Emerging Trends in Data and Insurance Underwriting
Since the advent of the modern industrial era, the most successful enterprises have leveraged spatial analytics, or data on the space and workstreams of our physical world, to upend industries. In the early 20th century it was Ford using data points to find a better way to manufacture high quality and low-cost cars — this was only refined by Toyota’s TPS system later in the 20th century. Now well into the 21st century, just what vertical applications are best suited to take up the baton and execute on the opportunity?
Immediately FinTech comes to mind based on the number of platforms utilizing data sets of all types to underwrite risk. Whether it be commercial or consumer loans or insurance policies, the more data a business has to assess risk, the more efficiently the platform can price its products, attract customers, and reduce write-offs. Within FinTech, insurance as a $5T global category has already seen traction in applying spatial data to practice. While data geeks have dominated the insurance industry for decades, the diversity of data sets used in incumbent models is surprisingly Neolithic. For instance, traditional underwriting algorithms only take into account a few dozen data points relative to the treasure trove of real time sets of spatial data around us. That said, new models are emerging to augment underwriting with data from the physical world, a notable example of which is Root Insurance, a full stack carrier employing dynamic driving behaviors to underwrite policies. Though founded just five years ago, the company recently IPO’d at a multi-billion-dollar valuation, largely attributable to growth associated with its novel underwriting capabilities. Root is not alone in this practice either as incumbents and emerging InsurTech platforms across lines are now unlocking billions in value via enhanced underwriting tools augmented with data sets from the physical world. The value of this type of information is even more topical when considering the substantive blind spots of legacy insurer underwriting models. Take that Consumer Reports findings suggest certain insurers charge minorities premiums as much as 30% higher than their white counterparts in areas with similar accident costs. Despite the historical correlation of zip codes to loss ratios, this underscores the importance of leveraging additional data points for a more holistic understanding of an individual’s and/or property’s risk profile.
All that said, how is the category evolving to address these blind spots? For the purposes of this discussion, the application of spatial data, be that via wearables or satellite imagery, can largely be segmented under two major umbrellas within the broader insurance sector — Property & Casualty and Health & Life lines. Beyond how to extract spatial data, though, is the question of how to best apply it in dynamic underwriting.
Health & Life Insurance (“H&L”)
Relative to other insurance lines, Health & Life lines are further along in their integration of spatial data into underwriting models. Historically, H&L insurance policies have been largely measured by examining medical histories of applicants. While this information is an important part of the picture, it does not accurately capture a full risk profile and it certainly lacks any real time data points when considering the average American sees a doctor of any kind just four times a year. To fix this gap, wearables have emerged as a natural source of real time updates for dynamic underwriting, most recently on the heels of the boom in fitness tracking devices such as Fitbit and Apple watches. Incumbent insurance players are moving fast to adapt too; in 2018, John Hancock announced it would stop selling traditional life insurance and instead only market interactive policies that record the exercise and data health of its customers through wearables. United Healthcare followed suite by adding the Apple Watch to its United Healthcare Motion program. Munich Re even recently released a study on the effectiveness of physical activity as measured by wearable sensors in profiling mortality risk of a U.S. population-based dataset. This is most important in the context of cost-containing, preventative treatment where an estimated 60% of high-spend members under coverage weren’t high cost the prior year. As a part of this study, Munich Re also uncovered that in addition to real time tracking and optimized insurance pricing, it also benefitted from continuous engagement via wearable devices as well as expanded insurability for underrepresented groups who otherwise would have been declined. All of this underpins the rapid growth of real time data sets with IDC estimating there will be 41.6 billion total IoT devices worldwide by 2025, enabling new channels of communication, engagement, and information sharing in the process. With the game changing health insights coming from spatial data, the value attributed to the global wearable healthcare device market, which is expected to surpass $29 billion by 2026, should not be surprising.
