Bitesize InsurTech: Cytora
September 8, 2017 Chris Sandilands
Cytora uses artificial intelligence and external data to improve the way commercial insurers quantify, select and price risk.
The business is an unlikely product of the Arab Spring. The founders, who have backgrounds including InsurTech (eBaoTech in China) and finance (World Bank) noticed that insurers had suffered substantial losses during the Arab Spring due to their inability to factor real-time information into their risk assessment models. Cytora was founded on the belief that online data sources such as social media, newspapers and government reports, can be used to understand and manage both traditional and emerging risks.
Originally spun out of the University of Cambridge, Cytora has developed a risk engine that continuously ingests data from thousands of online sources, and transforms it into risk intelligence for the commercial insurance industry.
There are two main use cases.
First, Cytora helps commercial insurers to improve their loss ratios by differentiating between good and bad risks at the point of underwriting, and issue granular technical prices. The company claims that backtesting showed that one UK insurer could have reduced its commercial property loss ratio for the previous six years by 33 percentage points if they had used Cytora’s insight at the point of underwriting (with a 10% reduction in GWP).
Second, Cytora helps commercial insurers to grow premium in new areas by enabling them to identify profitable segments. Because Cytora builds intelligence from sources which are not insurance dependent, they have broad coverage of a segment or geographic area – broader than the leading insurer who is restricted to its market share. Cytora therefore has better insight into new business applications than any individual insurer.
The Oxbow Partners view
In SME the case study we referenced suggests that the business has potential to create competitive advantage for its clients. We had a question mark about whether Cytora could replicate the model outside commercial property (where data could be trickier to find) but Richard Hartley, CEO, tells us that they are well advanced towards launching a BI/EL product in early 2018. We see great potential here.
Our second thought concerns the corporate segment. Data businesses like Cytora target personal lines and SME because there are sufficient risks to calculate granular risk prices. Corporate property is generally rated much more crudely than personal lines / SME. We think there could be an interesting opportunity for Cytora to help corporate insurers build a better understanding of multi-location businesses (e.g. restaurant chains), and thus move to more technical pricing. Public information about locations could be augmented with data mined from risk engineering reports (e.g. through natural language processing, see Risk Genius Bitesize). That could be really interesting.
Incidentally, this business is quite similar to Tyche, which we covered in Bitesize InsurTech in March 2017.