The Upside of Downtimes

Anniversaries have a way of sneaking up on even the best of us, but how many insurers know that 2015 marks 10 years since the U.S. had a Category 3-strength or higher hurricane make landfall on either the Atlantic or Gulf coasts?

That’s not to say there have not been damaging tropical storms. For instance, the Insurance Information Institute (I.I.I.) ranks 2012’s Superstorm Sandy third on the all-time list of costliest tropical storms in terms of cost to insurers, with nearly $19 billion in insured losses across a dozen states. Most claims professionals’ memories are still fresh from Hurricane Ike, too, which carved a path from the Gulf coast to Pennsylvania in 2008, resulting in more than $15 billion in insured losses.

But 2005’s Hurricane Wilma was the last time that a storm with winds topping 111 mph made landfall, which means insurers necessarily are focusing on other natural catastrophe trends and patterns while keeping a wary eye on the horizon.

What did insurers learn from 2014’s natural disasters? From the looks of it, extreme weather is poised to become the number one catastrophe headache for insurers, with different revelations emerging from two specific perils: hail and wildfires.

Revelation #1: Hail and Bad Dates

In 2014, severe thunderstorms—which include hail, tornadoes, and straight-line winds—accounted for the vast majority of natural catastrophe claims. I.I.I. and Munich Re reported that 62 of these events resulted in $12.3 billion in insured losses, making severe thunderstorms responsible for 80 percent of insured losses in this category. According to Verisk Analytics’ Property Claims Services (PCS) division, four of the top five catastrophe events were related to severe thunderstorms and accounted for nearly 850,000 claims (see sidebar, “Top Five Events”).

Insurers naturally are paying attention to the problems and challenges related to these claims. Specifically, a data quality and reporting problem has emerged that has far-reaching effects, especially as it relates to hail claims.

“When a tornado occurs and tears the roof off a home, it’s easy to define the date of loss because claims usually come in right away,” says Patrick Pollard, vice president of insurance solutions for Verisk Climate. “Hail claims, however, are different because some time may pass before evidence of damage appears. Additionally, many policies allow for claims to be made 12-18 months after an event occurs. Because of this, we are finding through our work with insurers and the analytics we compile that almost 25-30 percent of hail claims have an incorrect date of loss.”

Using totals for just the top four severe thunderstorm events of last year—each of which featured a hail component—means that up to 250,000 of those claims alone might be misclassified, which could have a significant impact on insurers of all sizes.

“It’s causing concern for several reasons,” says Pollard. “First of all, the designation between catastrophe and noncatastrophe claims is very important. Insurance companies report on whether claims are related to catastrophes or noncatastrophes, and if there is an increase in noncatastrophe weather claims, it’s going to get the attention of an insurer from an underwriting and risk management standpoint. But what might be happening, in reality, is a data quality problem related to the difference between the real and the reported date of the loss.

“On top of that, from a bottom line perspective, if these claims do not have the correct date of loss, they may not get assigned to the appropriate catastrophe reinsurance treaty in place, which means the insurer might not get the recovery it deserves,” continues Pollard. “Therefore, it’s imperative that the data quality is there so that insurers are assured that each and every claim associated with that catastrophe has the right date, since some of these treaties are very time sensitive.”

Another effect of bad dating? An insurer might pay a claim to which it’s not beholden.

“Insurers can inspect the property and find damage, but if they don’t get the date of the loss right, there’s a chance—however small—that they may not have been the ones insuring the property at the time of the damage,” says Pollard. “These are all reasons why date of loss is something that insurers are really trying to address and focus on.”

Revelation #2: Wildfires Expose Complacency Threat

On the other end of the spectrum, wildfires became an area of concern for insurers coming into 2014 because all signs pointed to a very active season. California, in particular, began its burn season earlier than usual, with dozens of wildfires appearing in January 2014, instead of during the typical May-December season.

