As of Friday, September 11, 2026, the Atlantic Ocean has not produced a single hurricane. Five storms have been named this season and all five are gone. The strongest of them topped out at 60 mph, which is 14 mph short of the 74 mph that turns a tropical storm into a hurricane. In the satellite era, no first Atlantic hurricane has ever formed later than September 11. That record is shared by Gustav and Humberto, in 2002 and 2013. Today the Atlantic ties it, and the National Hurricane Center's 8 a.m. outlook said plainly: "Tropical cyclone formation is not expected during the next 7 days."
Before anything else: a hurricane-free season is good news. Nobody on the coast is hoping for one. What follows is about how a market prices an absence, which turns out to be the hardest thing a market ever has to do.
Kalshi runs a contract on all 21 names on the 2026 list, asking which one belongs to the first Atlantic hurricane of the year. We handed the whole board to nine AI models and showed them no prices at all. They came back agreeing with the market about the order of the names and disagreeing sharply about something the market has no way to sell you.
The Quick Answer
Kalshi's board makes Fay the favorite to be the first Atlantic hurricane of 2026, with buyers bidding 35 cents and sellers asking 38. Gonzalo and Hanna follow. Our price-blind panel of nine models keeps that exact order but shaves every name down, landing Fay at a 29.5% blend. Its biggest departure from the market is not a name at all: the panel puts 30.1% on the Atlantic finishing 2026 with no hurricane whatsoever, an outcome with no contract on the board. Five of the nine seats made that the single most likely answer. — The Model Verdict panel
- Market Favorite: Fay, 35 cents bid and 38 cents asked
- Panel Favorite: no hurricane at all, 30.1%, a hair ahead of Fay at 29.5%
- Where The Spread Does The Talking: on Isaias and Kyle the panel lands close to the bid and nowhere near the ask, and those two quotes are each 11 cents wide
- The Number Underneath Everything: the panel gives any newly named Atlantic storm a 30% chance of ever reaching hurricane strength, against a long-run rate near 50%
- The Panel's Logic In One Line: names are handed out in strict alphabetical order, so this board is really a question about how many more storms form and how few of them survive a record-hostile Atlantic
Stay ahead of the markets.
Daily insights and expert picks on Kalshi, Polymarket, and what's moving markets.
Free forever. Unsubscribe anytime.
The Market At A Glance
| Market | Which name belongs to the first Atlantic hurricane of 2026 |
| Venue | Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of September 2026) |
| Ticker | KXFIRSTHURRICANE-26DEC01ATL (21 separate yes/no contracts, one per name) |
| Settles | YES only for the first Atlantic storm classified as a hurricane between May 15 and December 1, 2026 |
| If No Hurricane Forms | Every one of the 21 contracts settles NO. There is no contract that pays for the absence |
| Market Favorite | Fay, 35 cents bid and 38 cents asked |
| Panel Favorite | No hurricane at all, 30.1% |
| Prices As Of | Kalshi bids and asks read September 11, 2026 |
| Verdicts Generated | September 11, 2026, price-blind |
These are model estimates, not predictions of fact and not financial advice or betting advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate. Prices on this page were read live and refresh as the season develops, and the "as of" date above carries the timestamp. New to weather contracts? Our guide to how named-storm contracts actually settle walks through the rules before you read a single price.
Why Five Storms Came And Went Without A Hurricane
Every model saw the same fetched picture of the season, so it is worth laying out the facts that moved the numbers before you read them.
The 2026 Atlantic has named Arthur, Bertha, Cristobal, Dolly and Edouard. Arthur lasted a day in June and peaked at 45 mph. Bertha managed 60 mph across five days in July. Cristobal and Dolly each lived about 24 hours in August at 45 and 40 mph. Edouard, the most recent, reached 60 mph and fell apart on September 3. Not one of them came within 14 mph of the threshold, and the National Hurricane Center's 2026 advisory archive lists all five as tropical storms.
Five separate storms failing the same way is one condition doing the same thing five times.
