Tornadoes: A Costly Yet Complicated Risk for Midwest Metropolitan Areas

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 6 January 2015: 11:00 AM
131C (Phoenix Convention Center - West and North Buildings)
Megan E. Linkin, Swiss Re America Holding Corporation, Armonk, NY; and K. Orwig and D. N. Nelson

Due to their complexity, tornadoes are difficult to model using traditional catastrophe modeling approaches, but that does not mean the re/insurance industry can ignore them. There are 1,200 tornadoes each year in the US on average and event losses in excess of USD 1 billion are becoming more frequent, driven by growth in population, assets and values. The year 2011 saw record breaking tornado losses, with damages in excess of USD 25 billion.

By its nature, this elusive and deadly phenomenon defies traditional analysis. The volumes of data collected over the years fail to hold up to the task of probabilistic risk modeling. For example, tornadoes are more discrete and localized than tropical cyclones, so a much larger sample of simulations is required to determine the probability of impact at a given location.

Nevertheless, insurers must step up to the challenge and think the unthinkable: What if a tornado was to strike the downtown of a major metropolitan area? In recent years, the most notable tornado outbreaks have caused damage in smaller cities and towns such as Joplin, Missouri; Tuscaloosa, Alabama; Moore, Oklahoma and El Reno, Oklahoma. Downtown areas of cities like Chicago, St. Louis, Dallas, Indianapolis and Detroit have been spared the worst, but it's only a matter of time. Urban centers have experienced plenty of near misses and data suggests that historically, severe tornadoes have tracked through major cities. The downtown area of St. Louis, Missouri has been struck four times in the last 100 years, most recently in 1959, and downtown Chicago was directly struck by a tornado thought to be an EF3 in 1876. Therefore, it is imperative to not only understand the impacts that tornadoes have had historically, but what the potential impact of a large tornado striking a major city today.

Perhaps the most compelling reason to build our understanding of tornado loss potential is the growth in exposure to tornadoes, a trend that has accompanied population growth and accumulation of physical assets. Total insured values (TIV) in Dallas/Ft Worth, Oklahoma City, St. Louis, and Birmingham are now in the USD 50-100 billion range, with TIVs in some counties exceeding USD 100 billion and approaching USD 1 trillion.

Swiss Re uses a simple and logic-based methodology to construct scenarios to demonstrate the potential social and economic impact of a tornado striking the heart of a major city. For example, what would the economic impact be of a tornado like the 2013 Moore EF5 moving through the Chicago metro area? How many people could be affected if the 2011 Birmingham-Tuscaloosa EF4 went through the Dallas/Fort Worth metroplex? The authors will present their findings to these questions and more, while emphasizing the importance of planning and resilience to minimize the impacts of catastrophic events.