Property & Casualty Insurance (“P&C”)
Within the broader $1.5T Property & Casualty market, Auto lines have already ramped up use of spatial data via telematics and usage-based insurance for the purposes of risk modeling. In addition to underwriting, though, dynamic data also enhances other areas of the insurance value chain. Take for instance first notice of loss (“FNOL”) — while historically FNOL has required a slew of forms, phone calls, and from personal experience, tears, the filing process is being transformed with real time information extracted from IoT sensors, drones, and satellites. Case in point is CSAA Insurance’s partnership with Owlcam to send videos to a driver’s mobile phone when a car crashes or is broken into. Because of the availability of real time information facilitated via satellite, all parties now benefit from both a living communication channel and a reliable source of truth in determining loss outcomes.
Beyond Auto though, P&C also encompasses the enormous opportunity in Property lines which counts the industrial, commercial, and residential asset classes within its ranks. Like Health lines, adoption of spatial data for Property is quite verticalized with many technologies built with specific applications in mind. For instance, Cape Analytics, a software platform using aerial analysis via satellite imagery, focuses on evaluating roofing conditions such as size and age for home risk modeling. Along these lines, startup True Flood provides insurers with property level information on homes throughout a flood plain to evaluate structural durability of say a home’s foundation. Looking forward, many startups are now emerging to address global warming with preemptive, risk assessing technologies to manage growing unpredictability of property damages. These technologies are relatively nascent, particularly when considering the long sales cycles of frontier technologies into incumbent insurers and reinsurers.
Now to break this all down. With a massive market opportunity at hand, which innovative solutions are capitalizing on spatial data applications in insurance underwriting? To help navigate, the below graphic maps out select companies providing both the infrastructure needed to extract, organize, and analyze spatial data, including wearables and satellite-based aggregators and the InsurTechs utilizing this data for underwriting. Note, the chart is segmented such that applicable lines of insurance are distributed horizontally.
A takeaway here is that while the most familiar names in InsurTech such as Lemonade, Root, and Oscar function as full stack carriers, a sizable opportunity also exists in the infrastructure supporting them. As full stack and MGA models utilizing these data sets have scaled, so has the visibility of spatial data used in underwriting, and incumbents have moved to quickly catch up via investments, partnerships, and acquisitions. Over the last few years alone, American Family’s investment in Avinew and Teraki and State Farm’s investment in Cape Analytics speak to the emerging opportunities across lines on the P&C side. Meanwhile for H&L lines, investment has largely been focused on technologies targeting specific ailments or demographics such as New York Life’s investment in Carrot or TransAmerica’s investment in 100Plus. The largest outcomes, though, will go beyond specific applications and provide the most versatile infrastructure to augment risk modeling across lines.
With increasing availability of spatial data through IoT technologies, there will emerge winners and losers across industries. However, in InsurTech, the challenge lies not in the collecting of the data, or how this data can be useful, but rather which sets of spatial data will be most effective in proving out risk. In a multivariate equation, only time and investment will help discern where the highest impact data points by line lie. That said, to ignore the opportunity in spatial data would be to start a lap behind competitors in a land grab for market share. And with customer acquisition costs only increasing across the insurance landscape, the ROI of data that provides customization, pricing efficiency, and increasing engagement with customers will only become more important in driving winning outcomes.
Disclaimer: Views are my own and may not reflect not those of my employer.
McKinsey’s “State of property & casualty insurance 2020”
Consumer Report’s “Minority Neighborhoods Pay Higher Car Insurance Premiums Than White Areas With the Same Risk”
Institute and Faculty of Actuaries’ “Wearables and the internet of things: considerations for the life and health insurance industry” by A. Spender*, C. Bullen*, L. Altmann-Richer, J. Cripps, R. Duffy C. Falkous, M. Farrell, T. Horn, J. Wigzell and W. Yeap
S&P Global Market Intelligence
Munich Re’s The Future Is Now: Wearables for Insurance Risk Assessment authored by June Quah
Transparency Market Research (TMR)