partner login

BMS News

Insurance Industry Takeaways from 2014 Tropical Season


Sourc:e NOAA 2014 Atlantic Storm Tracks

Now that Arctic air has made its way across a good deal of the lower 48 states and snow had already covered 50% of the nation, the Atlantic hurricane season, and the threat of landfall along the U.S. coastline, is effectively over.
The season was forecast to be below the long-term average, and that is exactly what happened, for the most part. We will likely end up with a total of eight named storms, six of which became hurricanes. Two out of the six hurricanes made it to Category 3 or higher, with one, Gonzalo, reaching Category 4 – the strongest storm in the basin since the 2011 season. While we fell short of the average number of named storms (usually 10 per year), the number of hurricanes and major hurricanes is on par with what is expected in a typical season.

A better method of gauging seasonal activity, in my opinion, is the Accumulated Cyclone Energy index (“ACE”). By measuring the proportional energy of a named storm, it provides an accurate picture of how active a season might be and puts greater weight on the severity of the season as stronger storm accumulate a higher ACE than weaker storms. A typical season has an ACE value of around 104. This year the ACE score was 65.1, 62.5% of which was produced by the two major hurricanes of the season (Edouard and Gonzalo). This below-normal ACE activity follows the predictions made by most of the reliable agencies that produce seasonal hurricane forecasts. It should be noted, however, that the El Niño, which many of these forecasts believed would be the cause for a below-normal season, also failed to materialize. So, as discussed in previous postings, other factors that can lead to less storm activity, like African dust and cooler sea surface temperatures, actually did occur.

As I had previously blogged, Hurricane Arthur ended the absence of a hurricane stronger than a Category 1 hitting the U.S. coastline that had existed since 2008, and came close to ending the long-standing major hurricane drought, which has lasted 3,317 days and counting. However, many have already forgotten about Arthur due to its low impact. It is also worth noting that Florida alone has not had a hurricane of any intensity since October of 2005. This is simply an incredible statistic that, unfortunately, could negatively affect the Sunshine State if and when a hurricane hits in the future due to the fact that many may have become complacent regarding hurricane risk.

No other Atlantic hurricanes threatened the United States this season. However, there were indirect effects, such as high surf and rip currents from Bertha and Cristobal east of Cape Hatteras. Amazingly, at odds of 500-1, the small 20 square mile island of Bermuda was the unlikely recipient of two direct hits from named storms this season. Fay and Gonzalo both passed right over the island, causing power outages and some damage to buildings. Bermuda fared very well considering the impact from two systems less than a week apart, partially due to the strict building codes that have been followed and enforced, which is an important lesson for the insurance industry.

Another lesson from this event is that hurricane clustering is real, as was also the case during the 2004 hurricane season. At any time during a given season, prevailing wind currents that steer storms in a direction can get locked in place. This may cause storms to track repeatedly over or near the same area, triggering multiple storm losses in a given season. This storm clustering was also experienced in the East Pacific hurricane basin, which was exceptionally busy and produced a few hurricanes that impacted the Baja Peninsula and Mexico with either direct hits or leftover moisture. Three of these events, Norbert, Odile, and Polo, provide another great example of storm clustering. Storm clustering also occurred near the Hawaiian Islands and the south coast of Japan, which saw four typhoons this season.

Important insurance lessons also come from Hurricane Odile, a major hurricane that directly hit the resort city of Cabo San Lucas, Mexico. Although the overall wind swath and configuration of the Baja Peninsula seemed to minimize Odile’s impact, resulting in an industry event below $1B USD, there is a likelihood that the storm’s strength, with a preliminary pressure of 930 mb, will trigger the MultiCat 2012 cat bond, making it only the second cat bond to be triggered due to a tropical cyclone. This shows that alternative risk transfer products can work given the right insurance structure. The other insurance lesson learned from Odile is related to losses that go beyond basic building damage. There were numerous examples of looting in big box stores like Costco and Wal-Mart, as well as other small stores.

Walmart_Looting2  Walmart_Looting

Looting of Cabo San Lucas Wal-Mart after Hurricane Odile

The business interruption also was significant. I personally canceled a holiday in the beginning of October because of a lack of facilities, which adds up when thousands of others do the same, resulting in hundreds of millions of dollars in lost income to the tourism industry.

HotelDamage_Cabo  HyattPlace_Cabo (2)

Riu and Hyatt Hotel Damage from Hurricane Odile

While it’s too early to suggest what might occur during the 2015 Atlantic hurricane season, and although the current hurricane drought is exceptional, it is important to remember the long-term hurricane risk remains the same as it has been over the last century. BMS Analytics is here to help clients prepare for an active season by providing the newest tools and knowledge gained from past seasons, regardless of the overall low impact the past several seasons have had on the industry.