“Most wildfire experts thought 2014 was going to be a bad year for several reasons,” says Tom Jeffery, senior hazard scientist at CoreLogic. “Even though California has had an ongoing drought for years, 2013 was almost a record-setting low year in terms of number of fires and area burned. Many of us assumed that because there were not any fuel-reducing fires during 2013, it meant that 2014 would be a very bad season.”

But then a strange thing happened: The wildfires never materialized. In fact, CoreLogic’s study, “Natural Hazard Risk Summary and Analysis,” noted that 2014 had the second-fewest fires and the lowest wildfire acreage burned over the past decade. What happened?

As it turns out, the response to the threat of wildfires has become the poster child for the power of mitigation.

“Everyone in the state was concerned that a small fire could quickly turn into a large conflagration, so officials would hit flare-ups with everything they had,” says Jeffery. “Helicopters, tanker planes, and large numbers of responders would stomp these fires out immediately before they got out of control. This overwhelming response really did a good job of preventing damage, and that extended throughout the year.”

Jeffery also points out the power of getting the word out to policyholders in an effort to combat complacency.

“When policyholders hear in the media that there are drought conditions and fire bans in place, they become more aware of what they are doing and take preventative measures, such as removing ‘fuel’ from around their homes so a fire cannot get rolling into a huge catastrophe,” he says. “We cannot prevent these disasters, but certainly we can plan for them by taking preventative measures like these.”

This kind of mitigation and response has to be continual, though. One or two years of complacency can reignite the risk, something about which insurers worry.

“The one key word I keep hearing over and over from insurers is ‘complacency,’” says Jeffery. “Everyone is concerned about it because it’s human nature to forget. Insurers always are looking for ways to keep the public and their customers interested in being aware of what’s going on.”

Insurers also are using data to zero in on disaster sites for better intelligence. Gone are the days of broadly named regions like Tornado Alley, for instance. Instead, insurers are pushing for pinpoint exactness of potential exposures.

“They are demanding granularity as they dig into data, so we are always looking for ways to base the picture of risk on smaller geography,” says Jeffery. “Whether it’s wildfires or tornadoes, you can look within a state to find specific hotspots, which in turn gives insurers a better way to manage risk.”

In the end, predicting weather is tricky and often surprising, as is evidenced by the recent Northeast blizzard that caused the region to brace for the worst. Nearly 8,000 flights were canceled, schools and businesses closed, and officials declared massive travel bans. However, the storm shifted and brought far less snow than predicted to New York City while pounding New England. With that in mind, can any diaster season really be deemed “surprising”?

“There is really no way to predict how many of these events are going to occur with complete accuracy,” says Jeffery. “The only thing that surprises me anymore is when people are surprised when a natural catastrophe takes place.”

Top Five Events in 2014

According to Verisk Analytics’ Property Claims Services division, here are the top five natural catastrophes of 2014 and the states affected by them.

May 18-23, 2014

Wind and thunderstorm event (flooding, hail, wind)

437,000 claims

$2.94 billion insured losses

States affected: Colo., Del., Iowa, Ill., Ind., Mont., N.Y., Ohio, Pa., S.C., Va.

Jan. 5-8, 2014

Winter storm event (freezing, ice, snow, wind)

150,375 claims

$1.67 billion insured losses

States affected: Ala., Ga., Ill., Ind., Ky., Md., Mich., Mo., Miss., N.C., N.J., N.Y., Ohio, Pa., S.C., Tenn., Va.

June 3-5, 2014

Wind and thunderstorm event (flooding, hail, tornadoes, wind)

115,300 claims

$1.26 billion insured losses

States affected: Ark., Iowa, Kan., Neb., Wyo.

Apr. 27-May 1, 2014

Wind and thunderstorm event (flooding, hail, tornadoes, wind)

122,350 claims

$1.12 billion insured losses

States affected: Ala., Ark., D.C., Del., Fla., Ga., Kan., Md., Mo., Miss., N.C., N.J., N.Y., Pa., Tenn., Va.

Apr. 2-4, 2014

Wind and thunderstorm event (flooding, hail, tornadoes, wind)

172,150 claims

$1.08 billion insured losses

States affected: Ill., Kan., Mo., Texas

Source: PCS