- A Very Strong El Niño Is Sitting On The Basin. The Climate Prediction Center's September 10 diagnostic discussion has the Niño-3.4 index at +1.8 degrees Celsius for August and puts the odds of a very strong event through the coming winter above 90%. El Niño raises the high-altitude winds over the tropical Atlantic, and those winds tear apart a storm's structure before it can organize. Megan Borowski of the Florida Public Radio Emergency Network called it "potentially the strongest El Niño on record."
- The Wind Shear Has Been Historic, Not Merely Bad. Caribbean shear in July 2026 was the highest since records began in 1979.
- Saharan Dust Kept Arriving. Dry, dusty air blowing off Africa has crossed the Atlantic in repeated plumes this year, drying out systems before their thunderstorms can stack up.
- The Scoreboard Says 7%. Accumulated Cyclone Energy, the standard measure of how much storm a season has actually produced, sits at 4.4 units. A typical Atlantic season has banked about 58.2 by the middle of September. By the same date in 2024 the basin had four hurricanes, including Category 5 Beryl.
The other side of the same coin is worth knowing. While the Atlantic sat still, the Pacific produced 14 named storms and five hurricanes, and Kalshi's Eastern Pacific version of this exact market settled to Fausto back on July 24. El Niño suppresses one basin and feeds the other, so the two records are one story rather than two.
The Board
Here is the full list of contracts, narrow enough to read on a phone. The Kalshi column shows the bid and the ask, because on a market this thin the two are far apart and only the ask is a price you can actually pay. Only eight of the 21 contracts have any bid at all behind them; the other 13 sit at zero bid against a one-cent ask, which is a market that is displayed rather than traded.
The panel's single most likely outcome is not in the table below. No Atlantic storm reaching 74 mph by December 1 carries a 30.1% blend and has no contract at any price, so it cannot appear on a board made of contracts. The section further down walks through how the models got there.
| Contract | What has to happen | Kalshi | Panel |
|---|---|---|---|
| Fay | The next storm to be named becomes a hurricane | 35¢ / 38¢ | 29.5% |
| Gonzalo | Fay forms and fails, Gonzalo gets there | 21¢ / 29¢ | 19.1% |
| Hanna | The first two fail | 13¢ / 16¢ | 11.5% |
| Isaias | The first three fail | 7¢ / 18¢ | 6.3% |
| Josephine | The first four fail | 1¢ / 5¢ | 3.0% |
| Kyle | The first five fail | 1¢ / 12¢ | 1.3% |
| Leah | The first six fail | 0¢ / 1¢ | 0.5% |
| Marco | The first seven fail | 1¢ / 7¢ | 0.2% |
| Nana | The first eight fail | 1¢ / 3¢ | 0.1% |
| Omar | The first nine fail | 0¢ / 1¢ | 0.0% |
| Paulette | The first ten fail | 0¢ / 1¢ | 0.0% |
| Rene | The first 11 fail | 0¢ / 1¢ | 0.0% |
| Sally | The first 12 fail | 0¢ / 1¢ | 0.0% |
| Teddy | The first 13 fail | 0¢ / 1¢ | 0.0% |
| Vicky | The first 14 fail | 0¢ / 1¢ | 0.0% |
| Wilfred | The first 15 fail | 0¢ / 1¢ | 0.0% |
| Arthur | Already formed and dissipated below hurricane strength | 0¢ / 1¢ | 0.0% |
| Bertha | Already formed and dissipated below hurricane strength | 0¢ / 1¢ | 0.0% |
| Cristobal | Already formed and dissipated below hurricane strength | 0¢ / 1¢ | 0.0% |
| Dolly | Already formed and dissipated below hurricane strength | 0¢ / 1¢ | 0.0% |
| Edouard | Already formed and dissipated below hurricane strength | 0¢ / 1¢ | 0.0% |
These are model estimates, not predictions of fact and not financial advice or betting advice.
The last five rows are the cleanest thing on the board. Those storms are finished, so their contracts can no longer pay, and the panel was given that fact and priced them accordingly. The exchange still lists them at a one-cent ask, which is how a contract that can no longer pay continues to appear on a live board.
One more piece of arithmetic explains the whole page. Add up the bids on all 21 contracts and you get 80 cents. Add up the asks and you get 141. Those two sums show the book from both sides. Buying all 21 asks costs 141 cents and returns one dollar if any Atlantic hurricane forms, or nothing at all if none does. Selling into all 21 bids raises 80 cents. Neither figure says anything about a scoreless Atlantic, because no contract on the board pays for one.
