Impact of demand surge on insurance losses
Overview of the economic drivers of demand surge and the methodologies used to model its impact on insured losses following large catastrophe events.
Market forces generally ensure that the availability of materials and labor in any particular geographical area are sufficient to accommodate a normal level of demand without affecting price. However, if demand increases sharply and unexpectedly—as after a catastrophe event—the pressure on resources can cause prices to inflate temporarily. The relative scarcity of resources can also result in increases in the time required to repair and rebuild damaged property.
Coupled with other economic, social, and infrastructural factors, this can lead to insured losses exceeding “normal” expectations for a particular event and portfolio, which are based on the assumption that sufficient resources are readily available at constant prices. This phenomenon is referred to as demand surge. The greater and more widespread the damage from an event, the greater will be the resulting demand surge and thus the greater will be the insured losses.
Verisk engineers and statisticians have developed a mathematical function that relates the amount of demand surge—or the factor by which insured losses increase—to the amount of modeled industry insurable loss. Developing a demand surge function requires a high degree of judgment because there are few historical data points, with Hurricane Andrew (1992), the Northridge earthquake (1994), and the 2004 and 2005 hurricane seasons being chief among them.
We developed the Verisk demand surge function based on examinations of the historical data, statistical analyses, reviews of economic time series, and analyses of construction material and labor cost data. We augmented with detailed client claims data, publicly available economic data, and findings from post-disaster surveys conducted in the aftermath of the 2004 and 2005 hurricane seasons with high resolution, and component-level cost data for the 2004 and 2005 hurricanes from Xactware, an ISO subsidiary.
We included demand surge in the modeled losses for historical events in the Verisk hurricane and earthquake models for the United States.1 The modeled losses compare well with reported losses, validating both the Verisk models and the Verisk demand surge function.
We developed the default Verisk demand surge function using economic principles, and validated it based on U.S. loss levels and component cost analyses. Because demand surge is a phenomenon seen only with especially large catastrophes, there are relatively few events with which to validate demand surge functions outside of the U.S. This scarcity of data is further complicated by the relative paucity of cost indices and detailed data.
But despite the scarcity of data, we are currently developing country/region demand surge functions. These will depend on, among other things, the size of local and national labor markets and, therefore, their ability to accommodate excess demand. The functions will be augmented by other labor, material, and construction indices as available. The functions will reflect the interaction between supply and demand of rebuilding resources and will be scalable to suit local economies.
In the meantime, for countries other than the U.S., you may choose to apply the U.S. demand surge function or a user-defined demand surge function, at your discretion. You are also encouraged to perform sensitivity testing to better understand the scale of impact and uncertainty inherent in applying demand surge to non-U. S. models and perils.