Industry impacts of the Fort McMurray wildfire

Unlike the 2013 flood in southern Alberta, the Fort McMurray wildfire is a heavily insured event. Considering that property insurance got its roots in fire (indeed it used to be widely known as ‘fire insurance’, still is in certain circles, and many insurers still have the word ‘Fire’ in their names) most policies will respond.

Wade Wilson talks on the phone by his burnt up truck at his home in Saprae Creek near Fort McMurray Alta., on Friday June 3, 2016. Members of the community are now being allowed back into their homes. THE CANADIAN PRESS/Jason Franson

Giving this fact and the sheer size and intensity of the wildfire, Fort McMurray is a big loss – even by international standards.

But how big?

How much will this wildfire cost Canadian insurers?

PCS Canada has issued preliminary loss estimates of $4.6 billion. The number could go up or down when insurers are re-surveyed – it’s a hard one to call and there is no point in guessing.

On one hand, large complex losses like this can mean that claims could go up (possibly by a significant margin) as insurers get a better handle on such ‘wildcards’ as additional living expense and business interruption claims, the number of damaged homes (we know how many homes were destroyed, but do not know how many have been damaged, including by smoke, and how bad), debris removal, soil remediation, and the short building season in the region.

On the other hand, there are ‘wildcards’ that could see losses go down. For instance, there is an unconfirmed story that there was a 30 percent vacancy rate in Fort McMurray at the time of the fire. This could mean that many homeowners will take the lesser cash value and move elsewhere rather than take full replacement value and rebuild in Fort McMurray. The difference could be meaningful to insurers.

Much will depend on how insurers reserved for their losses – on the low, mid or high end. Most companies in Canada tend to reserve quite conservatively (better to take down reserves later than be forced to put up more), so if the total loss ends up going down (from the $4.6 billion), it may not be by much.

Modelled losses vs. company surveys

One investment bank riled things up early on by reporting that if all of Fort McMurray was razed by fire, insured losses would total $9 billion. And while the bank did give a more reasonable loss range of $2.6- to $4.7 billion, the mainstream press unfortunately clamped onto the $9 billion like a pit bull on an old tire.

Releasing what the number would be if all Fort McMurray was lost, in my mind, borders on irresponsible. You would never hear a (re)insurance industry analyst or cat modeller make such a statement. Indeed in its lost estimate, AIR Worldwide also mentioned $9 billion, but did not position it as being the total if all Fort McMurray was lost.

Rather than report a worst-case number, it would have been more acceptable to simply report the Total Insured Value (TIV) in the area.

The bank’s release appears to be a sensationalistic attempt to gain media coverage by using ‘shock and awe.’ If that was the bank’s goal, it worked. But if the goal was to inform, the bank failed miserably.

There are two lessons here.

First is beware of modelled losses as they are nothing more than that – modelled. They are useful for the range they give (they provide stakeholders a flavour for what can be expected) but nothing compares to the well-executed use of industry claims surveys. These numbers come straight from the horse’s mouth (i.e. from those actually paying the losses) and more fully reflect the ‘ground truth.’

Second, communication about catastrophe losses should be left to (re)insurance analysts (either insurers, reinsurers or reinsurance brokers), companies that collect data for such entities (such as CatIQ or PCS) or cat modelling firms. Such work should not be done by investment banks or bond rating agencies.

In a separate release, the investment bank mentioned above issued an estimate of the potential impact of the wildfire on the regional and Canadian economies. When commenting on a natural disaster, investment banks should restrict their commentary to this space, leaving the analysis of (re)insured losses to those whose business it is to do so.

This same bank estimated insured losses from the 2013 Alberta flood at between $2.25- to $3.75 billion. The final came in at about $1.8 billion.

How does this loss compare to other Canadian catastrophes?

Though it is too early to say with great certainty, the event will likely be greater than the previous costliest and second costliest insured losses in Canadian history combined: the 2013 Alberta flood and the 1998 Eastern Canada ice storm, and then some.

As noted above, the $4.6 billion preliminary could go either way, up or down, depending on several ‘wildcards.’

We should know more at the end of June.

How does this loss compare to other wildfire losses worldwide?

ICLR has gone onto to the record as saying that Fort McMurray will go down as the costliest wildfire loss in world insurance history, the first entity to make this statement. As noted by the risk intelligence service Artemis, there have been only six wildfires in world insurance history that have triggered claims of USD 1 billion or more, and Fort McMurray is the largest of the half-dozen.

ICLR calls #FortMcMurray largest #wildfire loss in world insurance history, outstripping 1991 Oakland Hills event (USD 2.96b 2015 dollars).

— Glenn McGillivray (@ICLRCanada) May 19, 2016

The size of this event and the amount of insured damage caused appears certain to put it up there with some of the large flood and moderate hurricane and earthquake losses in the international reinsurance industry’s top 50 costliest natural catastrophes, and possibly top 40.

ICLR: #FortMcMurray #wildfire set to be included in global (re)insurance industry's Top 50 disaster losses of all time and possibly Top 40.

— Glenn McGillivray (@ICLRCanada) June 5, 2016

At present, when converted to U.S. dollars, the event just makes it into Swiss Re’s global top 40 (1970 to present). If the $4.6 billion is later adjusted down, it will fall out of the top 40, but easily make the top 50.

We are also the first to make this call.

Is this an earnings event or a capital event?

In the world of insurance, an ‘earnings event’ is defined as a net insured loss (i.e. an insurers retained loss after reinsurance) that is small enough that it can be paid for out of the company’s near-term earnings. Such an event, therefore, can affect the immediate profitability of a company, but generally has no long-term impacts.

A ‘capital event’, on the other hand, is a net insured loss that is big enough that the company must dig into its surplus capital. Depending on the size of a company’s retentions, a capital event can have long term impacts on a carrier and, indeed, in some cases can be fatal.

With the typical conservative reinsurance practices of the Canadian insurance industry, and with risk spread over a large number of companies, it is generally agreed that the Fort McMurray wildfire is an earnings event and not a capital event (though there will be isolated exceptions).

Also a reinsurance event

Whenever there is a large insured loss event in Canada, the rule of thumb is that reinsurers tend to pick up roughly two-thirds of the loss. The Fort McMurray wildfire appears to be no exception.

According to official releases from some companies and stories surfacing out of the reinsurance sector, many primary companies have ‘blown through’ their cat reinsurance treaties. This means that they have exhausted their reinsurance coverage and must dig into earnings (and, in some cases, surplus capital).

According to several sources, quotes for reinstatement covers have been “flying out the door” as companies seek to restore their reinsurance programs as they roll into summer storm season. And as one reinsurance CEO put it: “They won’t be paying what they paid in January.”

Managing accumulations

The Fort McMurray event illustrates quite perfectly the importance of insurers properly monitoring and managing their concentrations and of reinsurers managing their accumulations. An over concentration of risk in one city, town or even neighbourhood can be very costly for an insurer, and sometimes even fatal.

The May 2011 wildfire in Slave Lake, Alberta triggered $775 million in clams (IBC Facts Book 2015) from the loss of one-third of a town of less than 7,000 people. The industry is currently sitting at $4.6 billion for the loss of 10 percent of a city of 88,000.

Canadian (re)insurers have to ask themselves: What if 20 percent of Fort McMurray was lost, or 30 or 50?

As horrific and heartbreaking this event is, insurers dodged a major bullet this time around.

Had it been worse, the industry would be looking at a significant capital event and the potential of some insurers getting into trouble.