Every Seat's Number
The full per-model view for the outcomes that are still live. Every seat put every remaining name below Kyle at under one point, and the five finished storms blend to 0.0%, so those rows are left out here rather than padded. Eight of the nine seats put Arthur through Edouard at a flat zero. The ninth kept 0.05% on Edouard, for the narrow case of a dissipated system regenerating and being given its old name back.
| Outcome | Claude Fable | Claude Opus | Claude Sonnet | GPT | Gemini | Grok | GLM | Kimi | DeepSeek | Blend |
|---|---|---|---|---|---|---|---|---|---|---|
| No Hurricane At All | 26.0 | 27.1 | 30.8 | 22.9 | 30.9 | 32.6 | 30.1 | 27.9 | 37.5 | 30.1 |
| Fay | 29.5 | 31.3 | 27.6 | 31.7 | 27.6 | 27.4 | 29.5 | 29.6 | 27.2 | 29.5 |
| Gonzalo | 19.6 | 19.4 | 18.6 | 20.5 | 18.6 | 18.3 | 19.1 | 19.4 | 17.4 | 19.1 |
| Hanna | 12.1 | 11.3 | 11.5 | 12.4 | 11.5 | 11.1 | 11.2 | 11.7 | 9.9 | 11.5 |
| Isaias | 6.8 | 6.0 | 6.3 | 6.8 | 6.3 | 5.9 | 5.8 | 6.3 | 4.8 | 6.3 |
| Josephine | 3.4 | 2.9 | 3.1 | 3.3 | 3.1 | 2.8 | 2.7 | 3.0 | 2.1 | 3.0 |
| Kyle | 1.5 | 1.2 | 1.3 | 1.5 | 1.3 | 1.2 | 1.1 | 1.3 | 0.8 | 1.3 |
Every seat on this panel is graded against real market settlements — records to date: Claude Fable 88% on 2,983 graded calls · Claude Opus 87% on 3,072 graded calls · Claude Sonnet 86% on 3,048 graded calls · GPT 85% on 9,722 graded calls · Gemini 86% on 8,779 graded calls · Grok 84% on 7,299 graded calls · GLM 83% on 4,020 graded calls · Kimi 86% on 4,394 graded calls · DeepSeek 84% on 4,083 graded calls. Recomputed daily; the full scoreboard is public.
These are model estimates, not predictions of fact and not financial advice or betting advice.
More live boards from the same panel: Hurricane season odds, the Eastern Pacific counterpart to this board, where "more than nine EP hurricanes" now trades at 45¢ bid against a 50¢ ask. How Kalshi weather markets settle, the house guide to the two clocks every weather contract runs on. Prices fetched September 11, 2026.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $215M traded
- 2027 Pro Football Champion · $94M traded
- Pro Baseball Champion · $70M traded
- 2028 U.S. Presidential Election winner? · $65M traded
- 2028 Republican presidential nominee · $64M traded
- Fed decision in Sep 2026? · $46M 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.
Fay, Gonzalo And Hanna: The Names Doing All The Work
The three names at the top hold roughly 60% of the panel's board between them, and the reason is mechanical rather than meteorological. Names are assigned in strict alphabetical order as each system reaches 39 mph, so Fay simply gets the first attempt. Gonzalo only wins if Fay forms and fails. Hanna needs two failures ahead of her. Each step down the list multiplies one more failure into the price, which is why the board collapses so fast.
Fay is where the market and the panel are furthest apart at the top. The ask is 38 cents, which is another way of saying a 38-cent contract turns $100 into about $263 if it hits. The panel says 29.5%. Both numbers are really a claim about the same hidden quantity: the chance that any one newly named storm gets to 74 mph. The market's price on Fay implies something close to 36%. The panel's median is 30%, and every one of the nine seats landed between 28% and 33% after the revision round. Long-run climatology would say 50%, since a normal Atlantic season produces 14 named storms and seven hurricanes. Nobody on the panel would go near 50 after watching five straight storms fail the same way.
