Canadian Underwriter

Slow Burn

September 1, 2003   by Glenn McGillivray

Print this page Share

At the writing of this piece in early September, it was beginning to look possible that the forest fires burning in British Columbia could usurp the 1991 Calgary hailstorm as the second-highest series of insured loss events in Canadian history (according to the Insurance Bureau of Canada’s (IBC) “2002 Facts Book”, the 1991 event triggered insured losses of $411.8 million, in 2002 dollars).

Neither the fires nor the hailstorm come close to the top “loss” spot, taken by the ice storm which hammered eastern Ontario, southern Quebec and Atlantic Canada in early 1998 – an event which triggered total insured losses of $1.82 billion (2002 dollars). But, while the ice storm and the fires raging in the interior of B.C. have almost nothing in common from a causal perspective (it is difficult to come up with two elements more characteristically different than fire and ice), the two do share commonality. Once the B.C. crisis is over, both series of catastrophes will have played out over an extended period of time (with the ice storm, freezing rain fell off and on for close to one week, and power remained off in parts of Quebec for more than one month. With the B.C. fires, many have been burning for more than three, four and even five weeks). And, historically, ice storms and forest fires have not caused significant insured losses in this country. During Ice storm 98, some areas where hit with 80 hours of freezing rain -nearly double the normal annual total.

Common threat

While forest fires are extremely common (according to the Ministry of Natural Resources, an average of 10,000 fires burn 2.5 million hectares of timber in Canada each year) they have tended to occur in either uninhabited or, at the very least, sparsely inhabited areas. According to Paul Kovacs, executive director of the Institute for Catastrophic Loss Reduction (ICLR), and author of “Wildfires and Insurance” (December 2001): “The Insurance Bureau of Canada has maintained a…catastrophic loss database for Canadian insurers since the early 1980s. Over this period, there were thousands of fires in the country, but there was not a major wildfire loss for the insurance industry. The industry has not yet experienced a Canadian wildfire loss in excess of $10 million.”

In fact, blazes in unpopulated areas are so common in Canada that, often, they are left to burn because no people, property or marketable timber is at risk. Indeed, of the hundreds of fires burning in B.C. up to the writing of this article, only a relatively small number – about 18 – have directly threatened homes or communities, triggering the evacuation of close to 50,000 people. But it is these 18 that are making all the difference.

Loss significance

Why will this loss will be significant? The impact of any loss event – natural or man-made – largely depends on the severity, frequency or duration, and locations of the occurrences. When placing these three measures up against the forest fires in B.C., severity has been high, and the duration long.

From a severity standpoint, more than 2,300 separate fires have been recorded in the province since the start of B.C.’s fiscal year (April 1, 2003 to March 31, 2004). As of September 4, the B.C. Forest Protection Branch reported that 691 fires were burning in the province and, to that date, more than 235,000 hectares had burned. By September 7, the number of reported fires grew to 763 with more than 243,000 hectares burned (according to 2002 Forest Protection Branch data, 1,729 fires were recorded in B.C. between January 1 and October 31 of last year, just over half of the annual average of 3,000 fires. Also, by the end of October 2002, 8,184 hectares had burned, much lower than the ten-year average of 23,058 hectares.) Thus, the number of hectares burned in B.C. so far this year has exceeded the decade average by more than tenfold. And with more than 760 fires still burning (as at September 7), and close to two months left to go in the high forest fire season, the number of hectares burned could no-doubt rise dramatically.

With regards to duration, at least one of the big fires was discovered in late July, and was still burning in early September. The blaze near McLure (located about 40 kilometers north of Kamloops) was first discovered on July 30. On September 7, though 95% contained, this fire was still burning. The McLure fire moved extremely quickly, growing from an estimated two square kilometers at 4 p.m. PST July 30, to 25 square kilometers just six hours later. Several other big blazes were discovered in early to mid-August, and were still burning at the writing of this piece.

Location impact

Looking at the overall picture: when the province first announced a state of emergency on August 2, 353 fires where still burning. When the state of emergency was extended on August 29, the number of fires had risen dramatically to 764.

