The 2026 North Atlantic hurricane season

Five storms, no hurricanes, and a market that has already moved

A season without a hurricane

The Atlantic has produced five named storms and not one hurricane by the start of September, the slowest opening to a season in almost four decades. The reinsurance market has spent the year pricing as though the pattern will hold.

Storm	Date	Landfall
Arthur	17 June	Texas
Bertha	19 July	Two Gulf landfalls
Cristobal	12 August	At sea
Dolly	27 August	At sea
Edouard	1 September	Louisiana

Climatological peak: 10 September. Average date of the first hurricane: 27 July. Latest first hurricane on record: 11 September (2002, 2013). Source: National Hurricane Center; NOAA tropical cyclone climatology.

Arthur, Bertha, Cristobal, Dolly and Edouard have all been tropical storms. None reached 74mph, the threshold for hurricane status.

The average date of the first Atlantic hurricane across the 60 seasons from 1966 to 2025 was July 27. The latest on record in the satellite era is September 11, a mark set in 2002 and matched in 2013. With nothing organised in the basin and the National Hurricane Center expecting no formation over the following week, 2026 is on course to break it.

Accumulated cyclone energy, the measure that combines the frequency, intensity and duration of storms, was running 84 per cent below the 1991 to 2020 average on August 20. Only four seasons in the satellite era have started more slowly: 1988, 1984, 1977 and 1967.

The cause sits in the Pacific. El Niño, the periodic warming of sea surface temperatures in the central and eastern equatorial Pacific, emerged in June and has strengthened since. It lifts vertical wind shear across the tropical Atlantic and the Caribbean, tearing storms apart before they can organise. Saharan dust and cooler than normal water across parts of the basin have added to the suppression.

Forecasters have cut their numbers twice. NOAA's August update called for 7 to 13 named storms, 2 to 6 hurricanes and no more than 2 major hurricanes, against a long-run average of 14, 7 and 3. It put the chance of a below-normal season at 75 per cent. Colorado State University was tighter still, at 9 named storms, 4 hurricanes and 1 major hurricane, with accumulated cyclone energy at 40 to 45 per cent of its long-term average.

None of that removes the exposure. The statistical peak of the season falls on September 10. More than 85 per cent of Atlantic tropical cyclones form between August and October, and about 85 per cent of major-hurricane activity lands in the six weeks from August 20 to October 10. Almost all of the season's capacity to do damage is still ahead of it.

What has arrived has been water rather than wind. Arthur formed off the mid-Texas coast on June 17 and came ashore southwest of Galveston, driving flash and urban flooding across Texas and the northern Gulf. Aon estimated the insured cost in the hundreds of millions of dollars. Bertha made two Gulf landfalls in July. Edouard came ashore near Johnson Bayou, Louisiana, on September 1 with sustained winds of 60mph and left 18 to 20 inches of rain across parts of Hardin County, Texas, in two days. Rescue operations were still running this week.

Every American East Coast hurricane this season has been a rainfall event.

What quiet has looked like before

Three of the most expensive storms in American history arrived in seasons that were, by the count, unremarkable.

1992 produced 7 named storms, 4 hurricanes and a single major hurricane, and ranks among the least active seasons on record. The major hurricane was Andrew, which crossed south Florida as a category five and caused about $26.5 billion of damage in the dollars of the day, making it the costliest tropical cyclone on record at the time.

1983 was quieter still. 5 named storms, the least active season of the satellite era and the least active of any since 1930. 1 of the 4 was Alicia, which came ashore over Galveston Island on August 18 with sustained winds of 115mph, blew out windows in downtown Houston and caused about $3 billion of damage.

1965 ran under a strong El Niño, the same pattern suppressing the basin this year. 6 named storms, 4 hurricanes, 1 American landfall. That landfall was Betsy, which struck Key Largo as a category three and Grand Isle, Louisiana, as a category four, and which carries normalised losses of more than $13 billion.

The counter-example is 2013, and it is the one the market tends to forget. 14 named storms, only 2 hurricanes and no major hurricane at all, the first season since 1994 without one. The first hurricane, Humberto, did not appear until September 11. Nothing of consequence made landfall. Reinsurance pricing then softened for four consecutive years.

Two lessons sit in that record. Storm counts are a poor guide to loss. And a run of quiet years is where pricing discipline tends to be given away, rather than where it is tested.

Season	Named storms	Notable
1965	6	Betsy — Cat 4 Louisiana landfall
1983	4	Alicia — Cat 3 over Galveston
1992	7	Andrew — Cat 5 over south Florida
2013	14	No major hurricane for the first time since 1994
2026 to date	5	No hurricane of any strength

1991–2020 average: 14 named storms. Source: National Hurricane Center; NOAA historical hurricane records.

The market has already moved

Checkpoint	Index (2024 peak = 100)
2017 soft-market low	58
2024 hard-market peak	100
1 January 2026 renewals	81
Mid-2026 renewals	77

Down 23 per cent from the 2024 peak, and still 32 per cent above the 2017 floor. Guy Carpenter global property catastrophe rate-on-line index. Source: Guy Carpenter, July 2026 renewals report.

Reinsurers have not waited for the season to finish. Guy Carpenter's global property catastrophe rate-on-line index fell 12 per cent at the January renewals and has fallen further since, leaving it down about 16 per cent across 2026. That is the steepest annual decline since the late 1990s and steeper than any single year of the soft market of the 2010s. The index now sits 23 per cent below its 2024 peak, though it remains about 32 per cent above the floor it reached in 2017.

Capital explains most of it. Global reinsurance capital reached a record $790 billion at the end of March, on Aon's measure, with alternative capital at $141 billion. The catastrophe bond market stood at $65.6 billion outstanding at the end of June, up from $61.3 billion at the close of 2025, after a first half of issuance of almost $18 billion. May was the most active month the market has recorded, with 20 transactions and close to $7 billion of new risk capital.

Losses have supported the trend, although the market cannot agree on their size. Swiss Re put global insured catastrophe losses at $42 billion for the first half of 2026, the lowest first half since 2020 and 16 per cent below the ten-year average. Gallagher Re put the figure at at least $46 billion and Aon at at least $47 billion. That is a spread of about 12 per cent on a six-month period that has already closed.

Severe convective storm, not wind, did the damage. Swiss Re attributed $28 billion of the half to it, roughly two thirds of the total. Aon put American severe convective storm losses at about $27 billion, against more than $40 billion over the same months in each of 2023, 2024 and 2025.

Sources: NOAA National Hurricane Center and Climate Prediction Center; Colorado State University Tropical Meteorology Project, August 2026; Swiss Re Institute, first-half 2026 insured catastrophe losses; Gallagher Re and Aon first-half 2026 catastrophe reports; Guy Carpenter Global Property Catastrophe Rate-On-Line Index, July 2026; Aon Reinsurance Market Dynamics, midyear 2026; Artemis Deal Directory and Q2 2026 catastrophe bond market report; Moody's Ratings reinsurance buyer survey, September 2026; Verisk global modelled catastrophe losses; Nature Communications, rapid intensification and marine heatwaves. All figures verified on 3 September 2026.

This document is published by BirdsEyeView Technologies Ltd for general information. It is not underwriting, investment or actuarial advice, and it is not an offer of insurance or reinsurance. Figures are drawn from the public sources listed above and are current at the date of publication. BirdsEyeView is reinsurer-independent and holds no position in any programme referenced.

BirdsEyeView Technologies Ltd | The 2026 North Atlantic hurricane season | September 2026