A blind approach
Throughout the 2016-2017 winter season, headlines read “Northern California get its wettest winter in nearly a century,” “The drought is over in Northern California after up to 20 inches of rain and 20 feet of snow,” or, my favorite, “Atmospheric river events will dump 15 feet of snow on California.”
After a truly historic winter, the thought of fire was the last thing on both Californians’ and insurers’ minds alike. Ironically, 6 months later, headlines would read “Northern California Fires are Deadliest and Most Destructive in State History.” How could these two extremes exist so closely together? For a rational thinking person, the answer is unclear.
A deeper, scientific inquiry, however, offers clarity: The influx of precipitation dramatically expedited the growth of seasonal grasses and weeds. Record setting precipitation lead to record setting vegetation heading into summer, and after 6 months of dry skies and blistering California sunshine, that same mass of vegetation had dried out and transformed into record setting fuel for wildfires. Enter the season change, when autumn causes California oak trees to drop their leaves – adding to the available surface fuel – and bringing seasonal winds, like the Diablo and Santa Anas, that inundate regions with hot and dry conditions. Connecting an a-typically wet winter to an inferno of a fall is counterintuitive. The natural assumption is that the water gained in the winter will lessen the dryness of the following seasons. But when reviewing the variables objectively, the outcomes make perfect sense. So if the experts knew this, why was everyone so caught off guard?
In short, misinformed data. Generally, data consumers are not subject matter experts, so they look to data analysis for expertise and to inform their decisions. If data does not reflect reality, then it can have disastrous consequences for both residents and insurers. Risk data needs to capture science rather than intuition, and it needs to reflect recent changes in the factors that cause perilous events to happen in the first place.
A better solution
Fire Risk Pro from Pitney Bowes understands these needs through and through. The Fire Risk Pro model accounts for variables including slope, aspect, precipitation, vegetation type, vegetation density, recent burn perimeters, special wind event regions, etc. to more accurately calculate wildfire risk. The result is a national database of “fire sheds” with corresponding risk scores (classified 0-49) that updates annually to capture recent changes. Wildfire risk type is further distinguished as either Wildland, Intermix, or Interface to account for surface type groupings at a macro level.
Above: A look at the latest Fire Risk Pro “fire sheds” within the burn perimeters of the 4 major fires.
29.78 – The average Fire Risk Score within the burn perimeters
13.20 – The average Fire Risk Score for all of California
14.26 – The average Fire Risk Score for the entire U.S.
These areas were highly likely to burn as indicated by the model. With scores more than doubling the US average and more than 225% of the CA average, our clients were aware of these hotspots, and they were able to take the necessary precautions to prepare ahead of time.
An informed approach
When preparing for loss events, more accurately identifying at risk regions is only half of the picture. Now that we understand what’s happened conceptually, let’s examine the fires from a financial perspective.
To assist with total exposure and maximum probable loss (MPL) assessments, Pitney Bowes provides property-, parcel- and address-specific information across the United States. By aggregating property attribute information within the burn perimeters, we can calculate a number of metrics:
- $5.3B – the total “property assessed value” within the burn perimeters
- 7,039 – the number of structures within the burn perimeters
- $750K – the average structure value within the burn perimeters
- 11,768 – the number of parcels within the burn perimeters
- $450K – the average parcel value within the burn perimeters
A simple match from policy holders to addresses makes it easy for insurers to quantify the total risk exposure and MPL specific to their book of business. This information is critical to both insurers and reinsurers who need to maintain enough liquid funds to support claims pay-outs in worst case scenarios.
We can’t stop the forces of nature, but if we understand the cause of perils, identify high risk regions, and then account for the financial implications, we’re doing everything we can to responsibly prepare. Smart and more accurate data makes this all possible.