AI used effectively should let the underwriter underwrite

The latest episode of the Voice of Insurance podcast features Simon Bird, Group Executive Underwriter and Active Underwriter of Brit Insurance’s Syndicate 2988.

Insurance market veteran Simon Bird, who leads Brit’s underwriting for its Syndicate 2988 at Lloyd’s, was the latest guest to feature on Mark Geoghegan’s podcast, The Voice of Insurance.

The conversation turned to the role of underwriters and algorithms in market pricing throughout the cycle, from a hard market turn through to the inevitably less disciplined soft market stages.

Among his many pearls of wisdom, Bird noted that here is one thing that people find tough to do, but for which technology can provide some additional discipline and consistency, when called upon.

“One thing human beings are very difficult doing is having drawn a line in the sand is respecting that,” he said. “If you programmed the algorithm and model correctly, you would have an absolute hard stop below which the market will not sink.”

This kind of discipline exists in theory, when plans are drawn up, but can be lacking in practice. Bird noted that insurance linked securities (ILS) and industry loss warranty (ILW) markets, when pricing has dropped below a certain level “the market, or the models, just put their pens down”, enabled a better chance of putting a pricing safety net in place.

The apocryphal story of “sending underwriters to the golf course” to stop them writing underpriced risks, can therefore apply to technology-led underwriting, such as Ki Insurance, ultimately underwritten by Brit, which underwrites following business as the first fully digital and algorithmically-driven Lloyd’s syndicate.

Ki is an example of using artificial intelligence (AI) to underwrite, albeit only following business, rather than leading on risks as they are presented by brokers, with pricing agreed by a lead underwriter, according to the subscription system at Lloyd’s. Mark asked the tricky question of how much of an underwriter’s job should be done by AI.

“AI is a challenge, obviously, because it’s technological progress on steroids, because it’s touching on human and philosophical aspects in a way that just technical digitisation doesn’t do. “It’s with us, it’s coming, and you can’t pretend it’s not there, so, you’ve got to embrace it and manage it somehow and look at how it can improve underwriting,” Bird said.

Any talk of AI replacing commercial and specialty lines underwriters is “a long way off”, Bird suggested.

He drew a line between commoditised SME business and larger property risks that require a lot more consideration among underwriters for insuring individual risks. However, AI can be of great use to big commercial lines business that is still beset with myriad manual processes.

“What we find is that underwriters are doing far too much mundane prep, before underwriting,” said Bird. “If you can refine the tools that triage those risks, then you are letting underwriters underwrite, add the value, add the selection process, and you use it as an ever increasingly sophisticated filter.”