Arthur’s Amazing Facts Are a Positive for the Insurance Industry

Two weeks ago, Hurricane Arthur made landfall along the North Carolina Outer Banks. Arthur was the strongest hurricane to make U.S. landfall since Hurricane Ike in 2008 and was just 13 mph shy of ending the U.S. major hurricane drought. However, the overall impact of Hurricane Arthur was diminished due to the strongest winds being on the right side of the storm as it crossed eastern North Carolina, as discussed in my previous blog post, resulting in less overall damage. While damage was reported, and up to six feet of storm surge was observed in parts of the Outer Banks, most damage seemed to be flood-related and will be picked up by the NFIP, resulting in a loss level that falls below PCS CAT designation guidelines. This is notable for several reasons.

When reviewing the extensive PCS records of both U.S. hurricane landfall and hurricane loss, Hurricane Arthur is the second Category 2 hurricane to make landfall and not have a PCS designation. The only other storm in which this situation occurred was Hurricane Gerda, which made landfall in the extreme northeast portion of Maine in 1969, making the lack of designation understandable given the limited exposure across this region. However, according to Corelogic, there are an estimated 23,215 residential properties in Kill Devil Hills and Morehead City, NC where Arthur made landfall, with a total replacement cost of $4.7 billion. Based on Verisk Climate Respond weather data found in the BMS iVision Historical Events Library and using the unique PCS shapefile for Arthur, it is remarkable that a Category 2 hurricane in this area that had three-second wind gusts over 70 mph would not cause a PCS loss of at least $25 million. Particularly since there have been previous storms that have taken similar tracks and caused PCS-designated losses in the past.

BMS iVision with Arthur’s track and estimated three-second wind gust swath.

Although each named storm has special attributes that may cause insured loss, the general characteristics that drive loss are similar. However, as the image below illustrates, there were five hurricanes that occurred between 1955 and 2012 that tracked within 30 miles of Arthur’s path across North Carolina’s Outer Banks. These five storms all produced PCS losses, even though they had similar or weaker storm strengths than Arthur at landfall.

Five historical storms that have tracked within 30 miles of Arthur’s track and caused PCS loss.

More significantly, when looking at the past named storms from 1955 to 2012, 35 have caused PCS losses in North Carolina, with many of the named storms making impact at or below a Category 2, and several storms tracking hundreds of miles away from North Carolina, such as Hurricane Sandy (2012), which tracked 273 miles east of the Outer Banks. Click here for a linked table to these storms, which can be reviewed using NOAA’s Historical Hurricane Tracks tool. The image below provides a view of four of the named storms that caused PCS losses in North Carolina.

Tracks of four of the 35 historical storms that have caused PCS to North Carolina according to PCS data.

The examples above illustrate that North Carolina’s Outer Banks are no stranger to named storm activity, with the expected landfall return period for this area being five years, and major hurricane return period being 16 years, according to the National Hurricane Center. This has allowed the Outer Banks to better prepare for future named storm losses. The good news is that after years of storms, a Category 2 hurricane making U.S. landfall and having minimal impact demonstrates that insurance companies are becoming more risk-averse and policyholders are either constructing or reconstructing buildings at standards that reduce loss. One can only hope that future hurricanes making landfall along the U.S. coast will produce similar results.

The Right Side of a Storm

The insurance industry often focuses on media graphics that depict a storm’s path and the “cone of uncertainty,” but many of these graphics fail to explain the physical structure of a hurricane. The extent of hurricane damage doesn’t solely depend on the strength of the storm. It is also greatly influenced by the way the storm makes contact with land, and whether the left or right side of a hurricane strikes a given area.

The “right side of the storm” refers to the storm’s motion. For example, if the hurricane is moving to the west, the right side would be to the north of the storm; if the hurricane is moving to the north, the right side would be to the east of the storm. In the Northern Hemisphere, the strongest winds in a hurricane are generally found on the right side of the storm because the motion of the hurricane contributes to its swirling winds. Therefore, the right side of a hurricane packs more punch, since the wind speed and the hurricane’s speed of motion align. Conversely, on the left side, the hurricane’s speed of motion subtracts from the wind speed. The National Hurricane Center (“NHC”) forecasts take this asymmetry into account and often predict that the highest winds are generated on the right side of the storm.

The image above illustrates why the strongest winds in a hurricane are typically on the right side of the storm.

