Expect the unexpected

During the last two decades there has been an alarming increase in economic and insured losses stemming from natural disasters (See Figure 1). This has been a global phenomenon dating back to Hurricane Andrew (1992, United States) and to some implies an increase in the occurrence of extreme events.

Figure 1

Annual insured losses of more than $20 billion, $35 billion and upwards to $70 billion, adjusted to 2012 U.S. dollars, have occurred within just the last decade alone and have left many insurers wondering about their catastrophe management strategy and technology. Extensive economic losses caused by major events such as earthquakes in Chile, Japan, New Zealand and Haiti, the tsunami in Japan, hurricanes and tornadoes in the U.S., wind storms and floods in Europe, along with flooding in Thailand and Australia, have exceeded our expectations. Even those events previously thought of as less severe (e.g., tornadoes) have surpassed our expectations.

Many insurance professionals have attributed the increase in losses to a “new norm” of increased natural hazard activity and severity.

Canada is one of the few developed countries exposed to many different severe natural hazard risks, including avalanche, freeze, drought, flood, geomagnetic storm, hurricane, wind storm, surge, severe convective storms, tornado, wildfire, earthquake, landslide, tsunami and volcano. Over the last 10 years, Canada has suffered through many natural catastrophes, but none of them approach the severity of what has been experienced elsewhere. Most of the larger occurrence losses in Canada have come from the perils of drought or flood; still none of these events have exceeded $1 billion (CAD).

Going further back over the last two decades, Canada’s largest – and the only event with losses exceeding $1 billion – was due to the 1998 winter storm. In today’s dollars (GDP-adjusted) this event caused over $3 billion in insured and $6 billion in economic damages.

Over the last two decades of global losses, even Canada’s largest loss event would not pierce the top 25 global natural disaster losses (See Figure 2). Yet when looking at the developed nations’ natural catastrophe losses in percentage of industry resources, the 1998 winter storm would rank as the 7th largest event (See Figure 3). Still, even accounting for industry resources on a global scale, Canada has been relatively quiet during the last decade. Does this mean that Canada is exempt from this new norm or has it simply been lucky?

Figure 2

Figure 3

OUTSTANDING QUESTIONS

The first question that must be answered is what is behind the global increased losses: Is it a change in hazard, exposure or something else? There remains some debate on whether or not some of the atmospheric hazards have or will be exaggerated by climate change. What we do know is that global temperature is increasing and sea level is rising, both of which create additional risk.

Although it is not possible to attribute any specific weather event to climate change, we cannot ignore the strong possibility that climate change may have already had an impact on flood, drought and wind – and that this influence will grow in future. Similarly, despite the recent string of damaging earthquakes affecting population centres, there is no strong evidence of an increase in earthquake activity worldwide.

The main reason why these losses have exceeded our expectation is due to our own success as a society. Our wealth and the value of our homes, despite the recent housing downturn, has doubled over the last two decades, notes information from the National Association of Realtors, Canadian Real Estate Association and Haver Analytics. Our infrastructure has become more complex, our urban areas more congested, and our buildings (including homes) have gotten larger. We have become more interconnected and dependent on technology as a society.

The winter storm of 1998 was one of the first modern disasters to demonstrate how vulnerable our technology and infrastructure is to natural disasters, and how costly in terms of time and money it is to replace. Since then, the recent events of the last decade have emphasized that our technology is very vulnerable while our lives and businesses are more dependent on technology than ever before.

Today’s technology (e.g.; computerized manufacturing, robotics, automated services, etc.) is more vulnerable, harder to replace or repair and, as such, can lead to massive costs (economic/ insured losses) before operations can be up and running again.

The world is getting smaller. Our businesses are a mosaic of other business intertwined as suppliers and consumers (supply chain). Many of the same business concepts that have made the world more competitive have increased our risk. We have outsourced much of our operations to areas outside of our home country in areas prone to natural disasters.

Although far apart, many businesses are symbiotic and can experience extreme disruption if any one or more of the links in a supply chain is disrupted. As a result, business interruption and contingent business interruption (CBI) claims are getting larger.

MODEL COMPONENTS

Since Hurricane Andrew, the insurance industry has devoted a tremendous amount of time and resources to developing catastrophe-modelling technology. These models combine science, engineering, statistics and financial components into a comprehensive model that estimates natural hazard losses on a probabilistic basis.

Over the years, these models have become more sophisticated in terms of information required for the specific risks to be modelled, and in terms of the scale and scope of the science and engineering information within the models.

However, despite our best efforts, modelling can be misleading. Sometimes the science can be wrong (e.g., the underestimation of maximum possible magnitudes in the Tohoku region by Japanese hazard maps), not fully recognized (e.g., widespread occurrence of liquefaction during the Christchurch earthquake), or missing from our models (e.g., tsunamis). Generally, our models perform well with regards to direct physical building damage, but have been less precise when modelling the non-structural components of risk such as operations and business interruption.

How does this impact Canada? Canada is no different than many of the other developed nations. Canada’s cities, its infrastructure and lifestyles have become more complex and expensive. As in other developed nations, Canada is also very dependent on technology. Canada and some of its major population centres are also exposed to a number of severe natural events, with British Columbia earthquake having the largest potential.

With respect to earthquake, B.C. is affected by a significant subduction zone and substantial crustal seismicity. As in the case of Japan, the hazard estimation could be understated. The B.C. coast could also be impacted by a tsunami, and much of the coastal or riverine areas are exposed to liquefaction. It is not unusual for large aftershocks to occur, some of which can be as devastating as the initial tremor.

There is also a chance that some hazards are not fully modelled, if mo delled at all. If a large B.C. earthquake were to happen today, we should not be surprised if the event damages are larger than we expect or are prepared for. Canada would not be exempt from many of the same conditions, increased wealth, supply chain (CBI), complexity of risk, and technology changes that have added significantly to the size of loss experienced by others.

Canada should learn from what the recent events have shown us and should be prepared for what can no longer be considered unexpected. Canada has been fortunate in recent years, much like New Zealand had been before 2011. A natural catastrophe may be around the corner, but are we prepared for it?

On February 7, 2013, Canadian Underwriter and Swiss Re hosted a free live webinar presentation, The Changing Canadian Risk Landscape, featuring Andrew Castaldi. The approximately 1-hour webinar is archived for free on-demand viewing at http://bit.ly/canadianriskwebinar.

Figure 4