Tag Archives: tornado

One Year Later: What We Learned from the Moore Tornadoes

This week marks the one-year anniversary of the severe weather outbreak that brought high winds, hail, and tornadoes to half of all U.S. states. The most damaging event in the outbreak was the Moore, Oklahoma tornado of May 20, 2013. Rated at the maximum intensity of EF5, it had maximum sustained wind speeds of up to 210 mph and was the most deadly and damaging tornado of the year for both Oklahoma and the U.S., causing roughly $2 billion in insured losses.

As we reflect upon the events that have taken place in Moore, the following can be discerned:

  • Understanding severe weather risks is key: According to the RMS U.S. Severe Convective Storm Model in RiskLink 13.1, the annual likelihood of a severe weather event causing at least $1 billion in insured losses in the U.S. is 92 percent, meaning it is almost certain to occur each year. For reference, from a loss perspective, the $2 billion 2013 Moore tornado loss represented a 1-in-50-year event in Oklahoma, or an event with a 2 percent chance of occurring in a given year. Similarly, a 1-in-100-year event, or an event with a 1 percent chance of occurring in a given year, would cause $4 billion or more in insured losses for Oklahoma. Events in excess of the 1-in-100-year return period would be driven by large, destructive tornadoes hitting more concentrated urban environments, such as a direct hit on Oklahoma City. Probabilistic severe storm models provide more perspective on these types of risks, and can better prepare the industry for the “big ones.”
  • What grabs the headlines doesn’t cause the most damage: Although tornadoes get all the news coverage and are often catastrophic, hail drives roughly 60 percent of the average annual loss in convective storms. This is mainly driven by the much higher frequency of hailstorms compared to tornadoes. Hailstorms also have a much larger average footprint size.
  • Tornado Alley isn’t the only risky place: Tornado Alley drives roughly 32 percent of the average annual loss for severe convective storms in the U.S., while the Upper Midwest drives 24 percent, Texas drives 16 percent, and the Southeast drives 12 percent. Buildings in affected areas need continued upgrades: For example, the Moore city council approved 12 changes to the residential building code after the Moore tornado, including mandates for continuous plywood bracing and wind-resistant garages (often the first point of failure during weak to moderate winds).

While we can never predict exactly when severe weather will occur, it’s imperative for communities, businesses, and individuals to understand its potential impact. Doing so will help people and industries exposed to severe weather be better prepared for the next big event.

Are you located in one of the regions affected by last May’s outbreak, or in another risk-prone area? Have you been affected by any recent severe weather events? If so, what did you learn, and what changes were made in your region to safeguard the community, businesses, and homes? Please share your experience in the comment section.

Jeff Waters also contributed to this post.

Severe Thunderstorm Risk: What You Don’t Know Can Hurt You

Are you using experience based rating to underwrite severe thunderstorm risk?

Many use this approach in North America, but if you do, you could be missing out on the full loss picture for this complex peril. Though tornado and hail are responsible for a major part of the annual insured loss, straight-line winds and lightning can also contribute to a material portion.Annual Insured U.S. Thunderstorm Losses by Sub-Peril

More importantly, recent trends in industry claims practices, event severity, and exposure concentration have indicated that the risk landscape is changing, suggesting that past hazard and loss patterns may not be reflective of those in the future.

Let’s take a close look at these trends.

From a Claims Perspective

Claims have been increasingly inflated and more severe in recent years, particularly in high-risk areas. RMS analysis of over $5 billion in new claims data has shown that the average size of a residential claim has increased over 9% per year from 1998 to 2012. Commercial claims have also increased around 9% per annum, while automobile claims have increased by 2%. These increases may not be captured solely by analyzing past hazard and loss patterns, hindering underwriters’ efforts to develop effective pricing practices.

From a Hazard Perspective

Major severe thunderstorm events have recently caused untold damage and loss, well beyond what was estimated using historical records. Between 2008 and 2013, the U.S. experienced over $80 billion in insured losses from thunderstorm related hazards, and two of these events each generated more than $7 billion in insured losses.

Similarly, events such as the 2010 Phoenix, AZ hailstorm, the 2011 Tuscaloosa, AL outbreak, and the 2013 Moore, OK tornado are redefining the way the industry sees tail risk. Such extreme events and their corresponding losses demonstrate the shortcomings inherent to using historical experience as the sole foundation for a view of thunderstorm risk.

From an Exposure Perspective

More people are living in high-risk regions like the Great Plains, the Midwest, and the Southeast, increasing the amount of insured exposure at risk. From 2007 to 2012, high-risk states like Oklahoma, Nebraska, and Kansas exhibited some of the largest increases in direct premiums written. With increasing exposure comes an increased likelihood of impact from severe weather, especially in high risk areas, making it imperative to understand the risk holistically, not just in large population centers. Relying on experience based rating alone will make it difficult to estimate losses in rural or newly developed areas, as such areas generate limited historical records.

Do these trends have a material impact on the North American catastrophe risk landscape? Absolutely; the impact is clear if we look at the annual loss numbers.

Annual Losses

In the U.S., average annual losses from severe thunderstorms are second only to hurricanes, causing over $10 billion in insured losses each year since 2003. In fact, losses driven by tornado, hail, and straight-line wind collectively contributed to more than one-third of U.S. annual insured losses between 1993 and 2012.

Where Can We Go from Here?

Historical experience, while important, may not be sufficient for fully understanding severe thunderstorm risk. Relying solely on this data could lead to poor underwriting practices, misinformed pricing decisions, and ineffective portfolio management. For a more complete picture, it is necessary to add a probabilistic element to the historical analysis, so you can estimate thunderstorm risk anywhere in the country (not just population centers), and differentiate the risk accurately across regions, lines of business, and risk characteristics.

What has been your experience with estimating severe thunderstorm risk?