Dublin based insurtech start-up Describe Data helps insurance underwriters develop a better understanding of risk, enabling them to price it properly and avoid the bad deals. “In the insurance industry there is a saying that there is no such thing as bad risk, just a bad price,” says Describe Data Chief Operating Officer Gerard de Vere.
The company uses data analytics to provide insights into financial lines risks such as directors and officers, employers practice liability, intellectual property, and mergers and acquisitions. “Our particular focus is on the directors and officers line,” says Chief Executive Michael Crawford. “This is insurance against companies being sued and provides protection for their directors and officers. Fifty per cent of the world market for that cover is in the US, due to the existence of class action suits there as well as the country’s highly litigious culture. We gather data on companies and their sectors and analyse it to help insurance underwriters assess the risks.”
This particular type of insurance is very specific to large corporations. “We are talking about very big risks,” Crawford adds. “The policies are being written by underwriters who analyse the risk in a qualitative way. We are putting in a quantitive overlay which helps them price the risk better and do it quicker.”
The insurance industry is not noted for its early embrace of technology, however. “That can be a challenge,” says Crawford. “The industry has been around for 300 years and it never really felt that it needed technology while there was still plenty of money to be made.”
But it couldn’t resist forever. “It has been happening in capital markets for the past 30 years,” says Crawford. “And we have seen technology and data analytics being applied to catastrophe insurance for quite a long time. A lot of those tools and techniques have matured and the cost of computational power has fallen through the floor. We now have access to computing power that would have been the preserve of large academic institutions or governments 15 years ago and we can deploy these advanced techniques and tools at an affordable cost. At the same time, open-source software has been developed that allows people to share the latest solutions. It’s been a perfect storm.”
The Describe Data product is a risk engine which produces insights across four levels. The first is the financial level where it gives financial information on the company and its sector. The next is litigation and the likelihood of lawsuits against the company or within the sector.
The risk evaluation level looks at what happens if different aspects of the risk are adjusted. “The underlying market for risk is highly sophisticated,” de Vere explains. “People in the industry trade bits of risk with each other all the time and it is possible to reduce exposure in this way.”
The final level is portfolio analysis. A company may have written hundreds of these policies and the system can slice and dice the portfolio to find risk hotspots. For example, companies in the same sector, or sectors with high levels of litigation, or companies with a poor track record, can be identified for the insurer either to increase premiums or cease cover at the next renewal date.
“It’s quite a dynamic market,” says de Vere. “We have met a lot of people in London who are very savvy at this sort of thing. They have an innate skill and are very good at it. We put technology in to make it better, quicker, stronger. We call it bionic underwriting.”
That emphasis on supporting the people in the industry is very important. “People come in with robotic process automation and try to say it will replace underwriters,” Crawford notes. “But you can’t replace 300 years of underwriting experience and the instincts and innate knowledge that comes with that. What we are offering is intelligent decision support tools which help people make better decisions faster.”
Crawford explains the value of these tools. “If a company takes 100 risks, four or five might claim. Three of those will be quite minor, one will be quite large, and one will be huge. We offer the ability to understand the risks and identify the bad ones. Avoiding the one with the very large claim can transform a portfolio from a margin of 5% or 6% to 20% or 25%.”
The company is now marketing the product following a delay caused by the Covid-19 pandemic. “We spent a couple of years developing the product after we started up in 2018,” says de Vere. “Towards the end of last year, we were looking for a large insurance company to partner with to take the product and prove it. We had a couple of promising leads and then the pandemic hit and the appetite for a pilot disappeared. We decided to double down and continue product development, and we completed that work during the lockdown. We now have a product ready to go to market and we are talking to a number of potential clients.”
Looking ahead, Crawford says the company will probably go to the market for a funding round next year. “We have been self-funded so far and will probably take on investment at the beginning of next year to sustain our growth. Enterprise Ireland has been amazing in helping us. We want to continue research and development and move into other lines of insurance. It’s a very interesting time to be doing this.”