Modeling of catastrophic events with applications to insurance

Citation data:

UNI ScholarWorks, University of Northern Iowa, Rod Library

Publication Year:
1999
Usage 6
Downloads 3
Abstract Views 3
Repository URL:
https://scholarworks.uni.edu/pst/18
Author(s):
Noga, Mary
Tags:
Tornadoes--Economic aspects; Risk (Insurance);
thesis / dissertation description
Tornadoes effect every area in the continental United States. There are great differences in the number and amount of damages caused by tornadoes in different regions. The mid-west is nicknamed tornado alley, because of the large number of tornadoes in this region each year. A tornado is commonly described as a "rapidly rotating, slender, funnel-shaped cloud."The National Climatic Data Center (NCDC) has data on tornadoes over the past four years. This data includes the state, date, and damage done by each tornado. For the purpose of this study a data-set including all tornadoes causing over ten million in damages was retrieved. Table 1 shows the forty data points found.The study of tornado damages is important for all insurers as all geographical locations in the United States have tornadoes. The damages done by tornadoes are rare and large. Such values are mathematically defined as extreme values. According to Emil Gumbel one purpose of studying extreme values "is to explain observed extremes arising of samples of given sizes ... " (qtd. in Hickman). Many times experience alone will not enable the insurer to accurately insurer against losses. Therefore, it may be appropriate to attempt to model tornado damages. This would better enable the insurer to prepare for future losses.