Auto Insurance Growth Opportunity #2: Write All Locations
Many insurers turn off certain ZIP codes or geographical territories. While this practice was effective in the past, it may limit future growth. With machine learning, the risk selection paradigm has changed, creating an opportunity to capture more profitable market share in every territory.
How Location Can Affect Exposure
Every auto insurance underwriter knows that location is a key piece of the risk puzzle. There are many possible reasons for this:
- Road Design. Some areas have dangerous intersections and roads, which contribute to higher collision rates.
- Congestion. Cities with more traffic typically have higher collision rates.
- Weather. Areas prone to slick roads and poor visibility may experience more collisions.
- Wildlife. Although many risks are higher in urban areas, rural and suburban areas with deer and other large animals may see higher rates of collisions with wildlife.
- Crime. Some areas have higher rates of car theft, vandalism, and other vehicle-related crimes.
The differences can be significant. To find out which cities have the highest risk of car accidents, Forbes examined data from the National Highway Traffic Safety Administration and Allstate. The study concluded that drivers in Atlanta have the highest risk of being involved in a collision, followed by Dallas, Baltimore, Detroit, and Philadelphia.
The National Insurance Crime Bureau publishes figures on car thefts, which likewise show significant differences by location. Vehicle thefts have been rising since 2019. In 2023, the number of cars stolen surged to 1,020,729, up from 1,008,756 in 2022. Although this was a nationwide increase, some areas saw much steeper increases than others.
- Between 2022 and 2023, the number of stolen cars rose by 64% in the District of Columbia – this was the jurisdiction with the largest increase. The increase was 63% in Maryland, 33% in Connecticut, and 18% in Nevada.
- California still has the most car thefts a year (208,668 in 2023), followed by Texas, Florida, and Washington.
- In terms of urban areas, the Los Angeles-Long Beach-Anaheim area in California had the most thefts (72,460 in 2023), followed by San Francisco-Oakland-Berkeley and Chicago-Naperville-Elgin.
How Auto Insurers Account for Regional Differences
To manage loss ratios, insurers need to account for regional loss differences. There are a few ways to do this. They can:
- Charge higher rates, based on the rate of accident and/or theft claims in an area.
- Turn off entire cities or regions, pulling out of areas because the losses are too high.
- Raise prices so high they price themselves out of the market.
All these approaches come with challenges. For one thing, some consumer watchdogs cry foul over ZIP code exclusions and price differences. That’s because loss rates aren’t the only thing to vary from one area to another – racial demographics also vary. In fact, a report from ProPublica found that minority neighborhoods tend to pay up to 30% more for coverage compared to other areas with similar losses. To fight potential discrimination, some states ban the use of ZIP codes as a rating factor. The Insurance Journal says Massachusetts recently considered legislation that would add this restriction to the state’s laws.
Additionally, it’s possible to give location too much weight in the risk assessment formula. It’s important to consider regional factors in the context of many other interrelated risk characteristics. For example, someone who lives in an area with low crime could still become the victim of car theft, just as someone who lives in a high collision area could go years without a claim. Insurers need to look at the complete picture.
A New Approach to Managing Profitable Growth
Insurers who avoid high-risk ZIP codes lose out on a lot of business. If multiple insurers avoid the same areas, the community becomes underserved. Insurers that find a way to serve these areas stand to see huge growth – as long as they do so while protecting their bottom line.
Risk factors are complicated. For every policyholder, there are dozens of intersecting factors that contribute to the overall risk profile. It’s so complicated that human underwriters might be unable to consider everything, which is why insurers have historically relied on segmentation.
Thanks to machine learning, this is no longer necessary. Insurers can use machine learning to evaluate dozens of risk factors, determine the likelihood of a claim, and produce a rate adequacy score. If the filed rate is adequate for the risk, the risk may qualify for straight-through processing. If the rate is inadequate for the risk, the program flags the application for further review.
This isn’t a hypothetical scenario for the future. It’s a solution that is available now through Soteris. With Soteris risk scoring, ZIP code exclusions become a non-issue, giving insurers the freedom to confidently write rate-adequate risks in all ZIP codes.
Interested in Other Auto Insurance Growth Opportunities?
