Takeaways:
- AI can compare portfolio underwriting data to real-time indicators
- Paired with algorithms, AI automates portfolio underwriting process
- Client and staff management benefits
AI capability is driving increased adoption of portfolio underwriting, according to underwriting executives who spoke in a recent
Portfolio underwriting is used by both personal and commercial lines insurers to manage risk for groups of policies by using data analysis to uncover trends and determine pricing.
Portfolio underwriters can use AI to extract information from insurance brokers submissions over email, explained Bijal Patel, co-founder and CTO of Aurora, an underwriting platform. The data extracted this way can be compared to real-time data such as geolocation peril models for flood or fire risks.
Algorithms can be used in conjunction with AI to automate portfolio underwriting processes, according to Patel. In this manner, more data points can be fed into underwriting decisions. "We can put predictive analytics on top of that to find new factors that could help you optimize pricing and underwriting," she said.
Aurora specializes in middle market commercial underwriting, which encompasses a variety of coverages. Such heterogeneous risks are more difficult to automate, according to Patel, which makes algorithms useful for underwriting.
Underwriters are pairing AI with algorithms to clean up data, Patel explained. "The opportunity we have is to start getting a more granular view, more consistent view, more structured view of data," she said. "A lot of insurers are hesitant about algorithms because they don't have that clean data, but there is an opportunity to start fresh. Even if you don't have this accessible, clean historical data, start collecting it now."
Greater use of AI and algorithms in portfolio underwriting helps insurers better handle clients, according to Tom Nasso, chief underwriting officer at Falvey Insurance Group, a cargo, marine and logistics insurer.
"It's separating out which clients fit, which insureds are okay being managed within a market – which ones really want to be seen as standing on their own based on their loss control, their risk management and all the other things that they're implementing to make their risk better than their peers," he said. "Mid-size, large risk management level clients don't want to be lumped into a portfolio. They don't want their account to be priced based on the results of everybody else."
The technology also helps insurers manage their staff internally for portfolio underwriting work, added Patel.
"If you're using an algorithm to underwrite at the case underwriting level, and you're letting underwriters free up their time to focus on portfolio management or building broker relationships, the volume of business you can write in a day is significantly higher, and you can scale that," she said.