But, while severity and duration are key to determining the impact of loss event(s) – with the forest fires being no exception – it is the location of the fires that is triggering much concern this time around. Having a fire threaten entire communities such as Kamloops, with a population of more than 86,000, and Kelowna with a population of close to 100,000, is a fairly new phenomenon in this country.

In the Swiss Re Canada publication “Inside an Ice Storm”, published in February 1998, the report stated “…it was evident quite early on that Ice Storm 98 would go down as the costliest natural catastrophe in Canadian history. This can be proven by looking at the amount of time it took to tally claims for the Calgary hailstorm. It was close to two years before the industry recorded $342 million in insured losses [1998 dollars] for the event while it took five weeks to tally $602 million in claims for the ice storm.”

While the current situation in British Columbia is far less clear, the loss could exceed that caused by the Calgary hailstorm partly due to the fact that a significant number of pricey homes and outbuildings have already burned (more than 240 houses/buildings were charred in Kelowna late in August), and while even more homes remain at risk. In the beginning of September, The Globe and Mail quoted Vern Huculak, president of Demara Insurance Brokers in Kelowna, as saying that claims will average about $750,000 a house all-in. The same article quoted Ken Snell, a fire investigator doing property examinations for a number of insurance companies, saying he has looked at some houses that were worth more than $1 million. In the Louis Creek area of B.C., a reported 72 homes and nine businesses, including a Tolko Industries sawmill were destroyed earlier in August.

Identifying cost

Because the fires are still raging, with many proving to be unpredictable, it is not possible to provide a complete inventory of structures destroyed or damaged to date. Suffice to say that more may be lost to the quickly-spreading blazes.

In reality, however, counting burned structures in a catastrophe such as this is the relatively easy part for insurers. The overall event will be made more costly (and complex) due to the combination of structures damaged or lost to fire. Other loss considerations include:

Any damage caused by smoke, which at this stage is a great unknown and;

The tally of all additional living expenses (ALE) for evacuated persons and those with damaged or destroyed homes.

If the total of all insured damages does approach or exceed $400 million, it would add over 1.5 percentage points to insurers’ combined ratio for this year – although a good proportion of this will, naturally, be borne by reinsurers.

Loss exposure

In recent years, insurers and reinsurers have been hit by a new phenomenon coming in the form of billion-dollar-plus insured natural loss events that fall outside the realm of earthquake, hurricane and flood. Not all that long ago, such events as billion-dollar hailstorms, tornadoes and ice storms were thought to be highly unlikely, if not downright impossible. But in the last few years alone, the roll call of such losses has proven na
y-sayers wrong.

The primary reasons for this paradigm shift (aside from the issue of climate change) comes from two somewhat inter-related sources: the building of communities near danger-prone areas, and an increased concentration in insured values (the two are inter-related in as much as both can be the result of an overall increase in wealth). There is an irony in this world in that many people live in danger-prone areas (on flood plains, near coastal areas or the base of mountains prone to mudslides) because they cannot afford to live anywhere else. Yet, many others choose to live in danger-prone areas (such as on wooded hills with sweeping vistas, on the bases of snowy mountains in ski chalets, or in ocean-side beach houses), because they can afford it.

In his paper “Wildfires and Insurance”, Kovacs makes the point that “more than ever people are moving into remote areas with the desire to ‘get back to nature’. Typically, homes built in these areas put more emphasis on ensuring a spectacular view and less on the dangers around them.”

Disaster mitigation

As with the many natural catastrophes that have preceded the B.C. wildfires, society learns from each event how better to protect people and property from natural catastrophic loss.

In the case of forest fire, for example, consideration of more and larger treeless green zones to serve as fire-breaks near heavily forested communities may be needed. More consideration may be put into the overall management of forestlands, including increasing prescribed burns and controlling the amount of forest litter that is accumulated over time. Similarly, property owners located near such forestlands may increase their vigilance against wildfires by ensuring, for example, that trees have been sufficiently cut back from structures and that properties are clear of forest litter and other potential forms of fuel. The insurance and reinsurance industries can be important contributors to this discussion. In the meantime, and to the extent that the risk cannot be prevented altogether, the industries will once again be there to distribute and finance the risk for the clients they serve.