When the western world awoke on Boxing Day, December 26, 2004, it was to sketchy reports of an earthquake off the shore of Indonesia which had apparently caused huge waves to come crashing onto land. But it would be many hours, even days, until the full magnitude of the disaster was understood. Even as this article is being written, thousands of people remain missing and the economic impact of the catastrophe is still being considered.
What is known is at about 7 a.m. local time, a magnitude 9.0 earthquake shook the floor of the Indian Ocean about 100 miles west of the Indonesian island of Sumatra. This was followed by a series of aftershocks, some registering as high as 7.2 on the Richter scale. Tidal waves produced by the earthquake took less than 30 minutes to reach the shores of Aceh province in Indonesia and the Phuket region of Thailand. In total, a dozen nations on three continents were hit by the massive waves, including Thailand, Indonesia, India, Maldives, Malaysia, Bangladesh, Sri Lanka and Somalia. In human terms, the extent of the disaster is barely comprehensible. Upwards of 155,000 people are dead, thousands still missing, and millions homeless.
Economic damages will be massive, but for insurers, the total claims will pale in comparison. In North America, it is not unusual for a natural disaster to produce insured losses roughly half of economic losses. This will hardly be the case in South Asia, India and Africa, where insurance penetration is low. Risk modeling firm RMS has put the insured loss estimate at US$4 billion.
Analysts note the level of the disaster for insurers will be more than manageable and no impact is expected on insurers’ financial strength ratings. Nonetheless, the tsunami poses many challenging questions for the industry, not least of which as it comes at the tail end of the worst year in recorded history for insured losses from natural disasters.
Chief among the questions posed by the Asian tsunami disaster is whether such disasters can be mitigated or are we doomed to remain at the mercy of Mother Nature’s wrath?
The Indian Ocean tsunami is a rarity – 80% of all tsunami activity takes place in the Pacific Ocean, with only 2% in the Indian Ocean. The lack of warning systems in place in the Indian Ocean, as compared with the well-known Pacific Ocean Tsunami Warning System (POTWS), becomes somewhat understandable in this light.
But tsunamis in and of themselves are not necessarily rare occurrences, partly because they originate from a variety of sources. The most common causes include earthquake, landslides either above or below the water, or volcanic explosion, explains David Smith, director of risk modeling firm Eqecat.
The question has been raised since December 26 of the potential for a tsunami to strike North America, with two infamous examples given in the Canadian context. In November, 1929, a rare Atlantic Ocean tsunami hit the Grand Banks of Newfoundland, resulting from a 7.2 magnitude earthquake under the ocean which caused a submarine landslide. Taking just 2.5 hours to reach the coast, the 10-meter tall waves were blamed for as many as 30 deaths.
Then, in 1964, a 9.2 magnitude earthquake in Alaska sent waves crashing into the B.C. coast at Port Alberni, killing more than 100 people as it moved down the U.S. West coast. About 260 homes were wiped out in Port Alberni, with economic damage there estimated at $10 million.
Canada’s West coast has a much higher risk of tsunami activity, explains David McCormack, head of the Canadian National Seismograph Network at Natural Resources Canada. The Port Alberni event was “slightly peculiar” in that two successive wave cycles hit the coast concurrently, he notes, but the 1929 East coast event was very rare, probably representing a less than one-in-1,000-year event.
Nonetheless, McCormack adds, “what we have seen in the Indian Ocean, although infrequent, shows that in principle tsunamis can happen anywhere, that they could potentially effect any of the ocean basins.”
West coast tsunamis are inevitable in some respects. “There is no question they will happen in the future; they’ve happened in the past,” notes Paul Kovacs, executive director of the Institute for Catastrophic Loss Reduction (ICLR), an insurance industry sponsored disaster research organization.
Much of this risk comes from what is known as the “Cascadia subduction zone”, where the eastward moving Juan de Fuca plate meets the westward moving North American plate off the coast of B.C. and northwestern U.S. The Juan de Fuca plate is continually trying to “slip” beneath the North American plate, and every 500 years or so it succeeds, with catastrophic results. One such slippage event sent waves all the way to Japan, notes David Wilmot, senior vice president, Canadian branch of the Toa Reinsurance Co. of America. He adds that reinsurers covering B.C. risks will certainly have modeled this kind of risk, but the province’s public auto insurer, ICBC, has also done its own modeling, noting the potential to wipe out yards of imported cars at harbors. “They were thinking of it as potentially their greatest exposure.”
“The potential for a large tsunami occurrence is low, but it’s a high consequence event,” comments Alan Pang, manager, p&c underwriting consulting services, CGI Insurance Business Services. The main lesson taken from the Boxing Day tsunami event is that even events as rare as one-in-1,000 or one-in-500 years can create a devastating impact lasting potentially for decades.
The Asian event has also turned insurers’ eyes to the U.S., which has the world’s highest level of insurance penetration. Coming on the heels of four major Atlantic hurricanes in 2004, insurers are more focused than ever on understanding “worst case scenario” risks.
The highest areas of tsunami risk, explains Smith and Eqecat senior vice president Tom Larsen, include: Alaska, as seen in the 1964 earthquake tsunami; Seattle’s Puget Sound, where landslide risk is high; the Oregon coast, where paleo studies have revealed five tsunamis in the past 2,000 years; and Puerto Rico.
The 1964 event caused more than US$84 million damage to several states including Washington, Oregon and California, on top of the damage sustained in B.C., notes Loretta Worters, vice president of communications for the U.S. Insurance Information Institute (III). The III says the value of property along the U.S. West coast is probably close to that of the East and Gulf coasts, about US$6.5 trillion. But the tsunami risk is clearly much higher on the West coast, where a damaging tsunami occurs every 18 years on average, and specifically a Cascadia zone earthquake, Worters says. “A Cascadia earthquake would be very large, would result in a tsunami, and would only give a few minutes of warning time to the residents along the Pacific Northwest coastline, in many cases not enough time to allow for evacuation, especially during vacation season.”
