How to Predict Fire Insurance Losses

With the coronavirus pandemic resulting in more Americans working, playing and schooling from home than ever before, many are now spending an additional 8.5 hours per day in their homes. And as college students and family members moved back in, home sizes have increased. As a result, homes across the nation are taking on significantly more wear and tear. More people spending their days at home means more cooking, more electrical and appliance usage, more space heaters and even more hot water usage, all of which can result in fire accidents.

Majority of Fires Caused by Non-Weather Risk

While catastrophic wildfires have been widely covered in the media, especially over the last few years, many are unaware that only 16% of structure fires are caused by wildfires and lightning while 84% are caused by non-weather risks. Non-weather fire damage accounts for $5.8 billion of annual losses, or approximately 14% of all homeowners insurance losses.

Data from the National Fire Protection Association indicates that cooking was the leading cause of home structure fire incidents in 2019 at 49% of all incidents. Following it was heating equipment malfunction at 14% and electrical distribution and lighting equipment at 10%.

As the potential for fire damage increases from greater time spent at home, it is critical for insurers to be able to properly predict the potential for fire losses.

How to Assess Fire Losses

The severity of loss given a fire event is sometimes calculated based on three metrics:

1. The accessibility of a responding fire department, or distance from the station to the property
2. The likely speed of response
3. The accessibility of water to douse the fire, likely via a fire hydrant.

While this sounds logical, these metrics are not entirely predictive of fire losses. Fire loss prediction has previously been focused on severity, with the underlying assumption that every property and location had an equal, stochastic chance of a fire incident. Thus, the focus of prediction was on speed of response and access to water in combination with the size and value of the property and its contents.

But 30 years ago, if your house caught on fire, you had 17 minutes to get out safely. Now, you only have 3 minutes. -That's right-houses burn 80% faster today than they did just thirty years ago.

Structures burn faster now than they used to because of the widespread use of petroleum-based furnishings in home and commercial buildings, including petroleum-based furniture foam, carpet, drapes and plastic piping and hoses. This, in combination with recent changes in home construction like manufactured trusses and the widespread use of composite materials, means that fire spreads from room to room much more rapidly than it did decades ago. If the fire department doesn't make it within 180 seconds of when the fire starts, the likelihood of a huge loss is significant.

This means that the likelihood of the fire event occurring in the first place is a far more relevant metric than it used to be. Focusing only on severity and not frequency means that the probability of the fire starting and the role that the composition of a structure plays in its spread, important factors for estimating expected losses, are both ignored.

Additionally, it can be incredibly complex to access the data for fire departments and water accessibility. Getting it right involves cooperation and precise coordination from localities, which can leave insurers to navigate a decentralized patchwork of local government frameworks. To be able to properly utilize this information, it would require rapid access to consistent information-not an array of survey forms, on-site visits and unique processes.

By adding the probability of the fire event occurring in the first place to the expected loss calculation, insurers can account for important variables such as a structure's location and the quality of external construction materials,

This is why it is critical for insurers to understand the unique risks which start fires at each local area, through an easy-to-access and consistent scoring system. By combining fire frequency and fire severity into one risk score, we can help communities be more prepared for when disaster strikes.

©2021 CoreLogic, Inc. All rights reserved.

Public link

Please use the above public link if you want to share this noodl on another website.