Hurricane Arthur is now less than 12 hours from impacting the North Carolina coastline, with a forecasted intensity of a strong Category 2 storm. Knowing the exact track of Arthur is critical to predicting the expected damage. If Arthur follows a more easterly track and skirts North Carolina’s Outer Banks, as suggested by the Geophysical Fluid Dynamics Laboratory (“GFDL”) model and current NHC forecast, it would mean the strongest winds (i.e., the right side) would remain away from the Outer Banks and offshore. However, forecast adjustments have been increasingly trending to the west, and with most U.S. models favoring a landfall near Morehead City, NC, the worst possible conditions would hit the Outer Banks as the storm tracks up Pamlico Sound.

Above is a view from BMS iVision, which, using model guidance from Verisk Respond, currently puts the right side of the storm and the strongest winds directly over the Outer Banks. This real-time wind forecasting information within iVision will enable clients to view the effect of Hurricane Arthur’s wind swath on their policy base, therefore providing a better estimate of exposed locations and possible losses. This westward track also increases concern for storm surge. The islands of the Outer Banks flood very easily, and the latest forecast by the NHC suggests up to three feet of water over US-64, which is one of two roads crossing the Outer Banks. However, Arthur’s forecasted approach along the North and South Carolina coastlines should limit the impact of a large storm surge.

While the Outer Banks is no stranger to hurricane-force winds, or even storms named Arthur (which occurred in both 1996 and 2002), this storm is forecasted to be one of the strongest to impact the area since Hurricane Emily in 1993. With an estimated return period of a hurricane passing within 50 miles of the Outer Banks occurring every five years, property has generally been upgraded to withstand such storms. However, the strongest winds staying to the right side of the current NHC track will determine the final outcome of damage and loss.

Tropical Update: Arthur

With a month of the Atlantic hurricane season in the books, one might think that the quiet Atlantic hurricane season is unusual. Historically, however, the year-to-date Atlantic hurricane season typically only experiences an Accumulated Cyclone Energy (ACE) index value of 1, based on the 1981 – 2010 climatology. Also, on average the first named storm typically does not form until the first week in July, with the first hurricane not showing up until mid-August. According to Roger Pielke Jr.’s normalized economic hurricane loss dataset, when looking at damage from tropical cyclones, historically only 2% of hurricane damage occurs in July, with 95% occurring in August and September. In fact, with the development of the first named tropical storm of the 2014 Atlantic hurricane season (Arthur) off the southeast coast of the U.S., the 2014 season is matching nicely with climatology, and by July 4 it should be ahead of climatology.

Earlier this spring, in our first look at the 2014 hurricane season, it was mentioned that not all El Niño seasons are the same. Even if an El Niño develops, it does not mean that the Atlantic hurricane season will have limited impact. In that post we highlighted past seasons, such as 2004, where El Niño had a high impact and we further detailed the importance the warmer- than-normal Sea Surface Temperatures (SST) off the East Coast could have on the upcoming season. Arthur is currently centered over these warmer-than-normal SSTs and is expected to strengthen into the first hurricane of the 2014 season.

Above is the National Hurricane Center (NHC) official track and intensity forecast, as of 11 AM EDT, showing Arthur tracking along the southeast coast of the U.S. over waters of at least 26 degrees Celsius. This water temperature is warm enough to support hurricane development. According to the NHC, Arthur is expected to just by pass the Outer Banks of North Carolina as a category 1 hurricane on Friday July 4th.

Another factor that will aid in hurricane development is the natural curve of the southeast coastline. Historically, the curve of the coastline has helped similar storms develop in this area, by providing a natural pressure/wind gradient that allows for counter-clockwise rotation. In 2004, Hurricane Alex battered the outer banks and strengthened in a 42-hour period from a minimal 35 kts tropical storm to a 85 kts hurricane, as it tapped into the warm waters of the Gulf Stream. Hurricane Alex produced light damage in the Outer Banks, primarily from flooding and high winds. Over 100 houses were damaged and damage totaled approximately $7.5 million (2004 USD) in economic loss.

As Arthur develops, an approaching trough of low pressure that is moving into the central U.S. will provide an atmospheric pattern conducive to low pressure development on the southeast side of the trough; this low pressure will allow for further intensification later this week. However, this approaching trough will not only keep the upper Midwest and parts of the East Coast cool for the July 4th holiday weekend, it will likely provide the steering flow to push Arthur off shore and provide minimal impact to insured property along the East Coast. This would be similar to the impact of Alex in 2004.

The greatest threat will be to the North Carolina Outer Banks on the 4th of July, as the storm tracks 50 – 100 miles east as a possible strong category 1 hurricane. It has been 1 year, 10 months and 1 day since the last hurricane hit the U.S. (Hurricane Isaac). With the understanding that Superstorm Sandy was officially downgraded miles off the NJ coastline, keep in mind that hurricane Sandy rapidly strengthened, due to a warm gulf stream and Arthur has access to similar warm waters to spur it on. It is these warmer-than-normal SSTs that need to be watched all season.