Since the turn of the market in 2023, reinsurers have benefited from the favourable economics, enabling them to outperform industry catastrophe trends, according to JP Morgan’s recent ‘Love Actuary’ report.
In 2023, reinsurance pricing experienced a substantial surge of nearly 30% based on figures from Guy Carpenter.
Beyond these significant rate hikes, structural adjustments within reinsurance programmes shifted in favour of reinsurers, marked by a notable average escalation in attachment points of up to 40%, according to the report.
Concurrently, the historical trajectory of insured catastrophe losses reveals a pronounced long-term, upward trend, driven by both shifting climate patterns and macro factors such as inflation.
The rolling 10 year average for insured catastrophe losses has climbed to approximately $100 billion, a level which has materially increased in recent years.
On an annual basis, insurers and reinsurers make assumptions on the level of catastrophe losses that they will incur in a year. Catastrophe modelling agencies, including Verisk, forecast the level of insured catastrophe losses that they would expect to see in an ‘average’ year called the Average Annual Loss.
This expected baseline, covering elements of tropical cyclones, earthquakes, storms and wildfires, has risen from $81 billion in 2020 to just over $150 billion for 2025, though diverse scenarios make a “usual” year difficult to define, notes the report.
While tropical cyclones, including hurricanes, typically cause material losses, 2011 showed that unexpected sources can also push insured losses. The year had insured losses well in excess of $100bn made up by a combination of events including the Tohoku earthquake, Thai flooding, and New Zealand earthquakes.
“Examining data from Verisk’s forecasts on the average annual expected loss from catastrophes, we can see that the expected level of insured catastrophe losses has increased materially in recent years. The level has increased from ~$80bn in 2020 to ~ $150bn in 2025 driven by trends such as inflation, higher prevalence of catastrophe losses and urbanisation,” JP Morgan highlighted.
By comparing ‘expected’ catastrophe losses to actual catastrophe losses, the analysis concludes that 2020 and 2023 were broadly in line with expectations, while there were higher than expected insured losses in 2021, 2022 and 2024 at the industry level.
In 2025 alone, insured catastrophe losses fell short of the year’s expected baseline. Even with this lower-than-anticipated outcome, the total still exceeded $100 billion, representing a figure that remains elevated on an absolute scale compared to historical levels.
By comparing expected industry losses against actual insured catastrophe figures, the analysis evaluates whether recent years qualified as light, normal, or heavy loss years, benchmarking reinsurer performance against these market baselines.
In 2021 and 2022, active hurricane seasons pushed annual industry catastrophe losses above average levels, leading to catastrophe losses above $120 billion in both years.
These levels compare to an expectation of $87 billion in 2021 and $100 billion in 2022, or 30% above average for these two years in aggregate.
During this high-loss stretch, European reinsurers heavily ‘underperformed’ their benchmarks, according to the report. Across 2021-22, their actual catastrophe losses outpaced their designated catastrophe budgets by approximately 37%.
JP Morgan concluded: “Looking at the evidence 2023-26, industry losses have been ~95% of expected levels using our methodology but the reinsurers have seen their catastrophe budgets ~75% of expected levels showing a step up in performance since the shift in attachment points.
“Reinsurers typically take on tail risk and therefore in a lighter year for insured losses, it might be fair to assume that reinsurers’ performance vs. the industry catastrophe benchmark should outperform and vice versa in a heavy year. But if we look at 2024, a year that had catastrophe losses above expected levels, the reinsurers actually had lower than budgeted levels of catastrophe loss when the benchmark suggested losses should have been higher than normal.”
