June 14, 2021 | By Matthias Weber.
Insurance companies need powerful analytics to translate data of the past into an understanding of the future. Since data can’t do it all, I emphasize good corporate governance which is essential as it lays the foundation for future growth.
Insurance has always been a data rich business. Data is used for risk selection, identification of fraud, or personalized and contextualized insurance distribution. Some of the data is proprietary and unique which provides a competitive advantage for Insurtech companies.
Derive understanding of the future from data of the past
Data derived learnings are often relevant for the future. Insurance related examples include:
- Policies that cover losses over the next 12 months.
- Lead generation that reflects the needs of future potential insurance buyers.
- Insurance fraud specialists who try to prevent future fraud.
However, collected data, by definition, reflects the past. And the past is not always a good predictor for the future. The risk of change impacts insurance even more profoundly as an insured event may result in a loss payment more than 45 years after the subject policy was written.
Therefore, insurance companies not only need data, but also powerful analytics to translate data of the past into an understanding of the future. The example of RMS immediately comes to my mind. The company recently announced the launch of a new suite of models to help customers assess near- and long-term impacts of climate change on physical assets and businesses. RMS states that the underlying analytics reflect the best forward-looking climate science consensus, including the one from the Intergovernmental Panel on Climate Change (IPCC).
Finally, if there’s anything I’ve learned in insurance, it’s that data and analytics must be combined with human judgement! After all, data can be incomplete, wrong, or fake. “All models are wrong but some are useful” according to the British statistician George E. P. Box. So examine both the data and the model yourself. See if they make sense. If not, ask why.
Build in transparency for risk assessment models
The ability to quantify insurable risk is a critically important core competency for insurance companies. Notably, I’m watching one Insurtech niche that offers software to assess risk faster, more precisely, and at lower cost.
These tertiary providers must demonstrate a high level of transparency on what happens inside their models. Any “black-box” solution might make the insurer feel like they are blindly outsourcing a core business competency.
Risk assessment analytics developers are tempted to provide limited modeling transparency to protect their intellectual property. However, the need for transparency is strong and ubiquitous. That’s why Oasis, an open-source modeling platform for natural catastrophe risks, is backed by the majority of the large global insurers, reinsurers, and reinsurance intermediaries.
Providers of insurance risk analytics should move away from the “black-box” approach and find alternatives to build up competitive advantage. I see an opportunity in securing a continuous flow of new, proprietary, and differentiating data that feeds into their models.
Insist on good corporate governance early
I conclude my three part blog series with a word about corporate governance in Insurtech. In our technology driven world, governance is sometimes overlooked which prevents promising startups from maturing into solid investments.
Corporate governance – the system of rules, practices and processes by which a company is managed and controlled – may not be super exciting. Some might even feel it curbs creativity. However, solid corporate governance builds trust with investors and creates a risk reducing culture. Since most insurance companies, agents, or reinsurance intermediaries perform highly at governance, they expect the same from their business partners.
Start-up companies are advised to establish good corporate governance as early as possible to have a sound foundation in place before scaling up. This is especially important for Insurtech brands as they often deal with massive data sets, some of which include personal information subject to data privacy and security laws.
Such personal information may include:
- The exact whereabouts of somebody’s car.
- Lifestyle choices, such as eating or smoking habits.
- Data that can help assess an individual’s health outlook.
- A score predicting an individual’s likelihood of having committed insurance fraud.
Given the sensitive nature of insurance related information, Insurtech entrepreneurs must make absolutely sure that the data entrusted to them is safe, trustworthy, and does not get misused.
Mighty Capital Partner, Matthias Weber is the former Group Chief Underwriting Officer of Swiss Re and current Board Member of Next Insurance US Company and CyberCube Analytics. With close to 30 years of activity in the insurance sector, his insight and guidance now help emerging brands and entrepreneurs to capitalize on new opportunities in Insurtech.