Gonzalo at 19.1% against a 21-to-29-cent quote is the panel's most ordinary disagreement. It is the same 30% conversion rate applied one rung further down, and it needs Fay to exist and then disappoint.
Hanna at 11.5% is where the arithmetic starts to bite. Reaching Hanna requires three more named storms to form during the back half of a season that has managed five since June 1, and it requires the first two to fail. Those two conditions pull against each other: a basin active enough to name three more storms is usually a basin capable of making a hurricane out of one of them.
Dark Horses The Panel Won't Dismiss
Below the top three, the panel flagged a few outcomes it specifically refused to round away.
- Josephine (3.0% Blend). The market will sell her for 5 cents and bid only 1. The panel sits at 3.0%, roughly the honest middle of that spread, and it is the last name on the board where the models still see a real path rather than a rounding error.
- Kyle (1.3% Blend). Kyle carries one of the two widest spreads on the board, 1 cent bid against a 12-cent ask. Anyone paying 12 cents for Kyle is paying 11 cents more than the best buyer in the market will offer, which is what a thin book looks like when you inspect it instead of reading the last trade.
- Leah (0.5% Blend). Worth naming for a reason that has nothing to do with weather. Leah is new to this list in 2026, replacing Laura, which the World Meteorological Organization retired after the 2020 Louisiana landfall. The 2026 names are the 2020 names recycled with that one substitution.
- The Back Half Of The Alphabet (0.0% Blend). Paulette through Wilfred are each offered at a cent. Every seat priced them at zero to one decimal place, because reaching them needs a burst of storm activity and a total failure to intensify at the same time.
The Outcome With No Contract
Kalshi's own settlement language is explicit about what happens if the Atlantic never gets there. If no storm is classified as a hurricane between May 15 and December 1, every one of the 21 contracts resolves No. The exchange does not offer the other side of that. A trader who believes the Atlantic finishes 2026 scoreless has no way to express it except by selling names one at a time.
The panel put 30.1% there, and five of the nine seats made it their single most likely outcome. A claim that large is worth showing the working on, and all nine seats arrived by the same two steps.
The first step is how many more storms form. The panel's median is 4.2. Colorado State University's updated forecast calls for nine named storms all year, which implies four more. NOAA's late-summer outlook spans 7 to 13 named storms. Both were issued before the Atlantic reached September without a hurricane, and every seat shaded toward the low end for that reason.
The second step is the 30% conversion rate. Every seat then combined the two the same way, treating the number of storms as itself uncertain rather than as a fixed set of coin flips. Run that way on the panel's median parameters, the chance that none of the remaining storms reaches hurricane strength comes out near 28%. The published 30.1% is a different calculation, the middle of the nine seats' own numbers, which ran from 22.9% to 37.5%.
The honest counterweight is that this has essentially never happened. Only two Atlantic seasons since records began in 1851 finished with zero hurricanes, 1907 and 1914, and both predate satellites and aircraft reconnaissance. Every one of the seven satellite-era seasons that reached September empty still produced at least two hurricanes before it ended. The panel knows this, and it is why nobody went above 38%.
"Satellite-era seasons never finished at zero hurricanes, and all seven late-start analogs eventually produced at least two, so NONE cannot dominate. But +1.8°C El Niño with >90% odds of a very strong event, ACE at 7% of normal, record Caribbean shear, five failed intensification attempts, a blank 7-day TWO past peak, and outlooks issued before this September-zero context all justify a real ~18–19% bucket rather than a historical near-zero." — Grok, in the first round, before revising to 32.6%
The analog the models kept throwing out is instructive. 2022 also reached September without a hurricane and then produced four of them in that month alone, including Fiona and Ian. Every seat that mentioned 2022 disqualified it for the same reason: 2022 was a La Niña year, the opposite background state. The honest comparisons are 2013, which finished with two hurricanes all season, and 2002, where Gustav broke the drought on this same calendar date and Isidore, Kyle and Lili followed.
Where The Panel Changed Its Mind
Each model priced the board alone, then read the other eight boards with the names stripped off and could revise. The revision round tightened this panel more than most: the per-storm hurricane rate started spread across 22% to 40% and finished between 28% and 33%. Two large moves and one refusal are worth reading.
