Leveraging data to better anticipate & plan for NAT CAT events

Hurricanes are the most violent storms on the planet. In addition to the extreme weather and destruction they bring, hurricanes are also notoriously unpredictable.

Hurricane Ian was no exception. Ian’s projected cone of uncertainty shifted considerably in the days leading up to its U.S. landfall on September 28, 2022. What was projected to be a tropical storm — or possibly a Category 1 hurricane — escalated within 24 hours into a serious Category 4 hurricane headed for the Tampa area.

This meteorological uncertainty exacerbates insurers’ claim and capacity readiness. By leveraging data from both internal and external, insurers can better anticipate and plan for actual loss exposure to events like this.

Here’s what our company did to prepare for a catastrophe and seize an opportunity:

Automated moratoriums

The prohibition on offering or adding new property coverage in territories that are expected to experience a CAT event customarily entails an underwriting moratorium to restrict binding authority. Typically, email bulletins and flash alerts are issued internally to underwriters and externally to agencies; programmers scramble to write code to restrict coverage in those ZIP codes, counties or states. Then, turning them off requires the same manual effort.

Coterie’s underwriting team worked with Opterrix, whose team includes self-professed insurance and weather geeks, to monitor weather events over a predefined threshold based on our underwriting guidelines. We built our proprietary automated underwriting platform, affectionately called Sleuth, to modify underwriting rules without programmer intervention. Using property-specific data from API integrations with numerous data providers, our director of underwriting implemented the moratorium in two clicks and worked with our marketing team to notify our distribution partners.

Real-time exposure monitoring

Coterie acquires over 3,000 data points on every location we insure — including a wide range of building, risk and policy-level characteristics. Our data team was able to pull this information, and working with Opterrix, we could instantly see Ian’s actual path, wind speeds and rainfall as it made its way across Florida. We isolated our exposure from state level or region level, down to the actual path level. This equipped us to make highly accurate, data-informed reporting to our reinsurers on the same business day as to our expected exposure to this event.

Insurers in Florida are required to report policy and claim data to the Florida Office of Insurance Regulation following a catastrophic event. We provided this data on the same day it was requested. This helped our reinsurers understand potential exposure and avoid delays in claim payments, capital allocation, and reporting to third parties.

Policyholder outreach

With businesses closed and residents evacuated, oftentimes policyholders don’t know the status of their property, much less who or how exactly to contact to initiate a claim.

As a result of the insights our team had into which customers were likely affected by Ian, we decided to reach out to those specific policyholders proactively to make it easier for our customers and agents to initiate the claims process. We sent out communications proactively to our policyholders who were in Ian’s storm path, as well as others who were possibly affected based on wind speed or other factors. Our in-house claim team was able to access the underwriting information to expedite the first notice of loss process, providing a superior customer service experience during a time of our customers’ greatest need.

Post-event assessment

A day after U.S. landfall, we received an email from David Tobias, co-founder and COO, at Betterview. They were offering their assistance to leverage NOAA aerial imagery to augment our existing data and his team quickly gave us access to their platform and ingested our policyholder data set.

Betterview’s product allowed us to see aerial imagery pre-event and post-event and identify actual property damage to potentially affected customers. It also provided indicators for our claim team where there was predicted damage based on their machine-learning models. So, within four business days, we now had credible data on the properties where our team needed to deploy claim resources, as well as more credible event exposure for our reinsurers.

Insurance relies on the law of large numbers in terms of pooling uncorrelated loss exposures to accurately predict and price risk. Catastrophes throw a wrench in this because the losses are highly correlated. Underwriters’ uncertainty pre- and post-CAT creates risk aversion and market gridlock. Ironically, these are the moments when consumers and business owners need insurers the most. By utilizing existing underwriting data, and collaborating with industry partners, insurance providers can more rapidly predict and plan for outcomes and improve responsiveness.

Pete Buccola, CPCU, is head of insurance at Coterie Insurance.

Opinions expressed here are the author’s own.

Related:

• FEMA logs 44K Hurricane Ian flood claims
• Hurricanes Nicole & Ian’s overlapping damage areas could muddy claims
• Ian’s toll: Munich Re warns on full-year profit target