Algorithmic underwriting set to boost options for corporate buyers

The adoption of algorithmic underwriting will increase competition and reduce frictional costs for commercial and specialty risk transfer, technology experts at WTW told Commercial Risk.

As the digitisation of the insurance market continues apace, more companies are investing in algorithmic underwriting, which uses technology to automate part or all of the underwriting process.

The emergence of algorithm-based portfolio underwriting is a positive for large commercial insurance buyers, said Richard Clarkson, global market leader for global specialty at insurance broker WTW in London. “As a corporate buyer, you fundamentally will want your broker to be able to access capacity through all these channels to get the best coverage. It will increasingly become a point of distinction, whether your broker can access these digital market places,” he said.

“Digital markets will have different price points and will perform differently. And this will become more and more the case. As a corporate buyer, as you look to transfer your risk, do you want to offload your risk into all the channels and market places that might present themselves,” Clarkson said.

He expects digitisation will bring more capacity to the market, by attracting more insurers and more capital to the insurance space. “As the ability for the insurer to describe their portfolio becomes greater – as data and technology create new insights – then your ability to effectively pitch your portfolio to a wider range of capital providers increases,” he said.

Algorithmic underwriting could also bring insurers and large corporates closer to the sources of capital, according to Clarkson.

“We are actively exploring, on behalf of [insurance companies], ways they can directly access capital markets – like securities, equities and debt markets – with this type of technology. It will take time, but we can see a method to describe the performance of a portfolio, tranche by tranche, which allows them to potentially access those capital markets,” he said.

In 2020, London market insurer Brit launched Ki, the first fully digital and algorithmically driven Lloyd’s syndicate. Ki uses an algorithm to assess submissions and automatically quote for business via a digital platform, which brokers access directly. Ki also began to offer follow capacity from other Lloyd’s syndicates, starting with Travelers and Aspen from 1 January 2024, and Beazley from 1 April.

“The success of Ki has catalysed how the market thinks about itself. We are now seeing deep appetite and conviction across the market that this will now be a channel within the London market, and people are now actively investing and deploying,” said.

“They see [algorithmic underwriting] presenting a whole range of opportunities. One opportunity to build, then to use it for their follow business, [Lloyd’s] coverholder business and facilities, and any other aggregated underwriting solution,” he said.

Going forward, insurers will increasingly distinguish between risks written on a case-by-case basis and those underwritten on a portfolio basis, either within an insurer or through separately branded and capitalised entities, predicted Clarkson.

“The distinction between case underwriting and portfolio underwriting will become more marked, and portfolio broking will meet portfolio underwriting. There will be whole parts of the market differently traded and managed,” he said.

The ability to make more granular, agile and forward-looking decisions has suddenly taken off, noted Clarkson.

“Algorithmic-based portfolio underwriting can bring about significant efficiencies. The bigger prize, however, is that the risk decisions within the value chain are more effective, so underwriters can make better decisions, set better prices, attract different forms of capital etc, which will lead to better outcomes for those that purchase insurance from the market, and create better returns for investors in the market,” he said.

Historically, risks are seen and underwritten on an individual basis in the London market, but portfolio underwriting uses data and algorithms to replicate decision making for a group of risks. Under the control and oversight of an underwriter, algorithms can help triage and select submissions, and then either trigger further action, such as referral to an underwriter, or generate an automatic quote.

Algorithmic underwriting can automate much of the traditional, time-consuming and manual underwriting processes usually undertaken, freeing up underwriters to focus on higher value work, explained Clarkson. Distinct from automated underwriting, where underwriting rules are fixed, algorithmic underwriting is more dynamic because it enables insurers to make decisions according to a set of rules. But decisions can differ depending on the performance of the algorithm.

Clarkson believes that London market risks will increasingly be “channelled” to either automated or algorithmic underwriting depending on their attributes. “We are already seeing this emerge in the market place, and I think as we go forward there will be more awareness of the distinction between these two, and the different types of business that fits into these two buckets,” said Clarkson.

Algorithm-based portfolio underwriting goes hand in hand with the digitisation of the insurance market. In the not-too-distant future, commercial and specialty risks will be traded electronically using algorithms to automate and speed up the underwriting process. To help facilitate this shift, last year WTW launched Neuron, an insurance trading platform that connects the systems of brokers and insurers, and facilitates algorithm-based insurance trading.

“We created Neuron because we saw a bigger shift in the market place. We think in this decade there will be profound change… Technology has suddenly matured, both in terms of people being able to manage their data, as well as connect their systems internally and externally, to bring about very different ways of trading,” said Clarkson.

Neuron will initially focus on algorithm-based portfolio underwriting for the London follow market, as well as lead business in the US small to middle market. In London, WTW has identified 18 priority classes of business that account for around 50% of the follow market – such as property, cyber, accident and health, directors and officers, errors and omissions, marine hull and cargo.

Digital trading will enable brokers to get more quotes, from a wider range of insurers quicker, and with lower frictional costs, according to Louise Smith, global leader of Neuron Digital Trading.

“You could potentially see more people enter the market because you will be able to access distribution and capital quicker. You may see people go into areas that they are not in today. If you can access areas at a lower cost, then you have more choice around your business model,” said Smith.

Algorithm-based portfolio underwriting could also increase the range of risks with direct access to capital markets, explained Smith. Today, insurance-linked securities mostly focus on peak natural perils, but a shift to portfolio underwriting and greater data transparency could see capital markets trade other risks, like cyber, she said.

“Insurers always want to expand coverages and the risk transfer envelope. But the thing that has stopped them from doing this in the past is the ability to model those risks with sufficient confidence, in order to set an effective price. But the more data we get, the more we can connect into other data services, the more it becomes possible for the insurance industry to offer that expanded risk transfer envelope, and to close the protection gap,” said Clarkson.

“This type of innovation [algorithmic underwriting] is a building block towards that outcome,” he said.