Optimise your portfolio management with strategic underwriting leveraging AI

Underwriting is the starting point for risk, but not the end point. To keep ahead of the competition, underwriters need to capitalise on artificial intelligence (AI) to level up and become a strategic, data-driven function able to quickly understand, assess and manage the concentration of risk across portfolios and feed into the business to create value and drive revenue-generation.

Simone Bohnenberger-Rich, senior vice president of product at Eigen Technologies, joins a panel of experts for our new webinar hosted by Intelligent Insurer: “ Optimize your portfolio management with strategic underwriting leveraging AI” on October 19, 2021. She believes that data rests at the heart of portfolio management. “Access to comprehensive and granular data which allows managers to slice and dice their portfolios by risk profiles, exposures, and channels is a competitive advantage and separates top performers from the rest of the industry,” she says.

With the panel she will debate how to:

Joining Bohnenberger-Rich is Guenter Kryszon, executive underwriting officer, global property at Markel; Claudio Morando, head of portfolio analytics and strategy (EMEA) at Swiss Re; and Will Roscoe, head of alternative portfolio underwriting at Beazley. The moderator is Marisa Murton, director of Solvlab Analytics.

Bohnenberger-Rich talked to Intelligent Insurer ahead of the webinar and stressed that portfolio managers need to answer a range of questions. For example, they need to consider why certain lines or segments and regions outperform others. They should also analyse the impact of policy language on the loss ratio, and they need to consider whether they are underwriting the best risk profiles and pricing them correctly. She also talked about how they can optimise their portfolios accordingly.

What are the current trends for optimising portfolio management?

Often portfolio managers can’t answer these questions effectively due to a lack of data accessibility, quality, and granularity. While there is a growing trend of incorporating external, alternative data sources to provide a more complete analysis of risks, most organisations struggle to utilise the unique dataset they already possess and sit on a wealth of untapped potential. Insurers today must understand and analyse traditional policy and claims data with non-traditional tools to transform raw information into insights.

In their search for technologies that can gather and make sense of this internal data, carriers have used rules-based technologies such as robotic process automation (RPA). This has proved successful in collecting data that is easily accessible and regularly templated. Carriers have also improved their digital processes, structuring data at source in high volume and heavily standardised retail lines, bypassing the data gathering problem.

However, for the high value, non-standard risks that dominate commercial P&C, most information today is largely unstructured and embedded in context such as policy agreements, engineering reports, emails, etc. Manual data processes simply aren’t scalable here and therefore carriers capture and analyse the bare minimum, leaving a lot of valuable data on the table. Next-generation, AI-driven technology such as natural language processing (NLP) is required to harness this data.

Underwriters can comprehensively digitise data during the submission-quote-bind process to improve underwriting insight and efficiency, giving them the edge in an increasingly competitive landscape. Portfolio managers and risk managers can leverage this same data to stay on top of new risk dynamics, and ultimately make more informed strategic decisions.

Take for example, complex, rapidly changing risk profiles such as pandemics, climate catastrophes, insurer downgrades, political instability, cyber-attacks, etc. This requires quick exposure assessment and portfolio optimisation. However, information needed to do so is not readily available as it is buried in policies or contracts and has not been captured at source. NLP can digitise risk data at scale as part of routine operational processes or as part of urgent, ad hoc reviews, providing carriers with the right data at the right time.

How can insurers transform underwriting into a strategic hub of value creation?

Underwriters are often overwhelmed with the volume of data that pertains to risks and needs to be analyzed (eg, submissions loss runs, engineering reports, previous year policies, etc). There is a constant need to strike the balance between thoroughness of analysis and speed of response. Subsequently, many portfolio managers prioritise the latter given the competitiveness of the market, capture and leverage only a small fraction of this information.

This accumulated dataset is unique to each carrier—a key example of ‘data as a differentiator’—but the vast majority remains unused as manual methods of structuring the data are unfeasible. Smart technologies such as NLP can digitise relevant risk data at scale and unlock this untapped potential. The underwriting function can act as a source of valuable data and insights that can be leveraged across the firm.

Improving the efficiency of data compilation using smart technology can lead to increased revenue generation through improved underwriting efficiency, insight, and capacity; better risk management through more granular data processing and better quality data (eg, internal audit checks); and cost reduction through greater operational efficiency of administrative processes.

What risks should insurers be aware of when adopting AI-driven technologies?

Smart technologies such as AI are often perceived as “black boxes” associated with a lack of control over outputs—it’s the fear of new algorithms running amok. Here it is vital to deploy platforms that have built-in “human in the loop” workflows.

For example, the technology clearly flags to the user where it may have missed information or where it doesn’t have the right information so “humans” can review this data and correct it before it is fed into downstream systems. Advanced technologies provide clear audit trails that allow users to trace what inputs the system had and why it created the outputs it did.

What key points would you like delegates to take away from the webinar?

Simone Bohnenberger-Rich, senior vice president of product at Eigen Technologies, is speaking at our new webinar “ Optimize your portfolio management with strategic underwriting leveraging AI” (Tuesday, October 19, 2021). Register here to learn what smart document technology enables, and how and why AI enables the optimisation of portfolio management.