Swiss Re: Advanced analytics enabling insurers' profitability and growth

Swiss Re: Advanced analytics enabling insurers' profitability and growth | Insurance Business

Swiss Re: Advanced analytics enabling insurers

In a challenging and highly competitive business environment, insurance companies around the world are seeking ways to improve profitability and enable growth. According to Swiss Re Institute’s latest sigma, property and casualty (P&C) insurers are increasingly looking to advanced analytics to unlock tangible value and gain a competitive advantage. The general theory is: those who use advanced analytics to gain insights into good customer behaviour, identify early signals of emerging risks, and optimize their business operations will have the best chances of success.

Insurers worldwide are now using behavioural science and predictive modelling to engage customers and improve customer retention, especially in lines of business where rates are hardening and markets are reducing their capacity. For example, one of Swiss Re’s United States clients - Anchor General Insurance, a West Coast US insurer that specializes in non-standard auto business – recently enlisted Swiss Re’s help in using advanced analytics to better understand what drove policy lapse after a rate change. Together, the two firms used behavioural economics driven by advanced analytics to reconsider Anchor General’s messaging so that it might better resonate with clients. This resulted in a 28% increase in retention rates among the policies subject to a rate change.

“Advanced analytics can also improve portfolio optimization,” explained Kassie Bryan, senior vice president, head solutions P&C Americas, Swiss Re. “We had a UK-based insurer who wanted to determine whether their commercial property book’s losses were the result of a short-term deviation, or a fundamental change. Advanced analytics helped this insurer benchmark its portfolios, loss frequency and severity against its peers, and it found its losses were driven by higher than expected losses in eight business segments. As a result, the insurer went about shrinking those segments, while at the same time expanding the six profitable segments identified. By implementing these changes, the insurer reduced its commercial property loss ratio by eight percentage points. They were also able to grow premium in the six target segments by about 8% annually.”

The availability of data, particularly from public sources, is continuing to increase. As insurers identify a need for advanced analytics, they’re looking for ways to transform unstructured data sets into structured information that can provide insights into individual risks or entire portfolios. In order to achieve that goal, insurance companies need “a strong data strategy,” according to Bryan.

“This depends on two aspects: legacy tech infrastructure, and data acquisition,” she told Insurance Business. “Legacy tech infrastructure is a major bottleneck for data ingestion and curation. There are large P&C carriers out there who can access better data with some predictive value, but it will take some significant time and effort to ingest and curate that data if they have an older tech stack. Meanwhile, there are next generation carriers who have a more modern tech infrastructure, who can quickly ingest these new data sources to test it and evaluate whether they want to use it or not.

“The second piece, along with how quickly a customer is able to ingest the data with their tech infrastructure, is around data acquisition. We think insurers should have a strategy to quickly and cheaply assess the effectiveness or efficacy of new data sources. Ideally, the insurer should start with a phased approach, meaning they start with a few intuitive data points to assess predictive power and correlation of data. Where the correlation is successful, then insurers can look to explain results in a clear and concrete manner, before buying additional data. But insurers with resource constraints can even avoid investing in data curation entirely, and buy highly condensed data with predictive features to input into their models.”

The opportunities for insurance companies are endless where advanced analytics is involved.