Glenn McGillivray is managing director of the Institute for Catastrophic Loss Reduction.
Like the 2011 Slave Lake wildfire, the 2013 flooding in southern Alberta and other costly, disruptive events that have affected Canadian communities in recent years, the Fort McMurray wildfire came as a surprise to everyone. This is largely because we do not have a very full and robust picture of natural catastrophe risk in Canada. So we continue to be surprised – and often unprepared.
But we needn't be.
In the (re)insurance industry, actuaries and modellers spend a lot of time considering plausible scenarios that, while low in probability, could have potentially devastating effects on a company. They then work to guard against those scenarios by ensuring that the premiums they charge are risk-adjusted, that the book of business is not overly concentrated in high-risk areas, that the company is well capitalized, that a proper program of reinsurance is firmly in place and so on.
The federal government, through the Office of the Superintendent of Financial Institutions, also requires insurance companies to conduct an annual stress test to ensure that they can survive a large shock or combinations of losses that could conspire to bring them down.
So Canadian property and casualty insurers are well prepared to deal with natural disasters. But imagine what can happen to companies in other industries that have yet to contemplate the impact of such events. The effects of large shock losses on business entities could range from physical damage to company assets, disruption to supply chains and distribution channels and loss of markets. Any one of these impacts can be fatal to a business enterprise that isn't prepared. Often, a large unforeseen shock loss from a natural disaster will nudge an already flagging company over the edge.
For the most part, it appears that insurers are among the few organizations doing stress-testing and planning for low-probability/high-impact natural disasters. A large number of players in other sectors do not model natural-disaster risk or conduct stress testing. Numerous entities fail to include natural disasters in their enterprise-wide risk-management frameworks (if they are doing such risk management at all) and many lack even basic business-continuity plans.
As for governments, these considerations often do not appear to even be on the radar.
But the wildfire in Fort McMurray should spur businesses and governments of all sizes and kinds and at all levels to do their homework to better understand what disaster scenarios are possible, what the impact can be and what can be done to guard against them.
At present, preliminary insured losses for the Fort Mac wildfire sit at $4.6-billion (this figure may be adjusted in coming months). This is from the destruction of about 10 per cent of a city of 88,000.
In the early days of the disaster, one investment bank estimated that if all of Fort McMurray was razed, insured losses would reach $9-billion. But simple arithmetic would indicate that this figure must be wrong if 10 per cent of the city is going to cost (re)insurers half that.
So what else don't we know?
There are several groups (including the Institute for Catastrophic Loss Reduction) that are diligently working behind the scenes to identify and prioritize various low-probability/high-impact scenarios and put credible numbers against them. Recently, some of this work has made it into mainstream media reports.
In May, the Fraser Basin Council released a study stating that a major flood of the Fraser River could result in up to $30-billion in damage to public and private buildings and infrastructure.
Swiss Re recently published a study stating that a 200-year flood in Canada (i.e. a flood with a 0.5-per-cent chance of happening in any given year) could trigger economic damage of $13.8-billion and insured damage of $5.7-billion.
And in June, an ICLR workshop featured a senior scientist from the Boston-based catastrophe-modelling firm AIR Worldwide who looked at several unlikely but plausible events that could drive large losses for Canadian insurers. The scenarios included a Category 4 hurricane scoring a direct hit on Halifax, shifting the 1998 Eastern Canada ice storm to hit present-day Toronto and a large hailstorm striking Calgary. The estimated loss figures were jarring: $32.5-billion, $26.2-billion and $13.6-billion, respectively.
This work is a helpful start. But what else aren't we considering?
Some answers can come from rerunning historic events to determine the impacts if they were to take place today in the same place and in the same way. How much damage would be caused if Canada's most powerful earthquake, the 1700 Cascadia temblor, and related tsunamis struck off the coast of British Columbia today? What about Canada's deadliest tornado, the 1912 Regina cyclone, or the country's most memorable storm, 1954's Hurricane Hazel?
The second kind of "what if" scenario involves the identification of wholly new losses to model. The trick here is to simulate events that, while plausible, are highly unlikely – the "Black Swan" phenomenon.
What if a strong tornado tore through downtown Toronto, or a magnitude 7 earthquake struck Montreal in the winter? What if cities such as Prince Albert, Sask., or Timmins, Ont., were ravaged by wildfire?
While identifying and planning for a series of low-probability/high-impact events will allow stakeholders to better prepare for the "big one," it ensures that they are also – by default – prepared for smaller, more frequent and less impactful events. The unthinkable only becomes thinkable after something happens.