A windstorm for the insurance market

Storm warning

As the market gradually starts to digest the recent European renewals, Francesca Nyman considers the increasing challenge that European windstorms pose and whether insurers are adequately protected against this risk.

Preliminary estimates out this week suggest the insured losses from Windstorm Dagmar, which tore through Russia and Scandinavia on Christmas day, could top €45m. While not a staggering figure when pitted against 2011’s record single-event cat losses, it is still a significant amount for the third storm in a season that still has two or three months left to run.

Depending on who you ask the answer to the question of whether European windstorm risk is increasing differs greatly. Most agree that there is no discernable upwards loss trend. On the contrary in fact, compared with the last decade of the previous century, which counted Anatol, Lothar and Martin, all in 1999, among its ranks, losses from the storms of the last decades have been relatively minor. That said there has been a key storm every year since 2007. There have also been storms in territories not traditionally thought of as windstorm danger zones.

Storm pattern
“Northwestern Europe- the UK, France and Germany to a lesser extent- tend to see storms. Windstorms can get up to Scandinavia, but they don’t get that far North that often,” says Ryan Ogaard, senior vice president at modelling firm Risk Management Solutions.

While many of the storms of the last twenty years have followed this formula, Dagmar is a recent example of a storm going as far as Scandinavia. And while coastal areas are often the worse hit, it is not the same as with Atlantic hurricanes. As AIR Worldwide ’s managing director Milan Simic explains: “Tropical cyclones take energy from warm waters and lose intensity over land, whereas extropical cyclones, which is what windstorms are, do the opposite. They can gain intensity as they go on. In Xynthia the damage was greatest inland, rather than in coastal areas.”

“Tropical cyclones take energy from warm waters and lose intensity over land, whereas extropical cyclones, which is what windstorms are, do the opposite.” Simic

Moving patterns

While Axa’s reinsurance head Philippe Durieux agrees that the two main exposed countries are the UK and France he has observed an increase in claims for the Southern part of Europe: “There are more storms. Not big storms, but storms where we did not see them before.”

While the lack of reliable historical data makes it impossible to be sure of any cycle, there is some indication of a cyclic nature of losses, correlated with North Atlantic oscillation, according to Simic. While he is not keen to infer anything from this, it is clear that if this is the case then at some point there will be another decade of severe storms like those of the 1990s.

Potential for losses
Even if storms do not increase in frequency or severity, the potential for large insured losses is increasing. The primary reason, according to Zurich’s head of reinsurance Reto Koller is the significant increase in insured values due to high levels of industrialisation along with the increased density of properties across Europe.

In many cases risk perception is as important as the actual risk, and it is a difference in risk perception that Andreas Schraft, head of cat perils at Swiss Re, credits with the current pricing environment in windstorm.

“There are more storms. Not big storms, but storms where we did not see them before.” Durieux

Artificial prices

“Prices are not where they should be. If you compare them to other cat exposures around the world they are too low. Maybe the buyer has a risk perception different from ours. Buyers do not buy cover up to the same return period as they do for other perils such as earthquake.”

Despite a general acceptance that compared to global cat exposures prices for windstorm are two low, few in the market believed that reinsurers can push for significant rate increases any time soon.

Little impact
“Many observers would have expected the intensity of natural catastrophes around the world, coupled with the financial market turmoil, to have had a significant impact on this year’s renewals. In fact, despite the largest sequence of major international losses on record, the European reinsurance market has remained very stable, with continued adequate capacity and only very modest risk adjusted increases on property catastrophe cover,” commented Willis Re’s European managing director David Rainbow.

Dan Beard, a member of Ernst and Young’s actuarial team, expressed a similar sentiment: “As in 2011 there is ample capacity in the European cat market. Some companies have assessed their non-cat business following the event in Japan and New Zealand and may scale back where this isn’t “core”, but even then there is still a significant amount of capacity in the European market, which will mean that rate changes are unlikely.”

“As in 2011 there is ample capacity in the European cat market.” Beard

Benign renewal

This pessimism seems to have been well-founded, after what Bryan Joseph, global actuarial lead at Pricewaterhouse Coopers, termed a “benign” renewal season. Though the rate increases in some pockets of the catastrophe market has led to a stabilising of windstorm rates, there seem to be no large increases on the horizon.

The other key factor which in many ways goes hand in hand with the excess of capacity is loss experience. “People were hopeful that they would get more increases out of it,” says Joseph. “But the way the cat market has become you are not going to get big rate increases unless there is an event with major losses that takes capacity out of the market. Although there has been a steady stream of storms in the last few years, the lack of major losses means that it is still an attractive market. You are not going to see a cat driven rate increase unless there is a major event that takes out one or two European territories and causes lots of insured damage.”

Cat bonds
While traditional reinsurance continues to be relatively inexpensive, increasingly insurers are choosing to supplement their cover with insurance linked securities such as catastrophe bonds.

According to Artemis, the alternative risk transfer and catastrophe bond portal, there were 12 European windstorm bonds placed in 2011, compared with eight in 2010 and four the previous year. This number includes the bonds for multiple territories, but there were also three European windstorm only-bonds placed last year.

