Lead Forensics


Maintaining an Even Keel through Data & Analytics

09.06.21 Alex Field

Last week, I attended the virtual roundtable All Aboard for Data Analytics, as part of the Marine Insurance Americas online conference. Moderated by our own Thomas Anderson, participants, Susan Carr, AVP Technology & Operations, The Great American – Ocean Marine Division; Drew Feldman, Head of Marine, CNA Hardy; Mike Karbassi, Chief Underwriting Officer and Founder, Corvus Insurance; Andrew Kinsey, Senior Marine Risk Consultant, Allianz; Vitor Ribeiro, Head of Data, AM Re Syndicate Inc. and Luke Wolmer, Senior P&C Analytics Manager, Swiss Re, explored how data and algorithmic underwriting are impacting the marine insurance market. They also discussed how marine insurers can drive efficiencies through harnessing analytics based insights.

Marine insurers today are focused on driving premium growth and increasing market share through portfolio diversification. As marine carriers increasingly leverage analytics tools to improve their risk selection and minimize exposure, participants in this roundtable shared their views and experiences on how data-driven decision-making drives underwriting profitability and opens up new frontiers for marine insurers.

The session kicked off with Tom asking participants how they currently harness data and analytics tools within their organizations. One panelist said that addressing and fully supporting the human element is the number one priority at their organization, ensuring that the talent using these tools are adequately educated and trained. The same participant mentioned that altogether “data aggregation and technology have helped us to become more precise underwriters and allow us to eliminate some of the subjective factors and rely more on objective analysis.”

As the conversation turned to the current economic climate, panelists discussed how distilling the correct data in real-time is fundamental to driving down combined ratios and getting better at managing losses and expenses. Commenting on this, one participant declared: “The data is coming at us in ever growing quantities and sifting through that is important; we need to ask the right questions, where should we be asking more questions and where can we afford to ask fewer questions.”

Data sharing

Last month, Tom moderated a similar roundtable at the Marine Insurance Nordics conference in which participants were divided over the value of data sharing amongst carriers. When the same question came up during this discussion, most participants agreed that data sharing would benefit the industry, with one participant noting that: “the marine industry probably does a better than average job of sharing information across whether it would be carriers, brokers or customers… the more information we all have the more we´ll be able to understand and quantify that and ultimately drive down costs and improve services we provide to our customers.”

The panel discussed the emergence of cutting-edge technology and whether there were any specific types of technology that enhance risk selection vs. plain noise. One participant suggested that it´s not so much the type of technology that makes the difference, but spending time with underwriters and data scientists to ensure that they are mining the right type of data and asking the pertinent questions, saying: “I spend time with some of our data analysts and its incumbent upon on me to be able to educate them on what we do, what are the risks that we’re concerned with, what are the issues… the more they understand, the more they’re able to look at the data and glean inferences vs. going down paths that maybe are ill-advised or incorrect.”

Continuing on the theme of internal education, one panelist raised the point that striking the balance of having the right data and the right people looking at that data is a challenge, saying that some brilliant data scientists lack domain expertise. Commenting further on this, one participant said: “I sit down with data people and sometimes I’m told ´well you can’t wish for that´ and yeah it’s like I can, I still believe in Santa Claus, I can still wish for that.” Another participant maintained that “as long as we continue to focus on keeping the subject matter experts working with the data analytics people, we can have success.”

Build vs. buy

The roundtable also discussed the merits of build vs. buy. One participant said that their company experimented with building their own software but now favors buying software from technology vendors as the level of configurability available in today’s insurance software products provides them with the ability to customize the software to their needs while also allowing them to focus more time on the customer. However, another participant said that their company prefers developing its proprietary software rather than buying it because they don’t have a single focus and their in-house team is agile and small enough to be able to design solutions, commenting: “data literacy is very important, making sure that the specialist knows the data you’re gathering, how you’re gathering and what it actually represents.”

Wrapping up the session, Tom asked participants for their views on how they see the industry evolving over the next few years. One participant responded by saying: “I definitely anticipate and foresee new markets and incumbent markets for that matter leaning more into the data in terms of becoming more of a factor for new insurance product line launches and in fact embracing that data much more regularly than ever before.” The same participants expanded further by stating that one way companies can harness and leverage new data would be through cross-selling purposes.

Looking to the future, one participant disclosed that their focus is on making their products more accessible to partners, driving out friction, increasing automation for internal efficiencies and looking at the value chain, and making life easier for their brokers. Acknowledging the training needed and the day-to-day use of data while they work towards using more predictive analytics were also critical areas for this participant.

Another panelist declared: “I do think that we’re going to have to raise our game, and more of us are going to have to look at things as more portfolio managers than as individual underwriters.” Commenting on the underwriter of the future, one participant shared the view that underwriters will have to become more adept at using granular data to have a better single risk view rather than a portfolio view.

Latest Insights