Claims Analysis Study of May 24, 2011 Hailstorms in Dallas-Fort Worth

Since the 1960s, considerable knowledge about damaging hailstorms has been gleaned from crop-hail insurance data. Hail-related losses were generally thought to be better documented by crop-hail insurance companies than by property insurers, who did not distinguish between hail, wind, tornado, rain, or lightning losses. Crop-hail loss data had historically been used by Dr. Stanley A. Changnon to understand the spatial and temporal aspects of economic losses attributed to hailstorms, and he used the crop-hail loss data as a proxy to estimate total economic loss data across the nation. The lack of property-hail loss data was noted by Dr. Changnon as a key problem in conducting economic analyses of hail losses.

The Insurance Institute for Business & Home Safety (IBHS), an independent, non-profit, organization wholly supported by property insurers and reinsurers, is focusing research efforts on understanding damaging hail events in order to reduce the economic impact of hail losses. Since opening the multi-hazard IBHS Research Center in South Carolina in 2010, the Institute has undertaken a major multi-faceted research effort to study hailstorms, which cause losses in excess of $850 million annually, based on Dr. Changnon's research.

The Institute's ongoing hailstorm research program includes in-situ hail measurement field research, laboratory hail impact testing of various building materials and component systems, and closed insurance property claims studies following damaging hail events. This paper presents the analysis of property claims and policy-in-force data for more than 67,000 residential properties located in 20 ZIP Codes in the area affected by a series of thunderstorms that produced significant hail in the Dallas-Fort Worth metroplex on May 24, 2011. The storm caused an estimated $876.8 million in insured damages to property and automobiles, according to the Texas Department of Insurance.

IBHS evaluated the importance of roofing material type during this event by determining the claim frequency and severity resulting from hailstone impacts. This information is valuable for insurers as they rely on data such as these to better understand the risks of underwriting specific buildings. The spatial distribution of damages to each roofing material type was also determined. The results indicated the claim frequencies from this event were highest for metal and wood roofs, but the sample sizes of those materials as compared to the sample size of asphalt shingle roofs, were quite small. The claim severities for asphalt shingle roofs and tile roofs were lower than the claim severities for metal, slate and wood roofs. Metal and wood roofs still had the highest claim severities, even after normalizing by the Coverage A limit of the insurance policy.

In recent years, there has been an increased focus on the effects of aging as it impacts the durability and performance of materials. Previous post-event studies from damaging hail, wildfire, and wind events have indicated that older buildings tended to have a higher claim frequency and severity than newer buildings. This could be due to changes in the building codes requiring stronger construction for new buildings, while in other instances it could be due to degradation of the building materials or poor maintenance practices for the older buildings. The life expectancy, frequency, and severity of damage to various roofing products as they age provides valuable information to insurers in evaluating underwriting risks and in determining the extent of losses that could be expected from a damaging hail event. The effects of age with regard to roofing resiliency to hail in this event were investigated by determining the claim frequency and severity of the various roofing material types when categorized by age.

In addition, the study also evaluated the claim frequencies and severities for various building component systems to determine relative damage costs associated with roofing systems versus wall, window, door, and other systems. The results showed roofing-related claims far exceeded those of other building systems for this particular event, with the cost of roofing-related losses more than 10 times higher than for any other component group. More than 90% of the money paid out for claims in this event was provided for roofing repairs.

Lastly, a spatial distribution of the claim severity data was overlaid with radar-estimated hail sizes to allow for a comparison between the two datasets. Results indicated some locations where there was reasonable agreement between the radar and damage datasets, while in some locations radar and damage datasets were contradictive, having either low damage rates associated with large radar-estimated hail sizes, or high damage rates associated with small radar-estimated hail sizes.

A brief explanation of the weather system and forecasting leading up to this damaging hail event is discussed, and the methodology for selecting the 20 ZIP Codes for inclusion in the study is described. Recommendations for improved data collection and quality of insurance claims data, and guidance for the conduct of future property insurance claims studies are summarized.