weather and Insurance

The information contained here provides an overview of my research in this area. A more detailed account of this research can found in my PhD thesis (18MB), and references therein.

One of the cornerstones of the (re)insurance industry is the managing of the risk associated with natural hazards, such as earthquakes and flooding. Insuring against adverse effects of weather is a core business. Increasingly, weather related events are dominating not only the news, but also losses to global non-life insurers.

Worldwide, the costs of catastrophic weather events have increased dramatically in recent years, as shown in Figure 1. Insured losses have risen from a negligible level in 1950 to an annual figure exceeding $10bn in the 21st century. When losses from non-catastrophic weather related events are included this figure is doubled. The trend exhibited is influenced by economic, demographic and geographic shifts and well as natural factors. Three quarters of the most expensive 40 insured losses of all time were windstorm related.

Losses from Natural Catastrophes since the 1950s

Figure 1 - Worldwide annual economic and insured losses from large events. Click on image to open in a new window.

The magnitude of losses associated with weather related catastrophes can carry serious economic consequences for the region and people affected, the political body responsible for that region as well as the insurance industry. All three parties have a vested interest in managing the risk posed by severe weather events, and mitigating against them. One of the keys of risk management is the identification and analysis of the risk. Catastrophe modelling is a tool that provides a thorough analysis of the risk, in this case a thorough understanding of the potential impact of weather events at a certain location.

Catastrophe modelling was not widely utilised until several insurers became insolvent following Hurricane Andrew in 1992. Currently, catastrophe models are in widespread use with several independent modelling firms serving the insurance industry. Leading companies, such as EQECAT, Applied Insurance Research (AIR) and Risk Management Solutions (RMS), have developed models covering most climate related perils worldwide.

Generally, a catastrophe model is comprised of four modules; hazard, exposure, vulnerability and loss. The hazard module contains information on the nature of the weather event, incorporating such things as storm track, intensity and duration. Within the exposure module details of the type of properties at risk are stored, including building specifications and age. The third module, vulnerability, is key to producing a realistic catastrophe model. Hazard and exposure are combined to produce a vulnerability curve, which describes the expected damage for a given magnitude or intensity of hazard. A fourth component, loss, is usually employed by the insurers, and incorporates the vulnerability curve and policy conditions, to produce estimates of insured loss.

Generally, insurers make underwriting decisions for risks of economic losses based on information from historical events. Should the pattern of these events be altered, such as a shift in climate, then the basis of these decisions is changed. Owing to this fact, within the finance sector, insurers have been prominent in addressing the climate change issue. However, the short-term nature of insurance means that the topic of potential climate change is met with some scepticism within certain sectors of the industry.

“insurers rely upon their ability to predict the economic consequences of future events. That's how premiums are set; that's the kind of assessment they do of their own exposure. In a period of changing climate, when the very basis of their decisions may be changing, then they need to have a better understanding of climate change. … the fact that future events mat not be a linear progression of the past, but in fact may have changed as a result of natural variability, or human activity or whatever, is an important thing to be taken into consideration”
Franklin Nutter, president, Reinsurance Association of America in a statement to Congress April 1998

"We are still at an early stage in our understanding and analysis of the opportunities and challenges presented by climate change. But work done today will repay the effort many times over in the decades to."

Insurance costs have been rising globally since 1970s (as shown in Figure 1), essentially due to increases in population in regions that are at risk, but also in part to the increase in frequency and severity of certain forms of extreme weather events. It is very difficult to disentangle the socio-economic impacts (such as increasing population, increasing population density in high risk areas, increasing wealth, land use change etc) and natural factors (such as changing patterns of extreme weather).

Global weather-related losses in recent years have been trending upwards much faster than population, GDP, premiums, non-weather related events. Some studies attribute this solely to increased vulnerability, but these studies often overlook the fact that human actions mask losses that would otherwise manifest (such as improved building codes, early warning systems, flood control, crop irrigation, etc).

Overall, the attribution of the increasing trend of insured loss from weather events has not been dealt with in the literature satisfactorily. However, future changes in the severity and frequency of extreme weather events are likely to accelerate the current trend, impacting further on the insurance industry.

"The costs of extreme weather events are already high and rising, with annual losses of around $60 billion since the 1990s (0.2% of World GDP), and record costs of $200 billion in 2005 (more than 0.5% of World GDP). New analysis based on insurance industry data has shown that weather-related catastrophe losses have increased by 2% each year since the 1970s over and above changes in wealth, inflation and population growth/movement. If this trend continued or intensified with rising global temperatures, losses from extreme weather could reach 0.5 - 1% of world GDP by the middle of the century. If temperatures continued to rise over the second half of the century, costs could reach several percent of GDP each year, particularly because the damages increase disproportionately at higher temperatures."

Consensus in the scientific world is that climate change is a reality for this and future generations (IPCC, 2007). Any shifts in the frequency and intensity of extreme events, associated with climate change, will have a direct effect on general insurance, with the greatest impact being on property insurance.

The importance of quantifying the risks posed by climate change to direct insured losses are well illustrated in the following figure. The graph shows a disproportionate increase in loss ratio (insured loss relative to premiums paid) with increasing windspeeds. For instance a 40% increase in windspeeds (from 100 to 140 km/h) produces a 100% increase in the loss ratio.

Figure 2 - Household insurance loss ratios in Great Britain, at various windspeeds recorded in the January and February 1990 storms.
Source: Munich Re (2002)

This illustrates the danger to the insurance industry; a small shift in climate (such as a slight increase in peak gust speed in storms) could have a dramatic impact on insured losses.

The insurance industry finds itself in a unique position with regard to climate change. On the one hand the potential affect of a shifting climate could result in significant financial losses. On the other hand the nature of insurers as risk assessors means they are ideally place to evaluate these impacts. Pre-existing tools, such as catastrophe models, can be adapted to estimate losses in future climates.

Recent hurricane losses in North America have prompted catastrophe modelling firms to reassess the “climatologies” used in their models (McGhee et al., 2007). In other words the storm set used in the hazard module of the catastrophe model, have been upgraded to reflect a period of high activity. RMS increased the landfall frequency of hurricanes in their model following a consultation with four leading hurricane climatologists, who predicted increases of about 20% in category 1-2 hurricane landfall, and 30% in category 3-5, over the next five years relative to 1990-2005 baseline. This had the result of increasing average annual losses within the model by 40% along the Gulf coast, Florida and South-east, and by 25-30% in the Mid-Atlantic and North-east coastal regions relative to those derived using the long-term (1900-2005) historical average hurricane frequencies. RMS, along with other companies, now issue two estimates of loss; one based on the long historical record and one based on this short period of high activity (McGhee et al., 2007).

IPCC (2007). Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden and C.E. Hanson (Eds). Cambridge, UK: 976.

McGhee, C., R. Clarke and J. Collura (2007). The Catastrophe Bond Market at Year-End 2006, MMC Securities: 42.

Munich Re (2002). Winter Storms in Europe (II) - Analysis of 1999 Losses and Loss Potentials. Munich, Munich Reinsurance Company: 76.

Weather and Insurance