Optimizing Reinsurance Strategy with Building-Based Geocoding

Accurately underwriting property risk for reinsurance

When consumers insure their property, they often don’t realize that the carrier they choose also has insurance on their property. Reinsurance exists so carriers can limit their overall risk and do not need to retain the full value of the properties in their book of business as liquid assets. Property and casualty (P&C) carriers assume an amount of exposure (also called “retention”) per risk below a certain threshold and above another. For example, in the treaty established with their reinsurer, an insurance carrier could be liable to pay property claims up to $1M and over $10M; any damages resulting in claims between $1M and $10M are assumed by the reinsurer. 

Harford Mutual Insurance Group, a regional P&C insurance carrier based in Maryland, has a similar reinsurance strategy that allows them to scale their business while reasonably managing their liability. Harford Mutual Insurance Group’s underwriting team found that the optimal way to manage their reinsurance required grouping buildings in close proximity to each other that could be damaged by a single occurrence. Using this methodology, the team is able to assess risk for a “value subject” containing buildings within 25-100 feet of each other; the treaty then applies to each group as an individual risk. This means that if one apartment building completely burns down, resulting in $3 million in damages, and the building next door incurs $800,000 of damages, the neighboring building is also covered by the company’s reinsurance policy because it is underwritten as part of the same value subject.

While grouping is ideal for optimizing Harford Mutual’s reinsurance strategy, the related underwriting process was manual and time-consuming, especially as the carrier expanded its book of business. To effectively group properties into value subjects, underwriters would feed the addresses into a homegrown mapping application, then inspect and correct every point to make sure the result was a building-based geocode. Once the geocodes were manually corrected, underwriters would carefully measure the distance between the edges of buildings on the property being insured and the nearest edge of any nearby buildings that were already insured by Harford Mutual to identify groups for value subjects. They would also manually measure the distance between the buildings on the property and any nearby buildings, data which can affect the property’s risk. Given the importance of accuracy in this process, a manual process was thought to be critical for this workflow. 

This manual process was justifiable for properties worth millions of dollars, but unfeasible for the majority of properties Harford Mutual underwrites. Additionally, it’s not only the new properties being quoted that need to be analyzed each time; existing groups need to be adjusted when nearby buildings are added to Harford Mutual’s book of business, or when policies are endorsed, lapse, or removed. As a result, the team was running the same manual process for over 12,500 P&C insurance quotes a year, taking up a material amount of underwriters’ time, and making it difficult to scale their book of business without significantly increasing the number of underwriters they employ.

Leveraging Ecopia’s Building-Based Geocoding to optimize property risk assessment

For years, Harford Mutual looked for a solution that could allow them to rapidly scale without sacrificing the accuracy and precision required for their reinsurance strategy. By partnering with Ecopia, Harford Mutual now leverages building-level geocoding and building footprints (derived by leveraging AI-based systems to mine high-resolution geospatial imagery) to group properties into value subjects. Their new process for underwriting policies to optimize for reinsurance has transitioned from manual geocoding and measurements to a fully scalable system powered by Ecopia’s Building-Based Geocoding. 

In this new process, Ecopia ingests individual exposures from Harford Mutual’s policy administration system for both prospective and existing businesses, generates building-level geocodes and building footprints for each one, and measures how far each is from other structures. When a property is measured to be within a certain distance of another property currently insured by Harford Mutual, they are automatically grouped as a single value subject and its combined value is assessed. 

The record of this group as a single value subject is provided to Harford Mutual and used to define a single risk under their treaty. If grouping the two properties results in their combined value falling above the reinsurance treaty, Harford Mutual’s underwriters are notified of the grouping as a single value subject and the potential need to purchase facultative reinsurance. The Ecopia platform is equipped with analysis tools specific to this grouping process. These tools allow for customized groupings based on individual risk characteristics, ultimately producing a digital twin of Harford Mutual’s book of business. 

The number of properties Harford Mutual underwriters need to manually assess has significantly decreased with the automation enabled by Ecopia. Instead of manually checking and measuring each property when a quote is generated or policy changes, the underwriting team is now only performing manual analysis of value objects that exceed Harford Mutual’s reinsurance treaty. This means the vast majority of property grouping for reinsurance purposes is automated, allowing the underwriting team to work more efficiently as the company expands.

“With Ecopia, Harford Mutual has been able to reduce time spent on manual property assessments by 75%,” said Tony Halloran, Project Lead at Harford Mutual. “Partnering with Ecopia allows our underwriters to move quickly without sacrificing quality as we grow our book of business - that alone makes the investment worth it.”

Why Ecopia?

Until connecting with Ecopia, Harford Mutual struggled to find a geospatial partner that could meet their requirements of both speed and accuracy. Other data providers offered part of the solution, but none could reliably deliver high-precision building footprints and building-based geocoding across urban, suburban, and rural communities. One mapping provider Harford Mutual spoke to claimed to provide building footprints that covered 90% of the US population - a statistic that quickly loses value when considering the number of insurable properties located in sparsely populated areas. 

Harford Mutual thoroughly tested each solution before making a decision. Ecopia was the only data partner to pass each test. To conduct these tests, Harford Mutual selected small buildings in rural communities to evaluate each provider’s coverage. Ecopia’s data was found to be the most accurate, up-to-date, and complete, with every property tested being identified by Ecopia’s AI-based systems during the trial. 

“Ecopia was the only data provider that could identify and match each obscure, rural structure we gave them,” said David Curtin, Vice President - Underwriting at Harford Mutual. “That gave us the confidence in its data quality and freshness that ultimately led us to choose them as our partner.” 

When comparing building footprint providers, most sources deliver around 130 million buildings at about 66% accuracy, while Ecopia’s dataset provides over 173 million footprints with at least 95% accuracy. Ecopia’s data is also updated each year to capture the dynamically changing world, which other providers struggle to keep up with.

The team at Harford Mutual was also assured by the fact that Ecopia develops their own data by leveraging AI-powered systems to mine the most recent commercially available geospatial imagery. Harford Mutual was not interested in working with a data provider that aggregates and packages up datasets from fragmented and inconsistent open-sources; they wanted a uniform consistent source they could trust.

“From the beginning, Ecopia was ready to build exactly what we needed to optimize our reinsurance strategy. They listened to our challenges and didn’t shy away from the complexities. The leadership team made it clear that no problem was too big for them to solve.” - Jeff Rink, Executive Vice President, Harford Mutual.

What’s next?

As Harford Mutual continues to expand into more US states, the underwriting team is able to take on the work of insuring more properties while optimizing their reinsurance strategy. Harford Mutual’s underwriters have the capacity to grow with the company and work more efficiently without having to manually analyze each of the more than 50,000 properties the carrier insures.

To learn more about Ecopia’s insurance solutions, get in touch here.