Tackling the toxic loss creep issue in European NatCat events
The recent trend of European NatCat losses climbing dramatically from initial estimates is an industry-wide phenomenon – and shows the pressing need for a more precise reserve setting after NatCat events. Achieving this requires up-to-date valuations of what is insured, knowledge of what is insured and a better flow of information throughout the insurance value chain.
Article information and share options
Northern Italy’s long and hot heatwave last summer was finally broken in dramatic fashion – with a succession of hailstorms of such severity that windows were smashed, roofs were damaged, and people were injured. Record-breaking hailstones that weighed 365 grams and measured 13 cm in diameter were observed, according to some reports.
CRESTA, an industry data organisation, estimated the storms would be the biggest international industry loss in Q3 2023 with an initial loss estimate of USD 2.2 billion. But for many insurers and reinsurers, the costs kept climbing months after the event and the insured losses ended up being nearly triple the initial estimates at close to USD 6 billion.
Loss creep as seen in Italian severe convective storms (SCS) from July 2023 showing reported losses over time in line with market loss estimations
This loss creep is an extraordinary example of an issue that is gripping the insurance industry – and one that we need to urgently tackle to safeguard the industry's reputation and resilience.
Loss creep – an industry-wide problem
The trend of losses climbing dramatically from initial estimates and resulting in higher-than-expected losses – or loss creep – is an industry-wide phenomenon. And while Italy’s situation presents a very tangible example, severe hailstorms are far from being the only European natural catastrophe where loss creep has occurred. We have seen unprecedented events in Germany in 2021 with flood Bernd and the French hail events from summer 2022 as well as the earthquakes in Turkey/Syria in 2023. All these events have been followed by severe loss creep.
An unsustainable scenario
When it comes to extreme weather events as we have seen in Italy, loss creep fundamentally boils down to the massive underestimation of loss exposure, risk values and inflationary impact. This was exacerbated by insufficient availability of loss adjusters, experts, builders and building materials as well as a lack of claims handling resources at these times of high demand. And although the industry has recognised the growing prevalence of loss creep for some years, it has largely failed to act on many of its drivers.
Given the magnitude of the loss creep issue, we believe the situation needs to be addressed. Especially given that severe convective storm-related insured losses are growing fastest in Europe, where losses grew above trend for a third consecutive year in 2023, according to the Swiss Re Institute’s NatCat sigma study.
“This is a market-wide problem,” says Rita Müller, Head of Claims Western and Southern Europe, at Swiss Re. “As an industry, we need to get better in the early estimation of the real loss exposure after an event. Ongoing underestimation of losses leads not only to underpricing but also to a loss of trust in the industry. As we expect more extreme weather events in the future, systemic loss creep is not acceptable and not sustainable for our industry."
Severe convective storms are both becoming more frequent and more severe, Swiss Re data show, which means that multi-billion event losses are becoming more common.
Annual insured losses from SCS in Europe (USD bn, 2023 prices), and trend
An incomplete data picture – utilising data to improve claims forecasting and reduce loss creep
The root cause for underestimating the cost of extreme weather events comes down to a lack of data about the up-to-date exposure and the current risk values. Frequently, many of the multiple data points re/insurers use to assess risk are missing or outdated. And as an industry, we consistently give too much weight to previous events when modelling losses.
In the case of Italy, for example, government incentives had helped drive up the number of solar panels on roofs. In many cases re/insurers were unaware of these and/or their presumptions around losses didn’t take into account the significant differences in costs in repairing a simple roof versus a smashed solar panel.
“Insurance companies report very little about their exposure to hail in Italy to their reinsurers, for example. For some, we don’t receive any information at all. For others, we just receive very broad information that covers a large area. Only a small number provide this information on a zip code level,” explains Balz Grollimund, Head of Catastrophe Perils at Swiss Re.
Add into the mix a lack of transparency between parties on the data available, and the waters become further muddied.
The rising cost of repairs
In recent years, inflation has been an added complication, raising the price of repair and adding to the argument against leaning too heavily on historic data.
As urbanisation continues apace there are also more high-value buildings concentrated in harm’s way when disaster strikes. Building in high-risk areas, such as flood plains, multiplies this problem.
“We have seen insurers base their initial reserve estimations in French severe convective storms in 2022 on a reserving model from an event that happened in 2014,” says Müller. “You can’t do that. You have to take repair cost inflation, the increase and concentration of values and all other up-to-date exposure factors into account.”
The French hail events of 2014 totalled USD 1.4 billion, and the 2022 hail events ended up at a total of USD 5.5 billion – an increase that is the result of a combination of frequency and severity.
Getting control of loss creep
The argument for limiting loss creep is clear – now we need coordinated action to tackle the problem.
The priority is tackling data and transparency across the value chain – this starts with keeping exposure information up to date at the point of primary underwriting. Valuations need to represent true replacement costs and new investments by homeowners must be captured in a timely fashion.
With a clearer view of the true costs, re/insurers can more accurately price the risks. And as a consequence, initial reserve setting after an event can be based on realistic exposure data rather than outdated historic events.
Linked to this is the need to update initial reserving models to be less biased towards historic data and a slow loss adjustment process. Instead, they must take account of the real exposure at risk, including up-to-date inflation assumptions.
“It’s about getting a few very basic things right from the start,” says Grollimund. “We need to be diligent from the beginning of the value chain to the end about insured risks. We need to have frequent reassessments of the sums insured. And reassessments of what is actually covered by a policy.”
By preventing loss creep from running unchecked, we can move forward with a fair balance: ensuring the sustainability of our exposure assessment and pricing in the industry, while also offering fair premiums and adequate self-retentions for insurers.
And why is this especially important for reinsurers? "As we have seen in the Italy July 2023 storm events, reinsurers are especially hit by underestimation of an event and the subsequent loss creep as they carry the majority of the total loss burden," says Müller.