The biggest single revision belonged to the seat that started lowest on the empty outcome, at 11%, and nearly tripled it.
"A naive reading goes wrong by treating 2022's September explosion as the template (wrong ENSO), or by assuming the 50% climatological conversion rate still applies after 0-for-5 with record shear, or by underweighting NONE because 'only 1907 and 1914' finished scoreless without conditioning on the driver state." — GLM, after moving the no-hurricane outcome from 11.0% to 30.1%
The second big move ran the other way, and it is the sharpest piece of reasoning the panel produced. One seat had cut its per-storm rate to 22% and had Fay down at 16.6%. Reading the others, it split the problem in half and raised Fay to 31.3%.
"Genesis is only moderately suppressed: five named storms by September 11 is roughly 70% of normal pace, which is why I do not cut the storm count to 3.5 as several others did. Intensification is severely suppressed: ACE at 4.4 versus 58.2, record 1979-onward Caribbean shear, +1.8C Nino-3.4 with over 90% odds of a very strong event, and 0-for-5 at reaching 74 mph." — Claude Opus
Storms forming and storms strengthening are two different failures, and the second one has been far the more severe of the two. That distinction is what pulled the whole panel toward the same shape.
One seat declined to move toward either camp and said so directly.
"The other forecasters' spread (NONE 11-38%) did not move me: C's 11% underweights the correlated-suppression regime, while B and H's 35-38% require storm counts or conversion rates below even 2013's realized season." — Claude Fable, holding near the middle at 26%
What Would Change The Panel's Mind
Every seat named the developments that would actually move its number, and they cluster on four checkable things.
- A Genesis Area Appearing In The Gulf Or Western Caribbean. Several seats set the bar at an NHC outlook giving a disturbance 40% or better seven-day formation odds. That pushes probability toward Fay and Gonzalo and away from the empty outcome.
- A Second Blank Seven-Day Outlook In Late September. If the basin is still quiet on September 20, the expected storm count falls and the no-hurricane number climbs. The outlook on this page already burns the next seven days.
- Any 2026 Storm Reaching 65 To 70 Mph. A near miss would tell the panel that intensification is possible after all, and it would lift the per-storm rate for every name at once. The current ceiling of 60 mph is what set the rate at 30 rather than 40.
- The Climate Prediction Center Softening Its El Niño Language. A Niño-3.4 reading peaking below +1.5 degrees, or shear anomalies relaxing in October, would raise the conversion rate toward 40% and cut the empty outcome to the high teens.
Settlement timeline.
| Record Tied | September 11, 2026, the satellite-era latest first Atlantic hurricane |
| Record Broken | September 12, 2026, if no Atlantic storm is at hurricane strength |
| Next Outlook Horizon | September 18, 2026, the end of the blank seven-day window quoted above |
| Season Ends | November 30, 2026 |
| Contract Expiration | KXFIRSTHURRICANE-26DEC01ATL settles December 1, 2026 |
| Settlement Source | The per-market rules govern; the exchange names the National Weather Service and NOAA's National Hurricane Center as its non-binding references |
This page is re-scored whenever the story moves, and the prices and blend above carry an "as of September 11, 2026" stamp. Every number here gets graded in public once the market settles; the running record lives on the full graded scoreboard.
More on this: How Kalshi's Rain Markets Decide It Rained · Kalshi High Temp Vs Low Temp Markets: Two Different Games · When Do Kalshi Weather Markets Close? The 24-Hour Clock, Hour By Hour · Kalshi Weather Trading: '100 In Phoenix' Isn't A Trade · What A Kalshi Temperature Contract Actually Is
The Bottom Line
The market and the panel agree about almost everything on this board. Fay leads, Gonzalo is second, the order runs straight down the alphabet, and the back half is worthless. They disagree about one thing, and it happens to be the thing Kalshi cannot list.
A board whose bids sum to 80 cents and whose asks sum to 141 was never asked to price a scoreless Atlantic. The question is not on the menu. The panel, handed the same facts and none of the prices, spent about a third of its probability on exactly that answer. Whether it is right gets settled by December 1, and by a basin that has spent an entire summer refusing to make the wind blow 74 miles an hour.
