The record share of global natural catastrophe losses that was insured in 2025 reflects the long-term trend of gradually expanding insurance protection. On average, more than 40% of annual global economic losses have been covered on average over the past decade, up from only 23% 30 years ago. 

Yet large protection gaps remain, in particular in many emerging economies where 80–90% of catastrophe losses are still uninsured. Even in advanced markets, underinsurance also persists, especially for low-frequency, high-impact risks. In California, for instance, only 12% of residential property insurance policies included coverage for earthquake risk in 2024, with the share having fallen significantly in the decades since the 1994 Northridge earthquake hit greater Los Angeles. 

As exposure grows, hazards intensify and secondary perils lift the annual baseline level of insured losses, a broad-based, multi-stakeholder approach can inform effective strategies aiming to narrow the protection gap.

Catastrophe models are the re/insurance industry’s primary tool for assessing risk. By extending beyond the historical record and simulating a wider range of plausible events, they provide a more complete view of loss potential. Today, the expanding role of secondary perils in driving losses presents a challenge, as modelling of what are often very localised events has begun comparatively recently, leaving gaps in historical data.  

While catastrophe models are getting better and better, uncertainty remains, adding to the urgency of improving models to ensure the most accurate underwriting. Sustained investment in data quality, proactive data sharing between clients, insurers and reinsurers, and regular model updates are all essential to capture the most accurate, up-to-date view of risk. Missing or incomplete data can lead to underestimation of loss potential, as observed in Italy in 2023, when final hail loss estimates were nearly triple initial figures. 

Adaptation can directly reduce losses and support insurability. Between 2021 and 2025, catastrophes accounted for just over half of US homeowners’ insurance claims, up sharply from two decades earlier. Without investment in adaptation and mitigation, underlying loss pressures will continue to affect premiums and underwriting. 

Flood protection has proven highly cost-effective. For instance, UK household flood losses would be 2.8 times higher without the country's existing defences, data from Swiss Re's water-risk intelligence unit, Fathom, indicate.   

In the US, data underscore similar adaptation benefits: in Alabama, evidence from Hurricane Sally in 2020 indicates homes built or retrofitted to the Fortified standard, developed by the US-based Insurance Institute for Business & Home Safety, experienced marked reductions in claim frequency and severity. In Louisiana, where Hurricane Katrina's 2005 landfall remains the costliest insured natural catastrophe event, homes retrofitted to the Fortified standard have seen lower insurance premiums.  

Government-backed schemes and pools that enlist strengths of public and private stakeholders have also enhanced protection in several markets. For instance, Flood Re in the UK and Italy’s recently introduced mandatory natural catastrophe insurance requirement for businesses demonstrate how structured schemes can increase insurance penetration and financial resilience. In Germany, recent proposals following the devastating 2021 floods also seek to align insurance coverage, risk reduction and public support to address lingering protection gaps.

Moreover, reducing non-physical drivers of loss escalation can also support insurance availability. In catastrophe-exposed Florida, reforms enacted since 2022 have materially reduced legal uncertainty for insurers and consumers, helping boost private market participation while easing pressure on premiums. 

A key conclusion from SRI's latest simga report is that stabilising the growth of losses and promoting long-term insurability demands a multi-pronged approach. This includes adaptation that keeps pace with development and risk; constructive partnerships that enlist multiple stakeholders; and accurate data that reflects evolving exposure, hazard and vulnerability.