Even quiet hurricane seasons can roar

Atlantic hurricane season may not begin until June 1, but hurricane forecasting season is in full swing: Colorado State University scientists recently issued their early outlook suggesting “somewhat below-normal activity,” with 13 named storms, 6 hurricanes, and two major hurricanes.

Behind their forecast are the irregular, multi-year variations in Pacific winds and sea surface temperatures called the El Niño–Southern Oscillation (ENSO). ENSO's phases – El Niño, neutral, and La Niña – affect weather and climate globally. El Niño typically suppresses Atlantic hurricanes, while La Niña does the opposite. With expectations of a moderate-to-strong El Niño during peak season, this implies less storm activity.

Still, even if these factors indicate slightly fewer hurricanes than average, the risk of a hurricane hitting the United States remains. CSU experts acknowledge this, too, cautioning that regardless of what early signals tell us, "it only takes one hurricane making landfall to make it an active season."

As I recently told a journalist, forecasts are no reason to lower our guard.  Averages don’t capture the impact of individual events. Efforts to strengthen our disaster defenses should continue at pace, with targeted investments in adaptation to make communities more resilient, no matter what each hurricane season actually brings.  

Hurricane forecasts provide hints, not crystal-ball certainty. For this, history is only too instructive. In 1992, when Hurricane Andrew struck Florida, El Niño was forecast to continue through peak hurricane season, and experts forecast “below normal” activity.

In both respects, they were right: El Niño persisted and the storm count trailed the average. Andrew was one of seven named storms and four Atlantic hurricanes, and the only one to make US landfall that year. Nonetheless, Andrew became an industry wake-up call, the costliest hurricane ever until it was surpassed by Hurricane Katrina in 2005.   

Category 5 Andrew highlighted the essential role of reinsurance to absorb shocks and support recovery. As some insurers were pushed into insolvency, companies were expected to better manage their exposures going forward, via measures including reinsurance and improved capital adequacy. This also accelerated development of more sophisticated catastrophe modeling and the catastrophe bond market.

Three decades on, the storm still serves as a case study for the necessity of better building codes and enforcement, as well as storm preparation measures like those promoted by the Insurance Institute for Business & Home Safety, in which Swiss Re is a member. Such measures work: our data indicate improved US building practices to reduce hurricane vulnerability mitigate expected loss development by 60-65%.   

In short, Hurricane Andrew galvanised efforts to narrow the natural catastrophe protection gap – efforts that have lost none of their urgency. Closing this gap should start long before an event, via planning, adaptation, and shared responsibility across public and private stakeholders.

In 2025, global insured natural catastrophe losses were USD 107 billion, a figure that actually was below the long-term trend. This was largely due to good fortune: despite 13 named storms and five hurricanes, three with Category 5 strength, no Atlantic hurricane made US landfall – for the first time in a decade.

In that sense, 2025 was an outlier, with lower losses because peak-loss events did not materialise, not because the risk had diminished. The reality is, average global insured losses have been rising 5–7% annually for decades and are likely to continue to do so, driven by structural factors such as economic growth and increasing asset exposure.

Even without a landfalling hurricane, Swiss Re Institute's latest sigma report documents how secondary perils like the Los Angeles wildfires in January and severe convective storms (SCS) are driving the annual loss baseline higher. Secondary perils have been the fastest-growing segment of natural catastrophe insured losses for the past 55 years, accounting for a record 92% of global losses and 99.9% of losses in the US 2025.

Against this backdrop, an on-trend scenario this year would mean insured losses of USD 148 billion. And that is by no means the upper bound: Swiss Re's proprietary risk models suggest a peak-loss year, typically driven by events like major hurricanes or earthquakes, could mean an insured total of USD 320 billion, a scenario with a 10% chance of occurrence.

CSU's forecast offers an early glimpse of what may lie ahead. It is by definition very uncertain and will be updated in coming months. By then, we'll have an even better understanding of how the 2026 season is shaping up, based on ocean temperatures, atmospheric instability, and wind conditions.

From my perspective, however, hurricane forecasts serve a valuable purpose beyond identifying possibilities. They capture our attention, generate risk-awareness, and reorient society's focus to where it should be: getting ready before the winds arrive.

We can't control how many storms form in the Atlantic basin, but we can prepare for them. The strongest lever we have to increase the insurability and sustainability of our communities is to double down on mitigation and adaptation efforts. Because even during a quiet season, hurricanes can roar.