WTW and Klarity are working together to help life insurers improve pricing accuracy by integrating wearable technology and data into their underwriting processes.
Executive summary
“Sitting is the new smoking” is a catchphrase often used to encourage people to get some level of physical activity. Medical personnel, underwriters, actuaries and mortality researchers understand activity level is an important measure to assess one’s health and expected longevity. Unfortunately, activity level information is often overlooked or has been measured only through self-reporting or correlation to other measures such as body mass index (BMI) in the life insurance risk selection process.
Shortcomings of traditional risk stratification approaches
Since the proliferation of multiple risk classes, companies have used traditional measures such as cholesterol level, blood pressure, BMI, tobacco usage, and personal and family history, to name a few, for stratifying and determining risk class criterion and placement for applicants. While each is an important health metric, these traditional approaches and metrics often misclass applicants because the measures only provide part of an individual’s health profile and often miss important individualized measures such as resting heart rate, heart recovery rate, sleep and activity versus inactivity levels.
Over the past 10-plus years, the industry has been moving toward changing the underwriting process. For life insurance, this has meant rethinking the data sources used, improving the customer’s experience and shortening the time from application to policy issue. This has added to the challenges for risk classification and difficulty in truly differentiating the risk profiles of the preferred risks as well as the healthier impaired.
The Klarity risk scoring model
The need to rethink the risk stratification process in the life insurance industry has become increasingly evident over the past decade. With the proliferation of new data sources and advancements in technology, there is a significant opportunity to enhance the accuracy and efficiency of underwriting processes. The Klarity model aims to address this need by leveraging nontraditional data to produce individual-level mortality scores that can predict and classify risks more effectively than traditional methods.
WTW’s analysis of the risk scoring tool
Over the past year, WTW’s Insurance Consulting & Technology team has analyzed a new risk scoring tool developed by Klarity, which incorporates data obtained from a wearable device such as a fitness watch, a smartphone or other device that captures activity levels, sleep patterns, heart rate and pulse data.
Key observations of WTW’s analysis show:
The Klarity model has the ability to better classify risks and reduce the overlap inherent in today’s risk classification systems. In our analysis, when activity level and the Klarity model risk score are considered:
34% of the second-best nonsmoker risks and 16% of the residual standard class risks are identified as having a better risk score and exhibit similar mortality risk profiles to the best nonsmoker risks
6% of risks currently classified in the best and 13% classified in the second-best preferred nonsmoker classes are identified as having a lower risk score and exhibit actual-to-expected (A/E) ratios more akin to a residual (not preferred) risk
The level of activity, including Step Count and Activity Duration, provide high correlation to mortality outcomes, even more than some traditional markers such as cholesterol, BMI, and family history of heart disease and diabetes
Though trained mostly on U.K. data, the Klarity model and model risk score are effective when applied to a U.S. population, and the model will continue to get stronger with increased U.S. insured data over time
Stratification improves even among those who would be classified under the same risk classification using traditional underwriting methods
There are many benefits to utilizing the Klarity risk scoring model, such as:
Improved stratification of risk and ability to improve the accuracy and assignment of risks to risk categories
Enhanced triaging or streamlining the triage process or risk class quoting to improve consumer and applicant expectations
Improved accuracy and selection of the healthier impaired risks
Reduced friction in the data collection process for measuring and analyzing mortality and longevity risk
In-force engagement
Potential for new products and pricing strategies focused on healthy longevity
Benefits of the Klarity model
The Klarity model demonstrates a promising approach to improving risk stratification in the life insurance industry. One of our key findings is that while the Klarity model validates the risk ranking of traditional underwriting classes used to assess mortality risk by insurers, the Klarity model can improve and further differentiate risks significantly even within a single risk class. We found the risk assignments by the Klarity model to be highly correlated with actual mortality results relative to mortality demographic baselines. By utilizing a broader range of data sources and advanced predictive techniques, the model can provide more accurate and individualized risk assessments, ultimately leading to better pricing, improved customer engagement and enhanced in-force management.
The white paper outlines WTW’s analysis and findings; it’s structured as follows:
Background — This section provides an overview of the current state of the life insurance industry, highlighting the challenges and opportunities presented by new data sources and methodologies
Model — This section details the Klarity model, including its data sources, methodology and the results of its application to the U.S. NHANES data set
Use cases and next steps — This section explores potential use cases for the Klarity model and outlines the next steps for its implementation and further development
About Klarity — This section provides background information on Klarity, the company behind the model
About WTW — This section offers insights into WTW, the global consulting organization that collaborated on the review and testing of the model
