Case study: Calculate demand surge

Construction cost data obtained from XactAnalysis following the U.S. hurricanes of 2004 and 2005 was used by Verisk to validate the demand surge function. In particular, Verisk first determined the change in cost of individual building components due to each storm, and then how each component contributed to aggregate demand surge.

Calculate changes in material and labor costs

For each of the 2004 and 2005 hurricanes, and for each geographical area affected by the storms, the change in cost for the materials and labor categories was calculated.

The percentage change was calculated separately for the quarter before the storm occurred, the quarter in which the storm occurred, and the quarter following the storm. To isolate the cost changes due to the storm from other economic factors, the change in U.S. average costs was subtracted from the cost change reported for the storm-affected region.

For example, the following table shows that the cost of roofing labor in Lee County, Florida, increased by 35.9% (42% minus 6.1%) between the third and fourth quarters of 2004 as a result, primarily, of Hurricane Charley.

Calculate component contributions to total repair costs

To estimate demand surge at the county level (the resolution most closely matching the resolution of the XactAnalysis data), engineers first estimated how physical damage from the storms was distributed across six building components: roof covering, roof sheathing, windows and doors, interiors, exterior walls, other (e.g. plumbing, electrical).

The estimated physical damage to each component was then translated into a dollar amount by applying a damage ratio to the replacement cost of the component, according to the data below.

Component damage ratios for values between the mean damage ratios (MDR)provided above were calculated by linear interpolation. This step was done at a ZIP code level.

As an example, ZIP code 33921 had a mean damage ratio of 17.36% from Charley. To get the component damage ratios for this ZIP code, values for each component were interpolated between the 5.40% and 20.20% mean damage ratio data shown in the table. Thus, the roof cover component damage ratio was 32.47%, roof sheathing 1.96%, and so forth.

This calculation was made at the ZIP code level and aggregated by county by weighting each ZIP within a county by its modeled loss for a given event. Each component’s repair cost was then calculated as a share of the total repair cost for the building. The table below illustrates this calculation for a typical home in Lee County, Florida, in the aftermath of Hurricane Charley. The top row provides the average replacement cost for each building component as estimated by Verisk construction engineers. For example, the average replacement cost of a roof covering in Lee County is estimated at $5,899.

The second row shows ’s estimate of the component damage ratio, or the percentage of each building component damaged by Hurricane Charley in Lee County. For example, it is estimated that 20% of the roof covering will be damaged. The product of the replacement cost and damage ratio is calculated for each component to achieve the repair cost for that component (e.g. $5,899 * 0.20 = $1,180). The sum of all the products (the sum across row three) is the total cost to repair the property ($5,713). The final row gives the percentage contribution of each component repair to the total repair cost.

Calculate the contribution of material and labor to repair costs

Because demand surge affects materials costs and labor costs differently, it is necessary to determine the breakdown of repair costs between materials and labor for each building component.

Results as estimated by construction engineers are summarized in the table below.

For example, it is estimated that 29% of the cost of repairing roof coverings is attributable to materials and 71% is attributable to labor. The Materials column indicates the type and composition of materials need for the repair. The Labor column indicates the type and composition of labor need to execute the repairs. Where the materials needed do not match the most common material types, the repair costs are attributed entirely to labor since, in any case, the demand surge effect on materials prices is minimal.

Calculate the effect on demand surge of individual components

Continuing with the example of the cost of roof covering, we know from the previous table that the repair cost breakdown between materials and labor is 29% and 71%, respectively.

From the analysis of the XactAnalysis cost time series data for Lee County, we know that labor costs experienced a 35.9% increase as a result of the storm . A similar analysis (not shown here) shows a decrease of 0.3% in the cost of roofing materials between Q3 and Q4—that is, virtually no change.

As shown in the table above, the total demand surge effect on roof covering repair costs from Hurricane Charley in Lee County is estimated at 25.4% for the third quarter (the quarter in which the storm made landfall). From the table below, we know that repairs to roof covering account for 20.6% of the total building repair cost. If we multiply the demand surge effect on the component cost by that component’s contribution to total repair costs, we arrive at the component’s contribution to total demand surge. This process is illustrated in the first column of the table.

These steps are repeated for each of the six modeled building components and the results summed. As the table shows, the total demand surge effect from Hurricane Charley in Lee County is estimated at 15.81% for the third quarter.

Calculate demand surge across the affected region

The demand surge calculation is repeated for each of the counties significantly affected by each storm.

The aggregate demand surge for the storm during the quarter is calculated as a weighted combination of the county-level demand surge factors, with weights determined by the share of the modeled county loss in total losses.

The table below shows the results for Hurricane Charley for the first quarter following landfall. Total demand surge for this quarter is calculated at 14.79%.

For each event, demand surge is calculated for the quarter in which the storm made landfall (capturing essentially the change in cost before the storm to after the storm) and for the next three quarters (essentially to beginning of the next hurricane season). The example for Hurricane Charley is shown below.

Calculate cumulative demand surge by event

To determine the total demand surge factor for each event, the cumulative demand surge for each quarter was weighted using the percentage of total claims paid during that quarter. An example, using Hurricane Charley, is provided below.

The aggregated demand surge for each of the storms analyzed is provided in the following table. Hurricane Charley has the highest demand surge despite the losses being lower than those from some of the other events. This is because post-Charley damage surge includes the incremental inflationary pressures of Hurricanes Frances, Jeanne, and Ivan. Similarly, demand surge for Hurricane Jeanne, which caused lower losses than either Charley or Frances, is quite high because it followed a path very similar to Frances less than three weeks after Frances made landfall, putting maximum pressure on local resources.

Aggregation of 2004 losses

The major storms of 2004—Charley, Frances, Ivan, and Jeanne—all made landfall in or near Florida within a 45-day period.

The demand created by events that occur in such close proximity has a bigger impact on the regional economy than any single event taken independently. Because of this and because the XactAnalysis cost data is provided quarterly (therefore making it difficult to isolate the effects of individual storms), we took the additional step of calculating a single combined demand surge factor for the four storms together. This factor is simply a weighted combination of the aggregate demand surge of the individual storms, with weights equal to the share of each storm’s modeled loss in total losses for the four storms.