She also highlights the risk in Hawaii, pointing to the 1946 tsunami which brought 20-32 foot waves to Hilo, Hawaii, and caused 159 deaths. In fact, Worters notes, “since 1945, more people have been killed as a result of tsunamis than as a direct result of an earthquake’s ground-shaking” in the U.S.
A tsunami anywhere in North America is likely to trigger extensive insurance exposure, despite some key exclusions in place in most homeowners’ policies. “All lines of the p&c business could sustain significant losses,” Pang says of the exposure associated with a tsunami off the B.C. coast. This includes the more easily quantifiable direct property damage, but also secondary losses, for example business systems shut down, including technology downtime. Following the Asian tsunami, it was noted that a significant portion of losses would be as a result of business interruption policies triggered by the closure of hotels and resorts in the Maldives and Phuket islan
Flooding risks are not generally covered in homeowners’ policies, with the exception of sewer back-up, but as was seen in last year’s Peterborough, Ontario floods, it is a challenge to separate out what damage has been caused by flooding and what has been caused by sewer back-up, confirms Walter Panasdir, manager, catastrophe service for Cunningham Lindsey Canada Ltd. He adds that in B.C. earthquake risk is not generally covered in homeowners’ policies, although it could be part of commercial policies. Panasdir notes that he has seen one Canadian policy which specifically excluded tsunami risk.
In the U.S., Panasdir adds, much of the flood risk from a tsunami would likely come under the National Flood Insurance Program. Worters confirms U.S. homeowners’ policies do not cover flood risk as a result of tsunami, nor damage caused by landslides, while earthquake is generally purchased on a stand-alone basis. Even when earthquake is merely the “proximate cause” rather than the direct cause of damage, because it is the original trigger of loss, associated losses would generally not be covered.
Lest the industry think that these exclusions would result in a local tsunami event leading to low insured losses, it is on commercial policies insurers would face significant payouts. Worters notes that commercial property insurance “would no doubt be triggered” by a tsunami, including business interruption. Also, insurance for infrastructure, port facilities, marine interests including ships, cargo, oil platforms and other offshore facilities, and of course, potentially life insurance, could all come into play following a tsunami, she adds.
Of course, the new reality of insurance is that local interests are no longer the only concern. Insurance has become globalized as much as any other industry and many carriers, specifically reinsurers in North America are part of global operations. A tsunami event anywhere in the world will impact this global market. At the same time, notes Smith, insurers need to understand their clients who may be impacted by disasters in other parts of the world. One of the most significant insured losses coming out of the Asian tsunami will likely be the destruction of the Lafarge cement operation in Aceh. And the shutdown of a “feeder plant” in a developing country such as Indonesia has the potential to send “echoes of the waves” across the world, possibly shutting down operations of domestic manufacturers and processors, Smith notes.
MITIGATION AND EDUCATION
Insurers may now face serious questions about what risks they are willing to assume in certain countries based on levels of disaster preparedness in place, and this includes domestically. It is likely that countries will make good on their commitment to extending the POTWS into the Indian Ocean, although the relative rarity of another tsunami will certainly bring into question whether this is the best use of limited funds in developing nations.
But here in Canada, the government has also committed significant funding to enhancing warning systems. The issue, says McCormack, is the “bottom level” of warnings, i.e. making sure that people get the warning, understand what it means and know how to respond to it.
Kovacs says the ICLR wants government to look beyond tsunamis to a “multi-hazard” warning system which would give different degrees of warning for events including hurricanes, tornadoes, hail storms and more. But, he adds, communication is really the key. “Most countries are not doing this very well.” ICLR research following the August 2003 blackout in Ontario found that even after that event many people were not taking steps to ensure emergency preparedness.
Kovacs adds that insurers have an important role to play in advancing this agenda with government. As the people who respond to catastrophes, “insurers should choose to influence this process, to be outspoken about this process”, he says. This means not only investing in research, but also pushing for education and dialogue around natural hazards and preparedness. The industry also needs to promote better funding arrangements at the national level to ensure that when ideas are brought forward they can be put into action, rather than being mired in red-tape.
Wilmot adds that he fears there will be a “knee-jerk reaction” to the tragedy of the Asian tsunami, which will see all the focus on tsunami threats rather than addressing a wider array of potentially more pressing disaster concerns in Canada. For example, he notes that in B.C. the more probable near-term flooding risk comes from the Fraser River, rather than from a tsunami.
And, he adds, while warning systems and other mitigation techniques are important, the first step needed to reduce the likelihood of people being impacted by disasters is to stop building in high-risk areas. “What is changing is that there are now six billion of us [in the world]. We’ve used up all of the easy-to-use and safe-to-use property. Now we’re moving to the not-so-easy-to-use and not-so-safe-to-use property.” This was seen most recently with the California and B.C. landslides, which impacted vulnerable homes built into the side of hills prone to such disasters.
Worters agrees with this perspective. “Public education, awareness, and a warning system can make a difference in community disaster resistance, but the only way to really mitigate the risk is not to build in these areas.”
Pang also notes the need for stronger infrastructure, specifically storm sewers, to go along with better land use planning. And, he adds, insurers need to be at the forefront pushing for more relevant, applied natural hazard research. At the insurer level, this means better modeling and use of geographical information systems (GIS) to understand concentration of exposure in the face of what appear to be ever more common natural disasters. “Mother Nature is very mad. Seismic activity and extreme weather events seem to be happening with extreme frequency. We must use more accurate information in the understanding and management of these kinds of events.”