agree it is strong, and they agree it raises what the East Pacific can do between now and December. They are disagreeing about a calendar.
Kalshi's ladder is priced close to the range NOAA published in May, before the season had produced a single storm. Since then the basin has made eight named storms and converted one of them into a major hurricane. The panel's read is that the pre-season forecast never got marked to what actually happened, and that the top of the ladder is now selling outcomes with no precedent in the record it settles against: seven more majors in 100 days at 19.5 cents, eight more at 13 cents, when the most any season ever managed after August 23 was six.
The panel has been wrong about a Pacific hurricane recently and publicly, and its weather record against the price is the weakest part of its résumé. Both of those things are on the scoreboard, and so is this. Come back December 1.
These are model estimates, not predictions of fact and not financial advice. Kalshi event contracts are CFTC-regulated and 18+. Availability varies by state; check the platform's eligibility page before assuming access.
Hero illustration: OddsShopper, in the house collage style. Hurricane Genevieve satellite photo by JPSS imagery: CSU/CIRA & NOAA/NESDIS, licensed Public domain; photos cropped, toned, and composited.
