According to a report published by Swiss Re, there has been an upward trend in natural catastrophe-insured losses over the past six years. The data shows a rise of about 5‒7% in average annual losses established over the last 30 years with the possibility of growth in this trend.
This has largely been driven by the rising loss severity of individual catastrophes resulting in the rise of exposures that comes alongside economic development, urbanisation, and population growth, often in areas exposed to natural hazards.
According to Swiss Re, a factor to consider is the mismatch between the assessment of the risks that natural catastrophes pose and actual exposures.
The re/insurance industry still remains in catch-up mode from last year’s losses. The report states that all throughout 2022 the primary and secondary peril events were driven by known risk factors, yet the industry’s valuation of potential losses was below actual outcomes.
This mismatch reflects in declining industry profitability over recent years. Since 2017, the re/insurance industry has paid out $650 billion for weather-related natural catastrophes claims. However, premium income has not kept pace, contributing to a decline in reinsurance sector profitability, with a return on equity down from an annual average of 12% in 2012‒2016 to 7% in 2017‒2021.
Swiss Re notes that the traditional focus of risk assessment has been on tail exposures and capital-threatening events, although recent years’ loss experience showcases a need for focus and discipline on the higher frequency end of loss distributions.
For the purposes of risk assessment, all perils should be given the same attention and resources as afforded to primary hazard exposures, finds the report.
There has been significant progress in natural catastrophe risk modelling capabilities over the last decade, and to further improve these Swiss Re’s report issues some guidelines that can help streamline the process.
Firstly, taking the insurance loss experience of 2022 into consideration, the sharing of peril-specific, granular exposure data is essential. Comprehensive data on existing exposures is the starting point for any underwriting process. Collection and transmission of exposure data of sufficient granularity, specifically for secondary perils (in particular for floods and hail storms), is vital, according to the report.
Insured losses from secondary perils have been on the rise for many years, sometimes reaching the magnitude of losses resulting from medium-size primary peril events. In the case of the flooding in Durban in South Africa in 2022, Swiss Re notes that it is time the industry affords these exposures the same discipline around the monitoring and sharing of exposure data and model results as primary peril risks.
Another important part of underwriting, according to the report, is to ensure that exposure data is updated to include the latest inflation developments. For instance, in the last two years, due to the rise in inflation, the costs of property rebuilds and reconstruction have risen. Inflation effects contributed to the large losses from the floods in Australia in February-March last year where the cost of rebuilding was higher than anticipated by re/insurers. This was due to the inflation impacts of lingering disruptions to global supply chains and pandemic-related border restrictions that had been overlooked.
The report also highlights how past loss experience can be beneficial for natural catastrophe risk assessment. However, it’s important that the assessment not be limited to the recent past but should be forward-looking to capture developments such as changes in weather regimes.
Additionally, Swiss Re argues that historical data points should be translated to represent the current-day risk environment. A holistic and representative trending/debiasing of historic losses should also be taken into account. According to the report, reconstruction and repair costs typically increase faster than consumer price inflation and physical asset values increase faster than the rate of economic growth.
In response, Swiss Re states that consideration of all relevant peril and region-specific loss drivers, including changes in urban development, migration to areas vulnerable to extreme weather events, and enhancements of risk mitigation infrastructure make for effective debiasing of historical loss data.
Finally, regular model updates lead to a gradual shift of risk perspective that has been predicted by the report. Most natural catastrophe events contain learnings, requiring the industry to incorporate these in its risk assessment practices. However, Swiss Re calls warn that bolder changes can be necessary while underwriting and adapting to fast-evolving perils. For instance, the 2022 loss experience from the hailstorms in France and flooding in Australia warrants a reassessment of the respective return period assumptions.