“The lack of major losses means that it is still an attractive market.” Joseph

Different pool

One example of this is Axa Global P&C’s €180m Calypso Capital bond, which closed in October 2011. The bond provides cover for European windstorms in Belgium, Denmark, France (excluding overseas territories), Germany, Ireland, Luxembourg, The Netherlands, Norway, Sweden, Switzerland, and the UK.

The primary motivation for seeking cover through catastrophe bonds is the opportunity to access a different pool of capital. The main European windstorm carriers are a fairly concentrated pool– Munich Re, Swiss Re, Hannover Re, Scor and Lloyd’s of London- so transferring the risks to the capital markets allows insurers to diversify.

Reduced risk
There are, however, significant other advantages: “Nearly all bonds are fully colaterised, meaning that there is reduced counterparty risk,” says Koller explaining this usually leads to a speedier payment in the case of a claim. It is extremely rare for traditional reinsurance to be collateralised because it limits the amount of business that a firm can write, which in Schraft’s view “defies the object of reinsurance”.

Cat bonds also tend to be multi-year deals, meaning that they protect the issuer against the market’s price volatility in the aftermath of an event. So why are they less popular? One factor that may previously have dampened the appetite for catastrophe bonds was the fact that as most operated on parametric triggers there was inherent basis risk.

“Nearly all bonds are fully colaterised, meaning that there is reduced counterparty risk.” Koller

Great innovation

However, this has changed with the advent of independent Zurich-based organisation Perils. Despite only being operational for two years, Perils claims to have been responsible for one of the greatest innovations in the European windstorm market: enabling insurers to use Perils market loss triggers, rather than parametric triggers, thereby reducing basis risk.

"Cat bonds are becoming increasingly tailor-made leading to a better matching of the pay-out pattern of the bonds with the losses actually experienced by the company,” says Eduard Held, head of products at Perils.

Great volume
One of Perils main aims is to facilitate risk transfer within the market, including insurance linked securities. While the complex derivatives are unlikely to replace the traditional insurance market anytime soon, many, like Schraft believe Perils’s presence will lead to a greater volume of cat bonds for European windstorm, emulating the success of similar US organisation Property Claims Service. “Perils will be the basis for a lot of innovation going forward,” he says.

There is currently very little sign of other innovation in the European windstorm market and Rupert Flastcher, head of Munich Re’s trading unit, does not think this is likely to change in the near future. “For general retrocession the structures are very well known. There could always be a shift between the different tools but there’s no need for much innovation.”

Going forward the most likely cause of rate increases if there is no major event is thought to be the Eurozone crisis. However, as Rainbow notes, while the debt crisis is beginning to have an effect on insurer’s financial ratings and risk appetite, this has had very little impact on reinsurers at this stage.

The original insured property losses based on the PERILS methodology are:


Storm name

Event start date

Oringinal loss*

Captured Territories


3 Dec 1999


Denmark, Germany


26 Dec 1999


Belgium, France, Germany, Luxembourg, Switzerland


27 December 1999


France, Switzerland


26 Oct 2006


Belgium, France, Germany, the Netherlands, Switzerland, the United Kingdom


18 Jan 2007


Belgium, France, Germany, Ireland, Luxembourg, the Netherlands, Switzerland, the United Kingdom

* Insured property loss as at the time of the event, exclusing government insurance schemes. Forex rates as at event start date. Lossses are available per country. Note that extrapolation to market level is performed on country loss data (as opposed to per-CRESTA zone loss data) and is based on a smaller market coverage than currently captured by PERILS. As a result, theindustry loss estimates for these five storm events will not be as accurate as the loss estimates produced by PERILS for windstorms Klaus (Jan 2009) and Xynthia (Feb 2010).



Storm name

Event start date

Oringinal loss*

Captured Territories


24 Jan 2009


CRESTA: France


28 Feb 2010


CRESTA: Belgium, France, Germany, the Netherlands, Luxembourg, Switzerland

• Insured property loss as at the time of the event, excluding government insurance schemes. Forex rates as at event start date. Losses are available per CRESTA and property lines of business.



Spotlight on clustering

The spate of European windstorms in the last few weeks has increased the focus on the phenomenon of clustering: serial storms hitting a particular region in one season with little time lapse between them. Clustering now is factored into the catastrophe models of all three main vendors, but it was not always given such prominance.

“Evidence of windstorms arriving in succession within one season, such as when Anatol, Lothar and Martin all arrived in 1999, supported our focus on the modelling of clustering some years ago. It is very important to understand the phenomenon because it can have a significant impact on the tail risk of a company’s exposure.”” says James Webb, Eqecat’s product manager and marketing lead.

Attritional losses from numerous smaller storms can become almost as costly a single major event, as AIR Worldwide ’s managing director Milan Simic , explains: “A number of smaller storms in a single year start to add up. Air’s model simulates storms in real years so the model can calculate yearly aggregates. The annual aggregate figure can be much higher than that for a single storm.”


  • LinkedIn  
  • Save this article
  • Print this page  

You need to sign in to use this feature. If you don’t have an Insurance Post account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an indvidual account here: